Value and examples

Check the math before you commit.

Estimate what AI could be worth to your business across five levers, from team time to pricing. Then walk through six industry scenarios and a worked example for each service, with every assumption on the page.

Run your numbers See the examples

Every number here is a modeled estimate or an illustrative example. None are client results.

01Value calculator

Start with your own numbers.

Your numbers, lever by lever

The defaults describe a typical 50-person business with about $12 million in revenue, 25 of them in desk or customer-facing roles. Tick the levers that fit your business and open any of them to put in your own numbers.

Agents take over repetitive desk work: routine email, data entry, reports and scheduling. Your people keep the judgment calls, and each person is capped at the hours a week the research supports.

People in desk or customer-facing roles

Office, admin, customer service, sales and management staff. Leave out roles that are almost all hands-on, like field technicians or kitchen staff.

Average loaded cost per hour

Hourly pay plus payroll taxes and benefits. If you only know the wage, multiply it by 1.4. $30 is the typical loaded cost of routine office work at a business with 50 or more staff.

Hours per person each week on repetitive tasks

Routine email, data entry, copying between systems, scheduling and follow-ups. Leave out time spent answering customers; the customer lever counts that.

Value of staff time, a year: $74,880 to $157,320
Hours back a month: 208 to 437 h
Full-time capacity: 1.3 to 2.7 people

See the mathModeled estimate

The math behind the Team time range, low case to high case
Repetitive work agents take on, after 20% human review	7 h × 50% to 75% × 80% = 2.8 to 4.2 h a week per person
Evidence ceiling per person	2 to 4.9 h a week (base case capped) → 2 to 4.2 h
Team hours back a month	25 people × 2 to 4.2 h × 50 weeks ÷ 12 = 208 to 437 h
Value of that time a year	208 to 437 h × 12 × $30 = $74,880 to $157,320

The base case uses the base rates in our research and the upside the optimistic ones, never beyond them. Each step is rounded down. Sources: lab-01 · lab-06 · ai-15 · ai-21 · ai-22 · time-01 · time-05 · ai-20

The Customer Agent answers calls, chats and forms in seconds, day or night, so fewer new jobs go to whoever picks up first. It also resolves routine questions without your staff.

Customer inquiries per month

Calls, chats, emails, texts and web forms from customers and prospects.

Share that arrive after hours or go unanswered

Inquiries nobody answers within a few minutes: evenings, weekends and calls that ring out on busy days. One large booking platform saw 48% of online bookings made between 5 p.m. and 9 a.m.

Average job or order value

Revenue from a typical new job, order or first visit. This is revenue; only your gross margin on it counts toward the total.

Gross margin

Revenue minus the direct cost of delivering it, as a share of revenue. New jobs and deals count only at this margin, and the pricing and vendors lever uses it to size your cost of sales.

Gross profit and staff time, a year: $36,720 to $103,824
Added revenue, for context (not added to the total): $82,800 to $236,160

See the mathModeled estimate

The math behind the Customer inquiries range, low case to high case
Inquiries after hours or unanswered	800 × 30% = 240 a month
New-business opportunities among them	× 30% to 38% = 72 to 91.2
Lost for good without an instant answer	× 40% to 60% = 28.8 to 54.7
Booked when the agent answers at once	× 40% to 60% = 11.5 to 32.8 jobs a month
Added revenue a year (context, not in the total)	11.5 to 32.8 × 12 × $600 = $82,800 to $236,160
Added gross profit a year	$82,800 to $236,160 × 40% = $33,120 to $94,464
Routine inquiries resolved without staff	800 × 20% to 35% = 160 to 280 a month
Staff time back, after 20% review	160 to 280 × 5 to 7 min ÷ 60 × 80% = 10 to 26 h a month
Value of that time a year	10 to 26 h × 12 × $30 = $3,600 to $9,360

The base case uses the base rates in our research and the upside the optimistic ones, never beyond them. Each step is rounded down. Sources: cx-27 · cx-23 · cx-26 · cx-21 · cx-25 · cx-24 · cx-16 · cx-09 · cx-10 · cx-34 · pnl-16

The Marketing Director audits every campaign, tests offers and audiences, and moves budget to what converts, within limits you set. The same leads cost less, which frees budget to put behind what works.

Monthly ad spend

Google, Meta, Microsoft and other paid media. Leave out agency fees and creative costs.

Budget freed to reallocate, a year: $12,000 to $24,000
Or, at the same budget: 11% to 25% more leads

See the mathModeled estimate

The math behind the Ad spend range, low case to high case
Ad spend a year	$10,000 × 12 = $120,000
Same leads for less: cost per lead down	10% to 20% after 60 to 90 days of testing and reallocation
Budget freed to reallocate, a year	$120,000 × 10% to 20% = $12,000 to $24,000
Or keep spending the same	about 11% to 25% more leads instead (not added on top)

The base case uses the base rates in our research and the upside the optimistic ones, never beyond them. Each step is rounded down. Sources: mkt-19 · mkt-25 · mkt-24 · mkt-15 · mkt-16 · mkt-14 · mkt-01

The Sales Agent researches each prospect and writes the email a thoughtful person would, then follows up. Cold email replies are rare, so we use measured reply rates, not vendor claims, and count only a deal’s first year.

New prospects contacted per month

Businesses that fit your ideal customer, each researched and emailed by the Sales Agent. We plan about 300 a month per sending mailbox to protect deliverability.

Average first-year deal value

What a new account brings in during its first year. We count only year one, even when accounts renew.

Gross profit on new deals, a year: $6,000 to $24,000
Added revenue, for context (not added to the total): $15,000 to $60,000

See the mathModeled estimate

The math behind the Sales outreach range, low case to high case
Prospects researched and emailed a year	1,000 × 12 = 12,000
Positive replies	3.5 to 6 per 1,000 = 42 to 72
Meetings held	× 20% to 25% = 8.4 to 18
Qualified opportunities	× 40% to 45% = 3.3 to 8.1
New deals won	× 20% to 30% = 0.6 to 2.4 a year
Added revenue, first year (context, not in the total)	0.6 to 2.4 × $25,000 = $15,000 to $60,000
Added gross profit a year	$15,000 to $60,000 × 40% = $6,000 to $24,000

The base case uses the base rates in our research and the upside the optimistic ones, never beyond them. Each step is rounded down. Sources: sales-07 · sales-08 · sales-09 · sales-10 · sales-01 · sales-04 · sales-05 · sales-22 · sales-23 · sales-24 · pnl-16

AI Consulting reads your books, quotes and vendor bills to find price leakage and overspending, then ranks the fixes by payback. A fraction of a margin point on real revenue adds up. Process fixes, like unused software or card fees, aren’t counted here.

Annual revenue

Last year’s total revenue. The pricing and vendor levers are sized against it.

Operating margin

Operating profit as a share of revenue, before interest and taxes. Shows what the margin gain means for your profit.

Gross profit from prices and costs, a year: $114,000 to $228,000
Margin gain: +0.95 to +1.9 pts

See the mathModeled estimate

The math behind the Pricing and vendors range, low case to high case
Pricing: quote and discount discipline	$12,000,000 × 0.5% to 1% = $60,000 to $120,000
Cost of sales, the spend vendors can touch (revenue less gross margin)	$12,000,000 × 60% = $7,200,000
Vendors and materials: renegotiation and rebids	$7,200,000 × 0.75% to 1.5% = $54,000 to $108,000
Margin gain	$114,000 to $228,000 ÷ $12,000,000 = 0.95 to 1.9 points
What that does to operating profit	$960,000 today (8%) → up 11.8% to 23.7%

The base case uses the base rates in our research and the upside the optimistic ones, never beyond them. Each step is rounded down. Sources: pnl-01 · pnl-30 · pnl-18 · pnl-17 · pnl-16 · pnl-08

Modeled estimate · 218 to 463 hrs a month$243K–$537K a yearBreakdown

Value a year · all five leversModeled estimate

$243,600 to $537,144 a year

$243,600base case, at typical rates

$537,144upside, at optimistic rates

Added revenue: $97,800 to $296,160 a yearcontext only: counted above at gross margin
Staff time back: 218 to 463 hours a monthabout 1.3 to 2.9 full-time staff of capacity
Break-even budget: $20,300 a month, at base ratesthe most the agents could cost and still pay off

Where the value comes from

Value a year by lever. Press a lever to count it or leave it out.

$74,880 to $157,320 a year
$36,720 to $103,824 a year
$12,000 to $24,000 a year
$6,000 to $24,000 a year
$114,000 to $228,000 a year

Modeled value a year by lever, base case and upside
Lever	Base case	Upside
Team time	$74,880	$157,320
Customer inquiries	$36,720	$103,824
Ad spend	$12,000	$24,000
Sales outreach	$6,000	$24,000
Pricing and vendors	$114,000	$228,000
Total counted	$243,600	$537,144

Modeled estimate: each bar adds up the levers you count. Base case at base rates, upside at optimistic rates.

Get your real numbers

Not a quote or a guarantee. Estimates exclude the cost of the engagement. A free assessment replaces these defaults with yours.

Modeled estimate, not a quote, a guarantee or a client result. Each range runs from the base rates in our research (low) to the optimistic rates (high), and never beyond them. Totals count profit only: staff time at the loaded hourly cost you enter, ad budget freed to reallocate, and gross profit on new jobs, deals and better prices. Added revenue is shown for context and never added to a total. The estimates assume the agents are fully up and running, so year one will be lower while they ramp up, and they exclude the cost of the engagement. Your results depend on your processes, systems and data. In a free assessment, we replace these assumptions with your real numbers. Methodology and sources

Turn on JavaScript to use the calculator, or email your numbers to hello@example.com and we'll run them for you.

02Industry scenarios

Find a business like yours.

Illustrative exampleLaw firm · 30 staff

Maple Row Law Group

Fictional businesssmall business$6M–$8M revenue

Modeled value, with 950 to 1,200 staff hours back a year.

Where the value comes from

Team time
Added gross profit

Where the value comes from, base case and upside case, per year
Source	Base	Upside
Attorney and intake time (Team time)	$37,200	$67,575
Billing leakage recovered (Added gross profit)	$70,000	$140,000
New matters from faster replies (Added gross profit)	$28,500	$57,000

Added revenue$127,000 to $254,000 a year

From new matters and recovered fees. Recovered fees are for work already done, so they count in full; new matters count at their 50% margin.

New inquiries that arrive while the team is with clients wait until the next business day, and some of those people hire another firm. The ten attorneys spend about 7 hours a week each on email, and finished work slips through billing and collections.

TodayWith agents

Reply time for new online inquiriesToday: Next business day after hoursWith agents: Under 2 minutes, 24/7
Inquiries fully handled without staffToday: 0%With agents: 35% of 1,000 a monthUpside 50%
Consultations booked per yearToday: 960With agents: 998Upside 1,036
Attorney time on emailToday: 7 hrs/week eachWith agents: About 5.8 hrs/week eachUpside 5.6

Agents

Customer Agent
Executive Assistant Agent

Time to value

Intake agent live in 2–3 weeks; full value in about 60 days

See the math

TodayWith agents

Reply time for new online inquiriesToday: Next business day after hoursWith agents: Under 2 minutes, 24/7
Inquiries fully handled without staffToday: 0%With agents: 35% of 1,000 a monthUpside 50%
Consultations booked per yearToday: 960With agents: 998Upside 1,036
Attorney time on emailToday: 7 hrs/week eachWith agents: About 5.8 hrs/week eachUpside 5.6

Maple Row is a 30-person estate and family law firm with ten attorneys. When the team is with clients, new inquiries wait until the next business day, and some of those people book with another firm. A Customer Agent now answers web, email and text inquiries within minutes, day or night, explains how consultations work using the firm's approved answers, and books directly into the attorneys' calendars. Anything that needs legal judgment goes to an attorney. An Executive Assistant Agent triages the attorneys' inboxes, drafts routine replies for approval and flags time that never reached an invoice. In this model the firm books 38 to 76 more consultations a year and recovers 1% to 2% of fees that used to slip.

What each agent does

Customer Agent. Answers web, email and text inquiries 24/7 from the firm's approved answers, checks that the matter fits the practice and books consultations; anything that needs legal judgment goes to an attorney
Executive Assistant Agent. Triages the attorneys' inboxes, drafts routine replies for approval, tracks deadlines and follow-ups, and flags unbilled time and overdue invoices

The math

Base case
Line item	Formula	Value
Intake handled by the agent	1,000 × 12 × 35% × 5 min ÷ 60	350 hours a year
Intake value (legal assistant $26.72 × 1.35 = $36/hr, counted at 50%)	350 × $36 × 50%	$6,300
Attorney email time returned (1.2 hrs/week each, measured)	10 × 1.2 × 50 weeks	600 hours a year
Attorney value (lawyer $76.76 × 1.35 = $103/hr, counted at 50%)	600 × $103 × 50%	$30,900
Extra consultations (40% of inquiries wait over an hour; 10% relative lift when answered in minutes)	960 × 40% × 10%, rounded down	38 a year
New matters (half of consultations become clients)	38 × 50%	19 a year
New-matter contribution ($3,000 flat fee at 50% margin)	19 × $3,000 × 50%	$28,500
Billing leakage recovered (1% of fees: unbilled time, invoicing and collections follow-up)	$7,000,000 × 1%	$70,000
Hours back	350 + 600	950 a year
Annual value	$6,300 + $30,900 + $28,500 + $70,000	$135,700

Upside case
Line item	Formula	Value
Intake handled (50%; counted at 75%)	1,000 × 12 × 50% × 5 min ÷ 60 = 500 h × $36 × 75%	$13,500
Attorney email time (20% of 7 hrs = 1.4 hrs/week each)	10 × 1.4 × 50 weeks	700 hours a year
Attorney value (counted at 75%)	700 × $103 × 75%	$54,075
Extra consultations (20% lift)	960 × 40% × 20%, rounded down	76 a year
New-matter contribution	76 × 50% = 38 × $3,000 × 50%	$57,000
Billing leakage recovered (2% of fees)	$7,000,000 × 2%	$140,000
Hours back	500 + 700	1,200 a year
Annual value	$13,500 + $54,075 + $57,000 + $140,000	$264,575

The upside case moves the rates that describe what the agents achieve to the optimistic end of the published evidence, all at once. The business’s own figures stay the same. Most results land in between.

Sourceslab-01 lab-06 cx-09 cx-10 cx-34 time-01 ai-15 cx-18 cx-19 cx-20 pnl-02 ai-08 ai-35

Illustrative exampleE-commerce brand · 40 staff

Fernway Home Goods

Fictional businesssmall business$18M–$22M revenue

Modeled value, with 3,198 to 4,620 staff hours back a year.

Where the value comes from

Team time
Added gross profit

Where the value comes from, base case and upside case, per year
Source	Base	Upside
Support and reporting time (Team time)	$47,307	$103,995
Ad optimization (Added gross profit)	$95,040	$216,000
Answers before purchase (Added gross profit)	$22,500	$67,500

Added revenue$391,800 to $945,000 a year

From ads and from answering shoppers before they buy. Ad revenue is attributed, not proven incremental, and only the 30% contribution margin is counted.

About 9,000 support tickets a month, mostly order status and returns, bury the support team, and first replies take about nine hours, so shoppers with a question before they buy often leave. Returns on $120,000 a month of ads have slipped because nobody has time to manage campaigns closely.

TodayWith agents

Tickets resolved without a personToday: 0%With agents: 35% of 9,000 a monthUpside 50%
First reply timeToday: About 9 hoursWith agents: Under 2 minutes
Attributed revenue per $1 of ad spendToday: $2.00With agents: $2.22Upside $2.50
Support team hours on ticketsToday: 750 a monthWith agents: About 488 a monthUpside 375

Agents

Customer Agent
AI Marketing Director
Analyst Agent

Time to value

Support agent live in about 2 weeks; ad gains build over 60–90 days

See the math

TodayWith agents

Tickets resolved without a personToday: 0%With agents: 35% of 9,000 a monthUpside 50%
First reply timeToday: About 9 hoursWith agents: Under 2 minutes
Attributed revenue per $1 of ad spendToday: $2.00With agents: $2.22Upside $2.50
Support team hours on ticketsToday: 750 a monthWith agents: About 488 a monthUpside 375

Fernway is a 40-person home-goods brand that sells through its own Shopify store. About 9,000 support tickets a month, mostly order status and returns, bury the team, and first replies take around nine hours, so shoppers who ask a question before buying often leave. Ad returns have slipped because nobody has time to manage campaigns closely. A Customer Agent now resolves a third to a half of tickets on its own, from tracking updates to return labels, answers pre-purchase questions in minutes and hands the rest to the team with the order history attached. An AI Marketing Director audits the ad accounts, tests new creative and audiences, and shifts budget toward what sells, within limits the founders set.

What each agent does

Customer Agent. Resolves order-status, return and product questions by email and chat, answers shoppers before they buy, and hands the rest to the team with the order history attached
AI Marketing Director. Audits the Meta and Google accounts, runs creative and audience tests, and shifts budget toward what sells within limits the founders set
Analyst Agent. Sends a margin and inventory report every Monday morning

The math

Base case
Line item	Formula	Value
Tickets resolved by the agent	9,000 × 12 × 35% × 5 min ÷ 60	3,150 hours a year
Support value (customer service rep $29/hr, counted at 50%)	3,150 × $29 × 50%	$45,675
Extra attributed revenue (starting return 2.0x; +11% from 60–90 days of optimization)	$120,000 × 12 × 2.0 × 11%	$316,800
Ad contribution (30% margin after product, shipping, fees and returns)	$316,800 × 30%	$95,040
Sales from answering before purchase (5% lift on the $1.5M a year bought by shoppers who ask first)	$1,500,000 × 5%	$75,000
Pre-purchase contribution (30% margin)	$75,000 × 30%	$22,500
Reporting time returned (10% less time on recurring reports)	40 hrs × 12 × 10%	48 hours a year
Reporting value (operations manager $68/hr, counted at 50%)	48 × $68 × 50%	$1,632
Hours back	3,150 + 48	3,198 a year
Annual value	$45,675 + $95,040 + $22,500 + $1,632	$164,847

Upside case
Line item	Formula	Value
Tickets resolved (50% of tickets)	9,000 × 12 × 50% × 5 min ÷ 60	4,500 hours a year
Support value (counted at 75%)	4,500 × $29 × 75%	$97,875
Extra attributed revenue (+25%)	$120,000 × 12 × 2.0 × 25%	$720,000
Ad contribution	$720,000 × 30%	$216,000
Sales from answering before purchase (15% lift)	$1,500,000 × 15%	$225,000
Pre-purchase contribution	$225,000 × 30%	$67,500
Reporting time (25% less; counted at 75%)	40 hrs × 12 × 25% = 120 h × $68 × 75%	$6,120
Hours back	4,500 + 120	4,620 a year
Annual value	$97,875 + $216,000 + $67,500 + $6,120	$387,495

Sourceslab-01 lab-06 cx-09 cx-10 cx-34 ai-26 mkt-05 mkt-06 mkt-19 pnl-09 pnl-31 ai-16 ai-24 ai-08 ai-35

Illustrative exampleHome services · 85 staff

Gablecrest Roofing & Exteriors

Fictional businessmid-size business$20M–$25M revenue

Modeled value, with 362 to 550 staff hours back a year.

Where the value comes from

Team time
Added gross profit
Cost savings

Where the value comes from, base case and upside case, per year
Source	Base	Upside
Office time on calls and texts (Team time)	$5,973	$13,612
Jobs won by replying first (Added gross profit)	$89,760	$179,520
Search spend cut (Cost savings)	$32,400	$64,800

Added revenue$264,000 to $528,000 a year

From extra roofing jobs. Only the 34% gross margin is counted.

About 4 in 10 of its 500 new leads a month wait more than an hour for a callback, and by then many homeowners have called another roofer. Estimators chase follow-ups at night, and no one has time to check where the $45,000 monthly search budget goes.

TodayWith agents

Time to first reply on new leadsToday: Over an hour for 4 in 10With agents: Under 1 minute
Inspections booked per yearToday: 2,400With agents: 2,496Upside 2,592
Search spend for the same 500 leads a monthToday: $45,000 ($90 a lead)With agents: $42,300 ($84.60 a lead)Upside $39,600
Calls and texts handled without staffToday: 0%With agents: 362 a monthUpside 550

Agents

Customer Agent
AI Marketing Director

Time to value

Lead response live in about 2 weeks; ad savings within 30–60 days

See the math

TodayWith agents

Time to first reply on new leadsToday: Over an hour for 4 in 10With agents: Under 1 minute
Inspections booked per yearToday: 2,400With agents: 2,496Upside 2,592
Search spend for the same 500 leads a monthToday: $45,000 ($90 a lead)With agents: $42,300 ($84.60 a lead)Upside $39,600
Calls and texts handled without staffToday: 0%With agents: 362 a monthUpside 550

Gablecrest is an 85-person roofing and exteriors company. About four in ten of its 500 new leads a month wait more than an hour for a callback, and by then many homeowners have already called another roofer. Estimators chase follow-ups at night, and no one has time to check where the ad budget goes. A Customer Agent now replies to every lead within a minute by text or chat, answers common questions, and books inspections straight onto the estimators' calendars. An AI Marketing Director finds search spend that never produces a lead and cuts it, so the same 500 leads a month cost $2,700 to $5,400 less. Booked inspections rise by 96 to 192 a year.

What each agent does

Customer Agent. Replies to every lead within a minute by text or chat, answers common questions, books inspections on the estimators' calendars and follows up
AI Marketing Director. Finds and cuts search ad spend that never produces a lead, within budget rules the owner sets

The math

Base case
Line item	Formula	Value
Extra inspections (40% of leads wait over an hour; 10% relative lift)	500 × 12 × 40% booked × 40% × 10%	96 a year
Extra jobs (25% of inspections close)	96 × 25%	24 a year
Gross profit ($11,000 average job at a 34% specialty-trade margin)	24 × $11,000 × 34%	$89,760
Search spend cut with no lost leads (6% of search spend)	$45,000 × 12 × 6%	$32,400
Calls and texts resolved by the agent	(750 texts × 35% + 500 calls × 20%) × 12 × 5 min ÷ 60, rounded down	362 hours a year
Office value (scheduler $24.20 × 1.40 = $33/hr, counted at 50%)	362 × $33 × 50%	$5,973
Hours back		362 a year
Annual value	$89,760 + $32,400 + $5,973	$128,133

Upside case
Line item	Formula	Value
Extra inspections (20% relative lift)	500 × 12 × 40% × 40% × 20%	192 a year
Gross profit	192 × 25% = 48 jobs × $11,000 × 34%	$179,520
Search spend cut (12% of search spend)	$45,000 × 12 × 12%	$64,800
Calls and texts resolved (50% of texts, 35% of calls)	(750 × 50% + 500 × 35%) × 12 × 5 min ÷ 60	550 hours a year
Office value (counted at 75%, rounded down)	550 × $33 × 75%	$13,612
Annual value	$179,520 + $64,800 + $13,612	$257,932

Sourceslab-01 lab-06 cx-18 cx-19 cx-20 pnl-16 mkt-01 mkt-19 mkt-18 cx-09 cx-16 cx-34 ai-08 ai-35

Illustrative exampleDental group · 120 staff

Willowmere Dental Group

Fictional businessmid-size business$18M–$22M revenue

Modeled value, with 1,800 to 3,204 staff hours back a year.

Where the value comes from

Team time
Added gross profit

Where the value comes from, base case and upside case, per year
Source	Base	Upside
Front-desk and reporting time (Team time)	$27,576	$74,250
New patients from answered calls (Added gross profit)	$136,243	$540,715

Added revenue$340,608 to $1,351,788 a year

From recovered new patients, in first-year revenue. Only the 40% contribution margin is counted.

About 28% of calls go unanswered at peak times, after-hours callers reach voicemail, and many of those callers book with another practice. Office managers in eight offices build reports by hand.

TodayWith agents

Calls that reach no oneToday: About 28%With agents: Answered by the agent, 24/7
New patients per monthToday: 240With agents: 272Upside 367
Answered calls fully handled without staffToday: 0%With agents: 20% (1,728 a month)Upside 35%
Office manager time on reportsToday: 60 hrs/monthWith agents: 54 hrs/monthUpside 45

Agents

Customer Agent
Analyst Agent

Time to value

Live in 3–4 weeks after testing with the practice-management system; full value in about 90 days

See the math

TodayWith agents

Calls that reach no oneToday: About 28%With agents: Answered by the agent, 24/7
New patients per monthToday: 240With agents: 272Upside 367
Answered calls fully handled without staffToday: 0%With agents: 20% (1,728 a month)Upside 35%
Office manager time on reportsToday: 60 hrs/monthWith agents: 54 hrs/monthUpside 45

Willowmere is an eight-office dental group with 120 employees. At peak times more than one call in four goes unanswered, after-hours callers reach voicemail, and some of those callers book elsewhere. A Customer Agent now answers overflow and after-hours calls and texts, books new patients, and confirms and reschedules appointments. Its access is limited to scheduling and patient messaging, every action is logged, and clinical questions go straight to the team. In this model it adds 32 to 127 new patients a month and fully handles one in five to one in three routine calls. An Analyst Agent pulls data from all eight offices into one weekly dashboard.

What each agent does

Customer Agent. Answers overflow and after-hours calls and texts, books new patients, confirms and reschedules appointments; access is limited to scheduling and patient messaging, and every action is logged
Analyst Agent. Builds one weekly dashboard of production, collections and schedule fill across all eight offices

The math

Base case
Line item	Formula	Value
Missed calls that are new-business opportunities	12,000 × 28% × 30%	1,008 a month
New patients recovered (40% would be lost; agent completes 20% of calls; 40% book)	1,008 × 40% × 20% × 40%, rounded down	32 a month
New-patient contribution ($887 first-year revenue at 40% margin)	32 × 12 × $887 × 40%, rounded down	$136,243
Answered calls fully handled by the agent	12,000 × 72% × 20% × 12 × 5 min ÷ 60	1,728 hours a year
Front-desk value (medical secretary $22.08 × 1.40 = $30/hr, counted at 50%)	1,728 × $30 × 50%	$25,920
Reporting time returned (10% less time on recurring reports)	60 hrs × 12 × 10%	72 hours a year
Reporting value (office manager $33.41 × 1.40 = $46/hr, counted at 50%)	72 × $46 × 50%	$1,656
Hours back	1,728 + 72	1,800 a year
Annual value	$136,243 + $25,920 + $1,656	$163,819

Upside case
Line item	Formula	Value
New patients recovered (60% would be lost; agent completes 35%; 60% book)	1,008 × 60% × 35% × 60%, rounded down	127 a month
New-patient contribution	127 × 12 × $887 × 40%, rounded down	$540,715
Answered calls fully handled (35% of calls)	12,000 × 72% × 35% × 12 × 5 min ÷ 60	3,024 hours a year
Front-desk value (counted at 75%)	3,024 × $30 × 75%	$68,040
Reporting time (25% less; counted at 75%)	60 hrs × 12 × 25% = 180 h × $46 × 75%	$6,210
Hours back	3,024 + 180	3,204 a year
Annual value	$540,715 + $68,040 + $6,210	$614,965

Sourceslab-01 lab-06 cx-22 cx-26 cx-23 cx-21 cx-25 cx-16 cx-31 pnl-10 cx-34 ai-16 ai-08 ai-35

Illustrative exampleRestaurant group · 180 staff

Bramblewood Restaurant Group

Fictional businessmid-size business$13M–$16M revenue

Modeled value, with 300 to 750 staff hours back a year.

Where the value comes from

Team time
Added gross profit
Cost savings

Where the value comes from, base case and upside case, per year
Source	Base	Upside
Manager time (Team time)	$6,900	$25,875
Catering and events (Added gross profit)	$21,600	$43,200
Food cost (Cost savings)	$43,200	$86,400
Labor scheduling (Cost savings)	$26,280	$78,840

Added revenue$72,000 to $144,000 a year

From extra catering orders and events. Only the 30% margin is counted.

Supplier prices change weekly, but food cost is reviewed only at month-end. Schedules follow habit rather than forecast covers, general managers spend about 10 hours a week each on invoices, inventory sheets and reports, and catering inquiries often wait a day for a reply.

TodayWith agents

Food cost (share of food sales)Today: 32.0%With agents: 31.6%Upside 31.2%
Reply time for catering and event inquiriesToday: Next dayWith agents: Within 5 minutes
Catering orders and private events bookedToday: 240 a yearWith agents: 264 a yearUpside 288
GM time on invoices, inventory and reportsToday: 10 hrs/week eachWith agents: 9 hrs/week eachUpside 7.5

Agents

Analyst Agent
AI Consulting
Customer Agent

Time to value

Invoice tracking live in about 3 weeks; food-cost gains within 60–90 days

See the math

TodayWith agents

Food cost (share of food sales)Today: 32.0%With agents: 31.6%Upside 31.2%
Reply time for catering and event inquiriesToday: Next dayWith agents: Within 5 minutes
Catering orders and private events bookedToday: 240 a yearWith agents: 264 a yearUpside 288
GM time on invoices, inventory and reportsToday: 10 hrs/week eachWith agents: 9 hrs/week eachUpside 7.5

Bramblewood is a six-location restaurant group with 180 employees. Supplier prices change weekly, but food cost is reviewed only at month-end, long after the margin is gone. General managers spend about ten hours a week on invoices, inventory sheets and reports, and catering inquiries often wait a day for a reply. An Analyst Agent now reads every supplier invoice, tracks price changes and food cost by location, and flags dishes that are losing margin. AI Consulting recommends menu, vendor and staffing changes for the owners to approve, trimming food cost by 0.4 to 0.8 points in year one and matching schedules to forecast covers. A Customer Agent answers catering and private-event inquiries within minutes.

What each agent does

Analyst Agent. Reads every supplier invoice, tracks price changes and food cost by location, and flags dishes that are losing margin
AI Consulting. Recommends menu-price, vendor and staffing changes for the owners to approve, with schedules matched to forecast covers
Customer Agent. Answers catering and private-event inquiries within minutes and books them

The math

Base case
Line item	Formula	Value
Food cost (0.4-point year-one improvement on $10.8M food sales)	$10,800,000 × 0.4%	$43,200
Labor scheduling (1% of $5.26M labor, 36.5% of $14.4M sales; half counted in year one)	$5,256,000 × 1% × 50%	$26,280
Extra events (all inquiries wait over an hour today; 10% relative lift)	20 × 12 × 10%	24 a year
Event contribution ($3,000 average at 30% margin)	24 × $3,000 × 30%	$21,600
GM time returned (10% less time on recurring reports and cleanup)	6 × 10 hrs × 50 weeks × 10%	300 hours a year
GM value (food service manager $33.36 × 1.40 = $46/hr, counted at 50%)	300 × $46 × 50%	$6,900
Hours back		300 a year
Annual value	$43,200 + $26,280 + $21,600 + $6,900	$97,980

Upside case
Line item	Formula	Value
Food cost (0.8 points)	$10,800,000 × 0.8%	$86,400
Labor scheduling (2%; 75% counted in year one)	$5,256,000 × 2% × 75%	$78,840
Extra events (20% lift)	20 × 12 × 20% = 48 × $3,000 × 30%	$43,200
GM time (25% less; counted at 75%)	6 × 10 hrs × 50 weeks × 25% = 750 h × $46 × 75%	$25,875
Hours back		750 a year
Annual value	$86,400 + $78,840 + $43,200 + $25,875	$234,315

Sourceslab-01 lab-06 pnl-11 pnl-12 pnl-18 pnl-01 pnl-15 pnl-25 cx-18 cx-19 cx-20 ai-16 ai-23 ai-08 ai-35

Illustrative exampleIndustrial distributor · 250 staff

Ironvale Industrial Supply

Fictional businessmid-size business$70M–$80M revenue

Modeled value, with 4,270 to 6,080 staff hours back a year.

Where the value comes from

Team time
Added gross profit
Cost savings

Where the value comes from, base case and upside case, per year
Source	Base	Upside
Team time on email and outreach (Team time)	$79,275	$169,200
Price discipline on quotes (Added gross profit)	$90,000	$270,000
Supplier price checks (Cost savings)	$75,000	$225,000

Added revenue$90,000 to $270,000 a year

From better-priced quotes. Costs stay the same, so all of it is gross profit.

Inside sales and customer service reps spend much of the day on about 12,000 order-status and quote-request emails a month. Quotes take a full business day, each rep prices from a slightly different spreadsheet, and supplier price increases slip into invoices unnoticed.

TodayWith agents

Routine emails resolved without a repToday: 0%With agents: 35% of 12,000 a monthUpside 50%
Quote turnaroundToday: Next business dayWith agents: Same day
Quotes and supplier invoices checkedToday: Spot checksWith agents: Every one
Lapsed accounts sent a personal noteToday: NoneWith agents: 100 a month

Agents

Customer Agent
Analyst Agent
AI Sales Agent

Time to value

Email agent live in 4–6 weeks; pricing findings within the first 30 days; full run-rate in 4–6 months

See the math

TodayWith agents

Routine emails resolved without a repToday: 0%With agents: 35% of 12,000 a monthUpside 50%
Quote turnaroundToday: Next business dayWith agents: Same day
Quotes and supplier invoices checkedToday: Spot checksWith agents: Every one
Lapsed accounts sent a personal noteToday: NoneWith agents: 100 a month

Ironvale is a 250-person industrial distributor with three branches. Its inside sales and customer service reps spend much of the day on order-status and quote-request emails, quotes take a full business day, and each rep prices from a slightly different spreadsheet. A Customer Agent now resolves a third or more of routine emails and turns quote requests into ready-to-price drafts. An Analyst Agent checks every quote against margin floors and current costs, and compares supplier invoices with the prices purchasing agreed. An AI Sales Agent researches lapsed accounts and drafts each buyer a personal note. Reps approve every quote and email, and get 4,270 to 6,080 hours a year back.

What each agent does

Customer Agent. Answers routine order-status and tracking emails and turns quote requests into ready-to-price drafts
Analyst Agent. Checks every quote against margin floors and current costs, compares supplier invoices with the prices purchasing agreed and flags exceptions before they cost money
AI Sales Agent. Researches lapsed accounts and drafts each buyer a personal note; reps approve every quote and outreach email

The math

Base case
Line item	Formula	Value
Emails resolved by the agent	12,000 × 12 × 35% × 5 min ÷ 60	4,200 hours a year
Email value (blended $37/hr: service rep $30 and wholesale rep $45; counted at 50%)	4,200 × $37 × 50%	$77,700
Price discipline (0.5% better realized price on $36M quoted; half counted in year one)	$36,000,000 × 0.5% × 50%	$90,000
Supplier price checks (0.75% of the $20M of purchases reviewed in year one; half counted)	$20,000,000 × 0.75% × 50%	$75,000
Outreach drafting time returned (10 min per note; 35% faster with AI)	100 × 12 × 10 min ÷ 60 × 35%	70 hours a year
Outreach value (wholesale rep $45/hr, counted at 50%)	70 × $45 × 50%	$1,575
Reactivated accounts		Not counted (no reliable benchmark)
Hours back	4,200 + 70	4,270 a year
Annual value	$77,700 + $90,000 + $75,000 + $1,575	$244,275

Upside case
Line item	Formula	Value
Emails resolved (50% of emails)	12,000 × 12 × 50% × 5 min ÷ 60	6,000 hours a year
Email value (counted at 75%)	6,000 × $37 × 75%	$166,500
Price discipline (1% better realized price; 75% counted in year one)	$36,000,000 × 1% × 75%	$270,000
Supplier price checks (1.5%; 75% counted)	$20,000,000 × 1.5% × 75%	$225,000
Outreach time (40% faster; counted at 75%)	100 × 12 × 10 min ÷ 60 × 40% = 80 h × $45 × 75%	$2,700
Hours back	6,000 + 80	6,080 a year
Annual value	$166,500 + $270,000 + $225,000 + $2,700	$664,200

Sourceslab-01 lab-06 lab-07 cx-09 cx-10 cx-34 pnl-01 pnl-30 pnl-15 pnl-18 pnl-17 ai-03 sales-11 ai-08 ai-35

Fictional businesses. Modeled estimates, not client results. Time to value is a planning estimate for a typical scope, not a track record.

03Worked examples

One worked example for every service.

Modeled estimate Illustrative example

AI Consulting

Corran Valley Metalworks

Fictional custom metal fabricator · 160 employees

AI Consulting analyzes the ledger, vendor spend, overtime, quotes against actual job costs and software seats, then ranks each opportunity and delivers a 90-day roadmap.

3 to 7 hours back a month

Team time
Added gross profit
Cost savings

$15,662 × 12 = $187,944Upside $45,628 × 12 = $547,536

Added revenue$60,000 to $180,000 a year

From repricing quotes on current material costs. The jobs cost the same, so all of it is gross profit.

Today

Revenue and EBITDA: $30,000,000; $2,400,000 (8.0%)
Materials; freight; software: $12,000,000; $1,200,000; $240,000
Overtime pay: $900,000 a year
Quoted from stale material costs: $6,000,000 of revenue a year
Controller on margin reports: 30 hours a month

With the plan

Savings identified: $348,000 a year (upside $708,000)
Counted in year one: $186,000 (upside $543,000)
EBITDA with counted savings: $2,586,000, 8.62% (upside 9.81%)
Controller hours back: 3 a month (upside 7)

Where the value comes from
Source	Base	Upside
Controller time (Team time)	$1,944	$4,536
Pricing on stale-cost quotes (Added gross profit)	$60,000	$180,000
Materials, freight and software (Cost savings)	$81,000	$228,000
Overtime (Cost savings)	$45,000	$135,000
Modeled value a year	$187,944	$547,536

The math

Base case
Line item	Formula	Value
Pricing: reprice quotes on current material costs	$6,000,000 × 2% = $120,000; count half	$60,000
Overtime: better scheduling and realistic lead times	$900,000 × 10% = $90,000; count half	$45,000
Materials: renegotiate and consolidate suppliers	$12,000,000 × 0.75% = $90,000; count half	$45,000
Freight: consolidate and rebid carriers	$1,200,000 × 2% = $24,000; count half	$12,000
Software: cancel unused seats	$240,000 × 10% = $24,000; in the client's control	$24,000
Identified	$120,000 + $90,000 + $90,000 + $24,000 + $24,000	$348,000
Counted in year one	$60,000 + $45,000 + $45,000 + $12,000 + $24,000	$186,000
Counted savings per month	$186,000 ÷ 12	$15,500
Controller: recurring margin report	30 × 10% = 3 hours × $54 (accountant)	$162
Monthly value	$15,500 + $162	$15,662
Annual value	$15,662 × 12	$187,944
EBITDA margin with counted savings	($2,400,000 + $186,000) ÷ $30,000,000	8.62%

Upside case
Line item	Formula	Value
Pricing at 4%, 75% counted	$6,000,000 × 4% = $240,000 × 75%	$180,000
Overtime at 20%, 75% counted	$900,000 × 20% = $180,000 × 75%	$135,000
Materials at 1.5%, 75% counted	$12,000,000 × 1.5% = $180,000 × 75%	$135,000
Freight at 5%, 75% counted	$1,200,000 × 5% = $60,000 × 75%	$45,000
Software at 20%, counted in full	$240,000 × 20%	$48,000
Identified	$240,000 + $180,000 + $180,000 + $60,000 + $48,000	$708,000
Counted in year one	$180,000 + $135,000 + $135,000 + $45,000 + $48,000	$543,000
Controller (reports 25% faster)	30 × 25% = 7 hours (rounded down) × $54	$378
Monthly value	$543,000 ÷ 12 + $378	$45,628
Annual value	$45,628 × 12	$547,536
EBITDA margin	($2,400,000 + $543,000) ÷ $30,000,000	9.81%

The upside case moves the rates that describe what the agent achieves to the optimistic end of the published evidence, all at once. The business’s own figures stay the same. Most results land in between.

Assumptions

Base lever rates: 2% repricing on stale metal quotes, 10% of overtime pay, 0.75% of materials spend, 2% of freight, 10% of software spend. Upside: 4%, 20%, 1.5%, 5% and 20%.
Half of each saving that depends on negotiation, customer acceptance or a ramp-up is counted in year one (upside 75%); cancelling unused software seats is counted in full.
Controller time valued at an accountant's $54 an hour, below a controller's.
Not counted: the rest of the identified savings, revenue growth from faster, more accurate quotes.

Sources

pnl-29 pnl-26 pnl-25 pnl-18 pnl-17 pnl-28 pnl-04 pnl-05 pnl-15 pnl-08 pnl-16 lab-01 lab-06 ai-06

Modeled estimate Illustrative example

AI Marketing Director

Juniper Lane Kitchen & Bath

Fictional kitchen and bathroom remodeler · 75 employees

The AI Marketing Director audits campaigns daily, cuts spend that produces nothing, runs controlled tests and moves budget to what works, within limits the owner sets.

7 to 14 hours back a month

Team time
Added gross profit

$12,237 × 12 = $146,844Upside $27,714 × 12 = $332,568

Added revenue$475,200 to $1,080,000 a year

From extra remodeling jobs at the same ad spend. Only the 30% gross margin is counted.

Today

Monthly ad spend: $27,000
Cost per lead: About $90 (home services median)
Leads a month: 300
Close rate and average job: 10% of leads; $12,000 at a 30% gross margin
Marketing team on reports and ad copy: 50 hours a month (30 reports, 20 copy and briefs)

With the agent

Cost per lead after 60-90 days: About $81, 10% lower (upside $72)
Leads a month, same spend: 333, +33 (upside 375)
Extra jobs: 3.3 a month (upside 7.5)
Marketing team hours back: 7 a month (upside 14)

Where the value comes from
Source	Base	Upside
Marketing team time (Team time)	$4,284	$8,568
More leads at the same spend (Added gross profit)	$142,560	$324,000
Modeled value a year	$146,844	$332,568

The math

Base case
Line item	Formula	Value
Leads today	$27,000 ÷ $90	300
Cost per lead after optimization	$90 × (1 − 10%)	$81
Extra leads at the same spend (rounded down)	$27,000 ÷ $81 − 300	33
Extra jobs at today's close rate	33 × 10%	3.3
Added revenue	3.3 × $12,000	$39,600
Gross profit	$39,600 × 30%	$11,880
Marketing team: recurring reports	30 × 10%	3 hours
Marketing team: ad copy and briefs	20 × 20%	4 hours
Value of time (marketing specialist, $51/h)	7 × $51	$357
Monthly value	$11,880 + $357	$12,237
Annual value	$12,237 × 12	$146,844

Upside case
Line item	Formula	Value
Cost per lead (20% lower)	$90 × (1 − 20%)	$72
Extra leads at the same spend	$27,000 ÷ $72 − 300	75
Extra jobs	75 × 10%	7.5
Added revenue	7.5 × $12,000	$90,000
Gross profit	$90,000 × 30%	$27,000
Marketing team (reports 25%, copy 35% faster; rounded down)	30 × 25% + 20 × 35% = 14 hours × $51	$714
Monthly value	$27,000 + $714	$27,714
Annual value	$27,714 × 12	$332,568

Assumptions

We model 10% lower cost per lead after 60-90 days of disciplined optimization (upside 20%); no study measures this directly.
Starting cost per lead about $90, the home-services search median.
Close rate stays at 10%; average job $12,000 at a 30% gross margin (the low end of specialty-trade margins). Only gross profit is counted.
Marketing team time at a marketing specialist's $51 an hour; reports 10% faster and first drafts 20% faster (upside 25% and 35%).
Not counted: wins from individual tests (most tests produce no significant winner), better lead quality, any choice to lower spend.

Sources

mkt-01 mkt-02 mkt-19 mkt-14 mkt-15 mkt-16 lab-01 lab-06 pnl-16 ai-06 ai-24

Illustrative example Modeled estimate

AI Agents · Executive Assistant Agent

Lanternfield Insurance Agency

Fictional independent property and casualty agency · 60 employees

The Executive Assistant Agent triages the team's inboxes, drafts replies for approval, writes meeting briefs and action-item summaries, and builds a first draft of the monthly leadership deck.

186 to 253 hours back a month

All of it from team time

$6,370 × 12 = $76,440Upside $8,711 × 12 = $104,532

Today

Owner email time: 28 hours a month (7 a week)
Owner meeting prep and briefs: 12 hours a month
Producer email time (12 producers): 336 hours a month (7 a week each)
Account manager email time (30 account managers): 840 hours a month (7 a week each)
Office manager: leadership deck: 10 hours a month
Office manager: notes and task follow-up: 12 hours a month

With the agent

Owner hours back: 7 a month (upside 11)
Producer hours back: 50 a month in total (upside 67)
Account manager hours back: 126 a month in total (upside 168)
Office manager hours back: 3 a month (upside 7)
Per producer or account manager: About 1 hour a week (upside 1.4)

Where the value comes from
Source	Base	Upside
Account managers' time (Team time)	$43,848	$58,464
Producers' time (Team time)	$25,800	$34,572
Owner and office manager time (Team time)	$6,792	$11,496
Modeled value a year	$76,440	$104,532

The math

Base case
Line item	Formula	Value
Owner: email workload	28 × 15%	4.2 hours
Owner: meeting briefs and summaries	12 × 25%	3 hours
Owner hours back (rounded down)	4.2 + 3	7 hours
Producers: email workload (rounded down)	12 × 7 × 4 × 15%	50 hours
Account managers: email workload	30 × 7 × 4 × 15%	126 hours
Office manager: deck	10 × 5%	0.5 hours
Office manager: notes and follow-up	12 × 25%	3 hours
Office manager hours back (rounded down)	0.5 + 3	3 hours
Hours back	7 + 50 + 126 + 3	186 hours
Value of time (owner $68, producers $43, account managers $29, office manager $30)	7 × $68 + 50 × $43 + 126 × $29 + 3 × $30	$6,370
Annual value	$6,370 × 12	$76,440

Upside case
Line item	Formula	Value
Owner (email 20%, briefs 45% faster)	28 × 20% + 12 × 45%	11 hours
Producers (email 20% lower, rounded down)	12 × 7 × 4 × 20%	67 hours
Account managers (email 20% lower)	30 × 7 × 4 × 20%	168 hours
Office manager (deck 20%, notes 45% faster; rounded down)	10 × 20% + 12 × 45%	7 hours
Hours back	11 + 67 + 168 + 7	253 hours
Value of time	11 × $68 + 67 × $43 + 168 × $29 + 7 × $30	$8,711
Annual value	$8,711 × 12	$104,532

Assumptions

7 hours a week of email per person (base; the measured large-firm figure is 11.65). It describes the team, so it stays the same in both cases.
Email workload 15% lower for active users, summaries 25% faster, slide decks 5% faster (net of review); 20%, 45% and 20% in the upside case.
Producers and account managers get back about 1 to 1.4 hours a week each, under the evidence ceiling on routine admin time returned (2 hours base, 4.9 upside).
Owner valued at a general and operations manager's $68 an hour; producers at a services sales rep's $43; account managers at a customer service rep's $29; office manager at an administrative assistant's $30.
Not counted: new policies and renewals written with the time back, fewer missed follow-ups.

Sources

lab-01 lab-06 lab-07 time-01 time-02 ai-15 ai-20 ai-21 ai-23

Illustrative example Modeled estimate

AI Agents · Analyst Agent

Greenhollow Landscape Co.

Fictional commercial and residential landscaper · 110 employees across three branches

The Analyst Agent pulls and cleans the monthly exports, drafts the report narrative, flags finished extra work that never reached an invoice and shows which kinds of jobs are underpriced.

16 to 36 hours back a month

Team time
Added gross profit

$5,883 × 12 = $70,596Upside $12,478 × 12 = $149,736

Added revenue$60,996 to $129,000 a year

From recovered billing and better-priced bids. The work is already done or costs the same, so all of it is gross profit.

Today

Accounting team on month-end reporting: 80 hours a month (48 cleanup, 32 report building)
Operations managers on spreadsheets: 50 hours a month (30 job costing, 20 owner questions)
Month-end close: About 6.4 days (the benchmark median)
Extra work done but not invoiced: $6,000 a month
Bid work priced from estimating spreadsheets: $10,000,000 a year

With the agent

Accounting hours back: 8 a month (upside 20)
Operations hours back: 8 a month (upside 16)
Month-end close: About 1 day sooner (upside 1.6)
Unbilled extra work recovered: $3,000 a month (upside $4,500)
Realized prices on bid work: 0.5% higher (upside 1%)

Where the value comes from
Source	Base	Upside
Accounting and operations time (Team time)	$9,600	$20,736
Unbilled extra work recovered (Added gross profit)	$36,000	$54,000
Better-priced bids (Added gross profit)	$24,996	$75,000
Modeled value a year	$70,596	$149,736

The math

Base case
Line item	Formula	Value
Accounting: data cleanup	48 × 10%	4.8 hours
Accounting: recurring report building	32 × 10%	3.2 hours
Accounting hours back	4.8 + 3.2	8 hours
Operations: job-cost spreadsheets	30 × 10%	3 hours
Operations: summaries for the owner	20 × 25%	5 hours
Operations hours back	3 + 5	8 hours
Value of time (bookkeeper $32/h, operations manager $68/h)	8 × $32 + 8 × $68	$800
Recovered billing (half of what is flagged)	$6,000 × 50%	$3,000
Better-priced bids (0.5% higher realized price; half counted in year one)	$10,000,000 × 0.5% × 50% ÷ 12	$2,083
Monthly value	$800 + $3,000 + $2,083	$5,883
Annual value	$5,883 × 12	$70,596

Upside case
Line item	Formula	Value
Accounting hours back (cleanup and reports 25% faster)	(48 + 32) × 25%	20 hours
Operations hours back (job costing 25%, summaries 45% faster)	30 × 25% + 20 × 45%, rounded down	16 hours
Value of time	20 × $32 + 16 × $68	$1,728
Recovered billing (75% of what is flagged)	$6,000 × 75%	$4,500
Better-priced bids (1%; 75% counted in year one)	$10,000,000 × 1% × 75% ÷ 12	$6,250
Monthly value	$1,728 + $4,500 + $6,250	$12,478
Annual value	$12,478 × 12	$149,736

Assumptions

Measured time savings are small for spreadsheet work: 10% on recurring cleanup and report building and 25% on summaries in the base case; 25% and 45% in the upside case (net of human review).
Bookkeeping clerk time at $32 an hour loaded; operations manager at $68 (OEWS medians × 1.35).
Half of the flagged unbilled work is invoiced and collected in the year (upside 75%); the work is already done, so it counts as margin.
Job costing shows which kinds of jobs are underpriced: realized prices on new bid work rise 0.5% (upside 1%), with half (upside 75%) counted in year one.
Close cycle 1 day shorter (upside 1.6) is a modeled target, not a measured automation effect.
Not counted: faster decisions from timely reports, the owner's own time.

Sources

lab-01 lab-06 ai-06 ai-21 ai-23 ai-20 time-06 pnl-01 pnl-30 pnl-15 pnl-18

Modeled estimate Illustrative example

AI Agents · Customer Agent

Larkfield Plumbing & Air

Fictional residential plumbing, heating and cooling company · 18 employees

The Customer Agent answers routine inquiries end to end, picks up calls nobody gets to and replies to after-hours web requests in under a minute, so fewer new jobs go to a competitor.

22 to 34 hours back a month

Team time
Added gross profit

$2,247 × 12 = $26,964Upside $4,818 × 12 = $57,816

Added revenue$48,276 to $114,984 a year

From jobs that no longer go to a competitor. Only the 40% gross margin is counted.

Today

Customer inquiries a month: 1,200 (600 calls, 600 online)
Staff time on inquiries: 100 hours a month (5 minutes each)
Calls nobody answers: 84 a month (14%)
Online inquiries after hours: 240 a month (40%), wait until morning
New-job requests: 30% of inquiries
Average service call: $341 at a 40% gross margin

With the agent

Resolved without staff: 330 a month (upside 510)
Staff hours back: 22 a month after review time (upside 34)
Replies after hours and to missed calls: Under a minute, 24/7
Jobs no longer lost: 11.8 a month (upside 28.1)

Where the value comes from
Source	Base	Upside
Staff time on routine inquiries (Team time)	$7,656	$11,832
Jobs recovered after hours (Added gross profit)	$18,000	$40,920
Jobs recovered from missed calls (Added gross profit)	$1,308	$5,064
Modeled value a year	$26,964	$57,816

The math

Base case
Line item	Formula	Value
Staff time on inquiries today	1,200 × 5 ÷ 60	100 hours
Resolved by the agent	600 × 20% + 600 × 35%	330 inquiries
Staff time those took	330 × 5 ÷ 60	27.5 hours
Less 20% for review and handoffs	27.5 × 80%	22 hours back
Value of that time (customer service rep, $29/h loaded)	22 × $29	$638
After-hours new-job requests	600 × 40% × 30%	72 a month
Jobs recovered after hours (40% lost for good; 40% of those book)	72 × 40% × 40%, rounded down	11 jobs
Missed calls that are new-job requests	600 × 14% × 30%	25.2 a month
Jobs recovered from missed calls (40% lost; agent completes 20%; 40% book)	25.2 × 40% × 20% × 40%	0.8 jobs
Added revenue	(11 + 0.8) × $341	$4,023
Gross profit from recovered jobs	11 × $341 × 40% + 0.8 × $341 × 40%	$1,609
Monthly value	$638 + $1,609	$2,247
Annual value	$2,247 × 12	$26,964

Upside case
Line item	Formula	Value
Resolved by the agent (35% of calls, 50% online)	600 × 35% + 600 × 50%	510 inquiries
Staff hours back after review	510 × 5 ÷ 60 × 80%	34 hours
Value of that time	34 × $29	$986
Jobs recovered after hours (60% lost; 60% book)	72 × 60% × 60%, rounded down	25 jobs
Jobs from missed calls (60% lost; agent completes 35%; 60% book)	25.2 × 60% × 35% × 60%	3.1 jobs
Added revenue	(25 + 3.1) × $341	$9,582
Gross profit	25 × $341 × 40% + 3.1 × $341 × 40%	$3,832
Monthly value	$986 + $3,832	$4,818
Annual value	$4,818 × 12	$57,816

Assumptions

The agent resolves 35% of online conversations and 20% of phone calls with no staff touch (base); 50% and 35% in the upside case.
5 staff minutes per routine conversation; customer service rep time at $29 an hour loaded ($21.53 median × 1.35).
14% of calls reach no one, 40% of online inquiries arrive after hours and 30% of inquiries are new-job requests. These describe the business, so they stay the same in both cases.
Of the new-job requests nobody answers, 40% are lost for good and 40% of those would have booked (upside 60% and 60%). On calls, the agent completes 20% (upside 35%).
Average plumbing service call $341 (revenue) at a 40% service gross margin; only gross profit is counted.
Not counted: estimate follow-ups, maintenance-plan renewals, review requests.

Sources

lab-01 lab-06 cx-09 cx-10 cx-11 cx-16 cx-34 cx-27 cx-22 cx-23 cx-26 cx-25 cx-21 cx-30 pnl-16

Modeled estimate Illustrative example

AI Sales Agent

Wrenfield Commercial Cleaning

Fictional janitorial company for offices, medical clinics and schools · 230 employees

The AI Sales Agent builds the prospect list, researches each business and drafts a specific first email; a rep reviews and approves every one, and sending stays within deliverability and CAN-SPAM limits.

31 to 36 hours back a month

All of it from team time

$1,333 × 12 = $15,996Upside $1,548 × 12 = $18,576

Today

Researched first emails a month: 540
Rep time per email: 10 minutes (research and writing)
Rep time on prospecting: 90 hours a month (30 per rep, about 19% of each rep's month)
Typical contract: $1,000 a month

With the agent

Rep time per email: 6.5 minutes to review and approve (upside 6)
Rep time on prospecting: 58.5 hours a month (upside 54)
Hours back: 31 a month across the three reps (upside 36)
Pipeline this outreach realistically yields: About 4.5 to 9.7 held meetings and 1.8 to 4.3 qualified opportunities a year

Where the value comes from
Source	Base	Upside
Rep time on prospecting (Team time)	$15,996	$18,576

The math

Base case
Line item	Formula	Value
Prospecting time today	540 × 10 ÷ 60	90 hours
Time per email with the agent	10 × (1 − 35%)	6.5 minutes
Prospecting time after	540 × 6.5 ÷ 60	58.5 hours
Hours back (rounded down)	90 − 58.5	31 hours
Value of time (services sales rep, $43/h)	31 × $43	$1,333
Annual value	$1,333 × 12	$15,996
For context: held meetings a year from this outreach	540 × 12 × 0.7 ÷ 1,000	About 4.5
For context: qualified opportunities a year	4.5 × 40%	About 1.8

Upside case
Line item	Formula	Value
Time per email with the agent (40% faster)	10 × (1 − 40%)	6 minutes
Hours back	90 − 540 × 6 ÷ 60	36 hours
Value of time	36 × $43	$1,548
Annual value	$1,548 × 12	$18,576
For context: upside held meetings a year	540 × 12 × 1.5 ÷ 1,000	About 9.7
For context: upside qualified opportunities a year	9.7 × 45%	About 4.3

Assumptions

10 minutes of manual research and writing per personalized first email; AI cuts the rep's time by 35% (upside 40%, the figure measured in a writing trial).
Rep time at a services sales rep's loaded $43 an hour ($33.65 median × 1.30).
Cold email is a slow channel: about 0.7 held meetings per 1,000 prospects for a quality-first approach (upside 1.5), and 40% of held meetings become a qualified opportunity (upside 45%). The reps send this volume today, so the agent changes how long each email takes, not how many go out: the pipeline is shown for context and not counted in the value.
Contract value $1,000 a month (small office), first-year value only if one closes. The calculator models new contracts with your own deal size and margin.
Not counted: any contracts won, renewals, referrals and upsells.

Sources

sales-20 sales-21 ai-03 sales-11 lab-01 lab-07 lab-15 sales-04 sales-07 sales-09 sales-22 sales-27 sales-16

04How we estimate

The rules behind every number.

Our rules

01
Conservative by default.
When an assumption could go either way, we take the one that makes the savings smaller.
02
Specific tasks, not whole jobs.
Agents take on defined workflows, like routine replies or data entry, and people keep the judgment calls. Time spent reviewing their work comes off first.
03
Fully loaded cost.
Staff time is valued at the median U.S. wage for the role plus payroll taxes and benefits, from BLS data. Nothing is added for office space or overhead.
04
Hours aren’t dollars until they are.
Time back is capacity. The calculator and the worked examples value it at what it costs you; the industry scenarios count half, because not every freed hour becomes paid work.
05
Gross profit, not revenue.
Extra jobs, patients or orders count only at their margin, never at their full price.
06
Before our fees.
Estimates don’t subtract our fees, because fees depend on scope. Your proposal shows net savings and payback.

How we model

The savings calculator

Counts two kinds of work: repetitive desk tasks handed to agents, and routine customer inquiries an agent resolves on its own. It values only the time returned, at the loaded hourly cost you enter.

Desk hours = team × hours a week × share × 80% × 50 ÷ 12Inquiry hours = inquiries × 5 min × 20% ÷ 60Value a year = hours back × 12 × hourly costCapacity = hours back × 12 ÷ 1,880

Worked examples, one per service

A fictional business for each service. Task time savings use base rates from published studies, already net of review. Work an agent completes on its own, like a customer conversation, has 20% set aside for review and handoffs.

Monthly value = hours back × loaded cost + gross profit or savingsAnnual value = monthly value × 12, at 4 working weeks a month

Industry scenarios

Six fictional businesses from 30 to 250 staff, with the agents each would use. Freed hours count at half their cost, and new business only at its gross margin.

Annual value = hours back × loaded cost × 50% + gross profit + cost savings

Key assumptions

Every value is the base case from the evidence, or the conservative one when a claim leads a page.

Key assumptions, their values and sources
Assumption	Value	Sources
Savings calculator
Share of the repetitive work agents take on (high case: 75%)	50% of repetitive hours	ai-15 ai-21 ai-22
Human review allowance, subtracted from the hours agents take on	20% of agent-handled hours	ai-20 ai-21
Evidence ceiling on routine-admin hours returned per worker, all agent tasks combined (high case: 4.9)	2 hours per worker per week	time-01 ai-15 time-05
Weeks per year the repetitive tasks recur (annualizing weekly hours)	50 weeks per year	lab-06 lab-09 lab-11
Default loaded hourly cost: routine office and admin work at 50+ staff (median wage × 1.40)	$30 per hour worked	lab-01 lab-06
Hours actually worked per full-time employee per year (for FTE capacity)	1,880 hours per year	lab-04 lab-05 lab-11
Inquiries an AI agent resolves end to end without staff: phone rates applied to every channel (high case: 35%)	20% of inquiries	cx-16 cx-09 cx-10
Staff minutes per routine inquiry, including wrap-up and logging (high case: 7)	5 minutes per inquiry	cx-34 cx-10
Inquiries that are new-business opportunities (high case: 38%)	30% of inquiries	cx-23 cx-26
Missed opportunities lost for good without an instant answer (high case: 60%)	40% of missed opportunities	cx-21 cx-26
Booking rate when a lead reaches someone promptly (high case: 60%)	40% of leads	cx-25 cx-23 cx-24 cx-26
Default share of inquiries that arrive after hours or go unanswered (you can change it)	30% of inquiries	cx-27 cx-21 cx-26
Default gross margin: the base case for service-and-repair work (you can change it)	40% of revenue	pnl-16
Cost per lead reduction from 60 to 90 days of disciplined optimization (high case: 20%)	10% of ad spend	mkt-19 mkt-25 mkt-24 mkt-15 mkt-16 mkt-14 mkt-01
Positive replies to researched outreach (high case: 60)	35 per 10,000 prospects	sales-07 sales-08 sales-09 sales-10 sales-01
Positive replies that become a held meeting (high case: 25%)	20% of positive replies	sales-04 sales-05 sales-09 sales-22
Held meetings that become a qualified opportunity (high case: 45%)	40% of meetings	sales-22
Opportunities closed and won (high case: 30%)	20% of opportunities	sales-23 sales-24 sales-22 sales-25
Realized price improvement, 0.5% of revenue (high case: 1%, or $100)	$50 per $10,000 of revenue	pnl-01 pnl-30
Vendor and materials savings, 0.75% of cost of sales (high case: 1.5%, or $150)	$75 per $10,000 of cost of sales	pnl-18 pnl-17
Default operating margin, before interest and taxes (you can change it)	8% of revenue	pnl-08 pnl-16
Scenarios and worked examples
Loaded cost: BLS median wage for the role × payroll taxes and benefits (scenarios and worked examples)	× 1.35 for 5–50 staff× 1.40 for 50–500 staff× 1.30 for sales roles	lab-01 lab-06 lab-07
Conversations an AI agent fully resolves without staff (scenarios)	35% of chats, emails and texts20% of phone calls	cx-09 cx-10 cx-16
Share of freed hours counted as value (scenarios)	50% of loaded cost	ai-08 ai-35

Each scenario and worked example lists its own assumptions and source ids next to its math.

Sources

Outside statistics appear only when we can cite them. Otherwise they stay out.

113Every source behind our numbersLabor costs (9) · Time on repetitive work (4) · AI time savings (12) · Customer conversations and lead response (17) · Paid marketing (11) · Sales outreach (16) · Industry margins and cost-saving levers (19) · Consulting costs (25)Show all

Labor costs

lab-01BLS Occupational Employment and Wage Statistics (OEWS), news release Table 1, May 2025 (median annual and percentiles from the BLS Public Data API), U.S. Bureau of Labor Statistics, May 2025 reference period (released Friday, May 15, 2026) bls.gov (opens in a new tab)
lab-04BLS OEWS Technical Note, May 2025 estimates; Table 1 footnote, U.S. Bureau of Labor Statistics, 2026 (May 2025 estimates) bls.gov (opens in a new tab)
lab-05BLS Employer Costs for Employee Compensation (ECEC), June 2026 news release, U.S. Bureau of Labor Statistics, June 2026 (released Wednesday, September 9, 2026) bls.gov (opens in a new tab)
lab-06BLS ECEC June 2026, Table 6 (private industry workers by establishment size and industry group), U.S. Bureau of Labor Statistics, June 2026 (released September 9, 2026) bls.gov (opens in a new tab)
lab-07BLS ECEC June 2026, Table 4 (private industry workers by occupational and industry group), U.S. Bureau of Labor Statistics, June 2026 (released September 9, 2026) bls.gov (opens in a new tab)
lab-09BLS ECEC Technical Note (2026 Q2); BLS Handbook of Methods, National Compensation Measures: Concepts, U.S. Bureau of Labor Statistics, 2026 bls.gov (opens in a new tab)
lab-11BLS chart: Paid leave benefits: Average number of sick and vacation days by length of service requirement (NCS), U.S. Bureau of Labor Statistics, March 2026 reference period (release date not shown on the page) bls.gov (opens in a new tab)
lab-15The Bridge Group, 2025 SDR Models, Motions & Metrics Report (10th edition), The Bridge Group, Inc. (sales consulting firm), 2025 (published February 6, 2025) bridgegroupinc.com (opens in a new tab)
lab-22Go Fractional, Fractional CFO Cost & Rates live benchmark, Go Fractional (fractional-talent marketplace), 2026 (page updated 2026-10-13 00:17 UTC, 'Last updated October 13, 2026') gofractional.com (opens in a new tab)

Time on repetitive work

time-01Dillon, Jaffe, Immorlica & Stanton, 'Shifting Work Patterns with Generative AI', arXiv 2504.11436v4 (13 Nov 2025, the latest version); also NBER Working Paper 33795 (May 2025), arXiv / National Bureau of Economic Research (authors at Microsoft Research and Harvard Business School), 2025 arxiv.org (opens in a new tab)
time-02Microsoft Work Trend Index 2023 Annual Report, 'Will AI Fix Work?' (May 9, 2023), footnote 2, Microsoft WorkLab, 2023 microsoft.com (opens in a new tab)
time-05Asana, Anatomy of Work Global Index 2023, press release (Business Wire, March 8, 2023), as reproduced by StockTitan, Asana (press release via Business Wire), 2023 stocktitan.net (opens in a new tab)
time-06APQC General Accounting Open Standards Benchmarking, reported in CFO.com 'Metric of the Month' (Perry D. Wiggins, March 5, 2018), APQC via CFO.com, 2018 (article date; data-collection period not stated) cfo.com (opens in a new tab)

AI time savings

ai-03Noy & Zhang, 'Experimental evidence on the productivity effects of generative artificial intelligence', Science 381(6654):187-192 (July 2023), AAAS (Science); abstract verified via the Europe PMC record (PMID 37440646), 2023 ebi.ac.uk (opens in a new tab)
ai-06Dell'Acqua, McFowland, Mollick et al., 'Navigating the Jagged Technological Frontier', Organization Science 37(2):403-423 (March 2026), INFORMS (Organization Science); abstract verified via the Crossref record, 2026 api.crossref.org (opens in a new tab)
ai-08Humlum & Vestergaard, 'Still Waters, Rapid Currents: Early Labor Market Transformation under Generative AI', NBER WP 33777 (May 2025, revised March 2026); previously circulated as 'Large Language Models, Small Labor Market Effects' (BFI WP 2025-56), NBER (revision); University of Chicago Becker Friedman Institute (May 2025 version), 2026 nber.org (opens in a new tab)
ai-15Dillon, Jaffe, Immorlica & Stanton, 'Shifting Work Patterns with Generative AI', NBER WP 33795 (May 2025, revised November 2025), NBER, 2025 nber.org (opens in a new tab)
ai-16Dell'Acqua et al., 'The Cybernetic Teammate: A Field Experiment on Generative AI Reshaping Teamwork and Expertise', NBER WP 33641 (April 2025), NBER, 2025 nber.org (opens in a new tab)
ai-20Cambon et al., 'Early LLM-based Tools for Enterprise Information Workers Likely Provide Meaningful Boosts to Productivity', Microsoft Research MSR-TR-2023-43 (December 2023), Microsoft Research (vendor), 2023 microsoft.com (opens in a new tab)
ai-21UK Department for Business and Trade, 'The Evaluation of the M365 Copilot Pilot in the Department for Business and Trade' (August 2025), Table 5, UK Government (GOV.UK), 2025 assets.publishing.service.gov.uk (opens in a new tab)
ai-22UK Government Digital Service, 'Microsoft 365 Copilot Experiment: Cross-Government Findings Report' (published 2 June 2025), UK Government Digital Service (GOV.UK), 2025 gov.uk (opens in a new tab)
ai-23Fitzpatrick, Kelly, Carey, Walsh & Nugent (Central Bank of Ireland), 'Assessing Generative AI value in a public sector context: evidence from a field experiment', arXiv 2502.09479 (February 2025), arXiv, 2025 arxiv.org (opens in a new tab)
ai-24Schwarcz, Manning, Barry, Cleveland, Prescott & Rich, 'AI-Powered Lawyering: AI Reasoning Models, Retrieval Augmented Generation, and the Future of Legal Practice', Berkeley Law (paper archive copy), 2025 law.berkeley.edu (opens in a new tab)
ai-26Fang, Yuan, Zhang, Donati & Sarvary, 'Generative AI and Sales Productivity: Field Experiments in Online Retail', arXiv 2510.12049 v6 (29 June 2026), arXiv, 2026 arxiv.org (opens in a new tab)
ai-35Suh & Oh, 'Generative AI and the Reallocation of Time: Productivity, Leisure, and Fulfilling Work', arXiv 2602.12695 (February 13, 2026), arXiv, 2026 arxiv.org (opens in a new tab)

Customer conversations and lead response

cx-09SaaStr (Jason Lemkin), reporting Gorgias's public AI Agent Benchmark (article published Oct 6, 2026), SaaStr, 2026 saastr.com (opens in a new tab)
cx-10Same Alibaba field-experiment paper (arXiv 2605.14830), section 3 and Table 1, arXiv working paper, 2026 (data 2024) arxiv.org (opens in a new tab)
cx-11Gartner press release (Aug 19, 2024), Gartner, 2024 gartner.com (opens in a new tab)
cx-16Gartner press release (July 9, 2024), Gartner, 2024 gartner.com (opens in a new tab)
cx-18Oldroyd, McElheran & Elkington, 'The Short Life of Online Sales Leads', Harvard Business Review (March 2011), Harvard Business Review, 2011 hbr.org (opens in a new tab)
cx-19Oldroyd (MIT) & Elkington (InsideSales.com), 'Lead Response Management Study', MarketingSherpa B2B Demand Generation Summit (Oct 16, 2007); PDF copy hosted by mortech.com, InsideSales.com / Dr. James Oldroyd, 2007 mortech.com (opens in a new tab)
cx-20Hennessey Digital, 2025 Lead Form Response Time Study, Hennessey Digital, 2025 hennessey.com (opens in a new tab)
cx-21Clio 2024 Legal Trends Report, as summarized by the Illinois Supreme Court Commission on Professionalism (2Civility), 2Civility / Clio, 2024 2civility.org (opens in a new tab)
cx-22CallRail ebook, 'From Conversations to Conversions: How Small Businesses Can Market Smarter' (released January 2025), CallRail, 2025 cdn.mediavalet.com (opens in a new tab)
cx-23Invoca Home Services Lead Conversion Benchmarks Report 2026, Invoca, 2026 invoca.com (opens in a new tab)
cx-24Invoca Call Conversion Industry Benchmarks Report 2025 (press release, June 10, 2025), Invoca (via Seeking Alpha / PR Newswire), 2025 seekingalpha.com (opens in a new tab)
cx-25ServiceTitan blog: 'Typical call-booking rate for ServiceTitan users? The number might shock you into action' (Oct 25, 2022), ServiceTitan, 2022 servicetitan.com (opens in a new tab)
cx-26Patient Prism, Dental Patient Access Report (July 2, 2026), Patient Prism, 2026 (data CY2025) patientprism.com (opens in a new tab)
cx-27TechTarget Patient Engagement, reporting Zocdoc's inaugural 'What Patients Want' report (Dec 13, 2023), TechTarget / Zocdoc, 2023 techtarget.com (opens in a new tab)
cx-30HomeAdvisor (Angi) cost guide: Hire a plumber (updated Jun 19, 2026), HomeAdvisor / Angi Inc., 2026 homeadvisor.com (opens in a new tab)
cx-31AHRQ MEPS Statistical Brief #555 (Olaisen & Manski, March 2024), Agency for Healthcare Research and Quality, 2024 (data 2019-2021) meps.ahrq.gov (opens in a new tab)
cx-34LiveChat Customer Service Report 2019 (PDF), LiveChat (Text Inc.), 2020 (data CY2019) livechat.com (opens in a new tab)

Paid marketing

mkt-01WordStream by LocaliQ, Google Ads Benchmarks 2026, WordStream by LocaliQ (USA TODAY Co.), 2026 (campaigns Apr 1, 2025 - Mar 31, 2026; page updated Sep 16, 2026) wordstream.com (opens in a new tab)
mkt-02WordStream by LocaliQ, Google Ads Benchmarks 2025, WordStream by LocaliQ, 2025 (campaigns Apr 1, 2024 - Mar 31, 2025) wordstream.com (opens in a new tab)
mkt-05Triple Whale, Facebook Ad Benchmarks by Industry (Updated 2026 Data), Triple Whale, 2026 (Aug 1, 2025 - Jul 31, 2026) triplewhale.com (opens in a new tab)
mkt-06Triple Whale, Google Ads Benchmarks by Industry (Updated 2026 Data), Triple Whale, 2026 (Aug 1, 2025 - Jul 31, 2026) triplewhale.com (opens in a new tab)
mkt-14Kohavi, Deng & Vermeer, 'A/B Testing Intuition Busters', KDD 2022 (Table 2), ACM SIGKDD, 2022 airbnb.tech (opens in a new tab)
mkt-15Optimizely, 'Top 10 takeaways from running 173,000 experiments', Optimizely, 2026 optimizely.com (opens in a new tab)
mkt-16Browne & Swarbrick Jones (Qubit), 'What works in e-commerce - a meta-analysis of 6700 online experiments' (2017), Qubit Digital Ltd (methodology assured by PwC UK), 2017 gwern.net (opens in a new tab)
mkt-18WordStream study as reported by Search Engine Watch, 'Are You Wasting 25% Of Your PPC Budget? New Study Says You Might Be' (Oct 9, 2013), Search Engine Watch (study by WordStream), 2013 searchenginewatch.com (opens in a new tab)
mkt-19Doctor Ads (Igor Ivitskiy), 'How Much of Your Google Ads Budget Is Wasted?', Doctor Ads (PPC consultancy), 2026 (data Sep 2024 - Feb 2025) thedoctorads.com (opens in a new tab)
mkt-24Google Ads & Commerce blog, 'Introducing AI Max for Search campaigns' (May 2025), Google, 2025 blog.google (opens in a new tab)
mkt-25Meta Newsroom, 'Introducing New Automation Tools to Increase Sales and Drive Growth' (Aug 2022), Meta, 2022 about.fb.com (opens in a new tab)

Sales outreach

sales-01Hunter, State of Email Outreach 2026 Report, Hunter.io (Hunter Web Services, Inc.), 2026 report (2025 data) hunter.io (opens in a new tab)
sales-04Belkins, 'What are B2B cold email response rates? Belkins' 2026 study' (updated 2026-06-26), Belkins (US B2B lead-generation agency), 2026 study (2025 data) belkins.io (opens in a new tab)
sales-05Sales.co, 'Cold Email Response Rates in 2026: The Real Numbers From 1,288,605 Sent Emails' (updated Aug 17, 2026), Sales.co, 2026 (Oct 2025 - Aug 2026 data) sales.co (opens in a new tab)
sales-07Sales.co, 'Cold email is dying and there is nothing you can do about it', Sales.co, 2026 (published Sep 27, 2026) sales.co (opens in a new tab)
sales-08Sales.co, 'Personalization no longer matters for cold email', Sales.co, 2026 (published Sep 27, 2026) sales.co (opens in a new tab)
sales-09OneAway, 'We Analyzed 1,533,399 Cold Emails. Here's What Actually Gets a Reply in 2026.', OneAway, Published June 2, 2026 (cumulative data as of June 2026; start date not stated) oneaway.io (opens in a new tab)
sales-10Sahni, Wheeler & Chintagunta, 'Personalization in Email Marketing: The Role of Non-Informative Advertising Content', Marketing Science 37(2), Stanford GSB (publication page) / INFORMS Marketing Science, 2018 (Feb 27, 2018) gsb.stanford.edu (opens in a new tab)
sales-11Noy & Zhang, 'Experimental evidence on the productivity effects of generative artificial intelligence', Science (July 2023), doi:10.1126/science.adh2586; abstract via Stanford SCALE, Science (AAAS); MIT, 2023 scale.stanford.edu (opens in a new tab)
sales-16FTC, CAN-SPAM Act: A Compliance Guide for Business, Federal Trade Commission, 2025 inflation-adjusted level, unchanged for 2026 ftc.gov (opens in a new tab)
sales-20Salesforce, State of Sales 7th edition (announcement, Feb 3, 2026), Salesforce Research, 2026 (survey Aug-Sep 2025) salesforce.com (opens in a new tab)
sales-21Salesforce, State of Sales 7th edition (report PDF), Salesforce Research, 2026 (survey Aug-Sep 2025) salesforce.com (opens in a new tab)
sales-22RevOps Squared and Revenue.io, 2021 Customer Acquisition Benchmarks Report, RevOps Squared / Revenue.io, 2021 report (survey 2020) knowledge.revenue.io (opens in a new tab)
sales-23Ebsta and Pavilion, 2023 B2B Sales Benchmarks Report, Ebsta, 2023 report (2022 data) ebsta.com (opens in a new tab)
sales-24Ebsta, 2024 B2B Sales Benchmarks: H1 Update, Ebsta, 2024 ebsta.com (opens in a new tab)
sales-25Loopio, 2025 RFP Response Trends & Benchmarks Report, Loopio, 2025 report (2024 data) link.loopio.com (opens in a new tab)
sales-27ISSA, 'Commercial Cleaning Rates per Square Foot: Full Pricing Guide' (ISSA Staff), ISSA (worldwide cleaning industry association), 2025 (Dec 4, 2025) issa.com (opens in a new tab)

Industry margins and cost-saving levers

pnl-01Marn, Roegner and Zawada, 'The power of pricing', The McKinsey Quarterly 2003 Number 1, Exhibit 1 'The power of one', McKinsey & Company (The McKinsey Quarterly); PDF copy hosted on a University of Colorado faculty course page, 2003 leeds-faculty.colorado.edu (opens in a new tab)
pnl-02Clio, Legal Trends Report, 'Law Firm KPIs / Key Performance Indicators & Benchmarks' page (2025 figures), Clio (Themis Solutions Inc.), 2025 clio.com (opens in a new tab)
pnl-04Zylo 2026 SaaS Management Index (report page), Zylo license-waste solution page and Zylo 'SaaS statistics' blog, Zylo (SaaS management software vendor), 2026 (2025 data) zylo.com (opens in a new tab)
pnl-05Help Net Security, '53% of SaaS licenses remain unused' (reporting Productiv's 2023 State of SaaS data), Help Net Security, reporting Productiv (SaaS management vendor) data, 2023 (article dated July 4, 2023) helpnetsecurity.com (opens in a new tab)
pnl-08Quarterly Financial Report for Manufacturing, Mining, Trade, and Selected Service Industries, 2026 Quarter 2 (QFR/26-Q2), Table D -- Profits per Dollar of Sales, U.S. Census Bureau, 2026 (issued September 2026) census.gov (opens in a new tab)
pnl-09Annual Retail Trade Survey 2022: Estimated Annual Gross Margin as a Percentage of Sales of U.S. Retail Firms by Kind of Business: 1993-2022 (gmper.xlsx); Estimated Annual Operating Expenses (exp.xlsx); Estimated Annual Sales (sales.xlsx), U.S. Census Bureau, 2022 data www2.census.gov (opens in a new tab)
pnl-10ADA Health Policy Institute, 'Income, gross billings, expenses, characteristics, selected staff wages: selected 2025 results from the Survey of Dental Practice' (XLSX), Tables 5 and 13, American Dental Association, Health Policy Institute, 2026 (2025 data) ada.org (opens in a new tab)
pnl-11National Restaurant Association, 2025 Restaurant Operations Data Abstract (2024 data), summarized in 'Restaurant operators kept food cost ratios in check in 2024', National Restaurant Association, 2025 edition, 2024 data (article dated September 10, 2025) restaurant.org (opens in a new tab)
pnl-12National Restaurant Association, 2025 Restaurant Operations Data Abstract, summarized in 'Elevated labor costs had a significant impact on restaurant profitability in 2024', National Restaurant Association, 2025 edition, 2024 data (article dated October 8, 2025) restaurant.org (opens in a new tab)
pnl-15Clowes, Hanson and Swannell, 'The Business Case for Reducing Food Loss and Waste: Restaurants', a report on behalf of Champions 12.3 (WRI and WRAP), Champions 12.3 / World Resources Institute / WRAP, 2019 (February 2019) champions123.org (opens in a new tab)
pnl-16IRS Statistics of Income, Corporation Complete Report, Table 5.1: Returns of Active Corporations, Balance Sheet, Income Statement, Tax, and Selected Other Items, by Minor Industry, Tax Year 2022, Internal Revenue Service, Statistics of Income Division, Tax year 2022 irs.gov (opens in a new tab)
pnl-17GAO-13-417, 'Strategic Sourcing: Leading Commercial Practices Can Help Federal Agencies Increase Savings When Acquiring Services', U.S. Government Accountability Office, 2013 (published April 15, 2013; publicly released May 15, 2013) gao.gov (opens in a new tab)
pnl-18Matthew Grennan and Ashley Swanson, 'Transparency and Negotiated Prices: The Value of Information in Hospital-Supplier Bargaining', NBER Working Paper 22039 (2016); Journal of Political Economy 128(4), 2020, pp. 1234-1268, NBER / Journal of Political Economy, 2020 (working paper February 2016) nber.org (opens in a new tab)
pnl-25Kesavan, Lambert, Williams and Pendem, 'Doing Well by Doing Good: Improving Retail Store Performance with Responsible Scheduling Practices at the Gap, Inc.', Management Science 68(11), 2022, pp. 7818-7836 (abstract via RePEc); Williams et al., 2018 Stable Scheduling Study report, Management Science (INFORMS), peer-reviewed; University of Chicago / UC Hastings WorkLife Law (report), 2022 (report 2018) ideas.repec.org (opens in a new tab)
pnl-26BLS CES series CEU3133200009 (average weekly overtime hours, production and nonsupervisory employees, fabricated metal product manufacturing) and CEU3133200007 (average weekly hours); BLS ECEC series CMU2103000000000D/P (overtime and premium, manufacturing) and CMU2023000000000D (wages and salaries, manufacturing), U.S. Bureau of Labor Statistics, 2026 data.bls.gov (opens in a new tab)
pnl-28BLS PPI series PCU484121484121, 'PPI industry data for General freight trucking, long-distance TL, not seasonally adjusted', U.S. Bureau of Labor Statistics, 2026 data.bls.gov (opens in a new tab)
pnl-29BLS PPI series WPU1017 (Metals and metal products - Steel mill products) and PCU332312332312 (Fabricated structural metal mfg), U.S. Bureau of Labor Statistics, 2026 data.bls.gov (opens in a new tab)
pnl-30Michael V. Marn and Robert L. Rosiello, 'Managing Price, Gaining Profit', Harvard Business Review, September-October 1992, Exhibit 1, Harvard Business Review (authors from McKinsey & Company); PDF copy hosted on a UNC Charlotte course page, 1992 pages.charlotte.edu (opens in a new tab)
pnl-31National Retail Federation and Happy Returns, 2025 Retail Returns Landscape (press release), National Retail Federation / Happy Returns (a UPS company), 2025 (press release dated October 15, 2025) nrf.com (opens in a new tab)

Consulting costs

cons-06GSA CALC+ API, keyword 'Ernst', U.S. General Services Administration, 2026 (data as of 2026-10-12) api.gsa.gov (opens in a new tab)
cons-07GSA CALC+ API, keyword 'KPMG', U.S. General Services Administration, 2026 (data as of 2026-10-12) api.gsa.gov (opens in a new tab)
cons-09GSA CALC+ API, keyword 'FTI Consulting', U.S. General Services Administration, 2026 (data as of 2026-10-12) api.gsa.gov (opens in a new tab)
cons-11GSA CALC+ API, keyword 'management consultant', filter business_size S / O; full download filtered on labor_category, U.S. General Services Administration, 2026 (data as of 2026-10-12) api.gsa.gov (opens in a new tab)
cons-12NYS OGS Award 23269 Attachment 1 - Pricing, McKinsey & Company, Inc (last updated 6/16/2025), New York State Office of General Services, 2025 online.ogs.ny.gov (opens in a new tab)
cons-13NYS OGS Award 23269 Attachment 1 - Pricing, The Boston Consulting Group, Inc. (last updated 6/16/2025), New York State Office of General Services, 2025 online.ogs.ny.gov (opens in a new tab)
cons-14NYS OGS Award 23269 Attachment 1 - Pricing, KPMG LLP (last updated 10/7/2025), New York State Office of General Services, 2025 online.ogs.ny.gov (opens in a new tab)
cons-15NYS OGS Award 23269 Attachment 1 - Pricing, Ernst & Young LLP (last updated 6/16/2025), New York State Office of General Services, 2025 online.ogs.ny.gov (opens in a new tab)
cons-16NYS OGS Award 23269 Attachment 1 - Pricing (Deloitte Consulting LLP; Accenture LLP), New York State Office of General Services, 2025-2026 online.ogs.ny.gov (opens in a new tab)
cons-18Application of the Official Committee of Unsecured Creditors to retain FTI Consulting, Inc. as financial advisor (Doc 385), U.S. Bankruptcy Court, E.D. Missouri (via CourtListener RECAP), 2025 storage.courtlistener.com (opens in a new tab)
cons-25McKinsey & Company, National Cooperative Purchasing Alliance 'Strategic Management Consulting' price quotation, Exhibit 4: McKinsey team bundle rates, McKinsey & Company (document hosted by Ohio procurement, procure.ohio.gov price lists), 2021 (Year 1 prices) dam.assets.ohio.gov (opens in a new tab)
cons-26Austin Jenkins, '$165K a week for COVID analysis: That's how much WA paid a global consulting firm', Northwest News Network, Aug 10, 2020, Northwest News Network (public radio), 2020 nwnewsnetwork.org (opens in a new tab)
cons-27Arkansas Department of Transformation and Shared Services news release, Feb 9, 2024, State of Arkansas, 2024 sas.arkansas.gov (opens in a new tab)
cons-28Houston Chronicle reporting as quoted by Off the Kuff (Charles Kuffner, Mar 31, 2025); scope from Community Impact (Kevin Vu, Feb 12, 2025), Houston Chronicle (via Off the Kuff); Community Impact, 2025 offthekuff.com (opens in a new tab)
cons-29Nick Castele, 'Cleveland to hire Ernst & Young to audit how City Hall works', Signal Cleveland, Apr 17, 2023, Signal Cleveland, 2023 signalcleveland.org (opens in a new tab)
cons-30Rod McGuirk, 'Deloitte to partially refund Australian government for report with apparent AI-generated errors', Associated Press, Oct 7, 2025 (KSAT syndication), Associated Press, 2025 ksat.com (opens in a new tab)
cons-31Minutes of the Ogden City Council regular meeting, Dec 16, 2025, Ogden City Council (Utah Public Notice Website), 2025 utah.gov (opens in a new tab)
cons-32Stephi Smith, 'Consultant recommends Jefferson City establish grants department', News Tribune, Aug 16, 2024, Jefferson City News Tribune, 2024 newstribune.com (opens in a new tab)
cons-33'Village Considers Hiring Consultants to Conduct an Organizational Evaluation of the Assessor's Office', Scarsdale10583, Mar 1, 2017, Scarsdale10583 (local news), 2017 scarsdale10583.com (opens in a new tab)
cons-34Workplace Strategy Consulting proposal, Waterbury Select Board meeting packet (May 2026), Town of Waterbury, Vermont, 2026 waterburyvt.com (opens in a new tab)
cons-36Go Fractional, 'Fractional COO Cost & Rates', Go Fractional (fractional-talent marketplace), 2026 gofractional.com (opens in a new tab)
cons-37Riverside County Board of Supervisors, Item 3.12, June 9, 2026, with Amendment No. 1, County of Riverside, California, 2026 media.rivcocob.org (opens in a new tab)
cons-38K. Michelle Moran, 'Grosse Pointe Park hopes to find cost savings through utilities review', C&G Newspapers, Feb 25, 2026, C&G Newspapers, 2026 candgnews.com (opens in a new tab)
cons-39Cost and Revenue Review Agreement, NYCOM USA Program (rev. 09-30-2024), Village of Croton-on-Hudson, NY (meeting attachment, Nov 2024), 2024 play.champds.com (opens in a new tab)
cons-41America's SBDC, 'About Us', America's SBDC (Association of Small Business Development Centers), 2026 (page checked Oct 12, 2026) americassbdc.org (opens in a new tab)

The fine print

Questions about a number? Ask us and we’ll walk you through it.Ask about the math

05Our approach to proof

Proof you can check yourself.

You won’t see

Testimonials or reviews
Client logos or names
Guaranteed ROI
Statistics without a source

You will see

Working demos of every agent
Assumptions next to every number
A written success measure for your pilot
Straight answers, including “AI isn’t the fix here”

See how a pilot works Watch the demos

Free AI assessment

Get an estimate built on your numbers.

On a free 30-minute call we size the opportunities task by task, then send a written summary with every assumption shown.

Book a free AI assessment See how we work

No cost, no obligation.

What you leave with

A written plan for your business

Your top three opportunities, ranked
By estimated value and effort, specific to how your business runs.
A rough savings estimate
A modeled estimate, with every assumption shown.
A suggested first step
The first job to hand an agent, and what it would take.

Yours to keep, whether or not we work together.

Check the math before you commit.

Start with your own numbers.

Team time01AI Agents$74,880 to $157,320a year · value of staff time

Customer inquiries02Customer Agent$36,720 to $103,824a year · gross profit and staff time

Ad spend03AI Marketing Director$12,000 to $24,000a year · budget freed to reallocate

Sales outreach04AI Sales Agent$6,000 to $24,000a year · gross profit on new deals

Pricing and vendors05AI Consulting$114,000 to $228,000a year · gross profit from prices and costs

Find a business like yours.

Maple Row Law Group

Fernway Home Goods

Gablecrest Roofing & Exteriors

Willowmere Dental Group

Bramblewood Restaurant Group

Ironvale Industrial Supply

One worked example for every service.

The rules behind every number.

Conservative by default.

Specific tasks, not whole jobs.

Fully loaded cost.

Hours aren’t dollars until they are.

Gross profit, not revenue.

Before our fees.

The savings calculator

Worked examples, one per service

Industry scenarios

Labor costs

Time on repetitive work

AI time savings

Customer conversations and lead response

Paid marketing

Sales outreach

Industry margins and cost-saving levers

Consulting costs

Proof you can check yourself.

Get an estimate built on your numbers.

A written plan for your business