Headline test
RunningGoogle Search: Dentist near me (one ad group per office)
- Current
- “Family Dentist in Eastgate | New Patients Welcome”
- New
- “Same-Week Appointments | Most Insurance Accepted”
AI Marketing Director
The AI Marketing Director audits every ad you run on Google, Meta and beyond, cuts the waste, runs its own tests and moves budget to what works, inside limits you set. Every week, it tells you what changed, in plain English.
Works for your marketer or agency, not instead of them. They approve strategy, creative and budget.
This week
The first audit checks for the patterns that quietly drain ad budgets, and puts a dollar figure on each one it finds.
Clicks from people looking for jobs, free advice or a service you don’t offer, paid for every month.
Ads shown outside the area you serve, at hours nobody can answer, or for products that are out of stock or too thin on margin to advertise.
Two campaigns bidding on the same searches and driving up your own costs.
The same ad shown to the same people too many times, with results sliding.
Slow pages, unclear offers and forms that ask too much.
Conversions counted twice or not at all, so every decision rests on bad data.
Google and Meta can both claim the same sale. We check their numbers against your real orders and booked jobs, and judge campaigns on margin, not revenue.
Every finding comes with the fix, and nothing changes without your approval.
The agent connects to the accounts, finds the waste, launches tests and moves budget, then reports what changed.
See all six demosCampaigns · last 28 days
Waste found: $4,364 per 4 weeks(18% of spend)Out-of-area clicksDentist near me
14% of clicks came from 15+ miles from any office. None of the 40 bookings did. $896.
Broad-match wasteDental implants
41% of spend ($2,132) matched searches that never book:
Creative fatigue$99 new-patient exam
Frequency 7.4. CTR down from 1.9% to 0.8% in 4 weeks.
Weekly cost per booking: $95.45 → $131.25 → $175.00 → $210.00
$1,336 spent above its week-1 cost per booking.
Booking = a new-patient appointment booked online or by phone. Est. ROAS = bookings × average first-year collections per new patient for that service ÷ spend (general $800, emergency $650, implants $3,600, clear aligners $2,000).
Experiments · 3 live
A test wins at 90% probability or better.
Headline test
RunningGoogle Search: Dentist near me (one ad group per office)
Audience test
RunningMeta: $99 new-patient exam (new ads)
Landing page test
RunningGoogle Search: Dental implants
Results · week 4
Winner threshold: 90%
Chance the new version books more patients per dollar. We call a winner at 90%.
Weekly budget · week 4 → week 5
Cap: ±20% per campaign per week
Queued for week 6: Dentist near me +12.5% to $2,160 (guardrail)
Cost per new-patient booking · weekly
Spend held at $6,000 a week
Experiment results · week 4
Guardrails· set by Stillbrook
Week 8 report · for Dr. Park
39 new-patient bookings at $154 each (week 1: 35 at $171). Same $6,000 a week.
What worked: cutting search waste and moving budget to the lowest cost per booking.
Still testing: the “Same-Week Appointments” headline and the implant consult page.
Next: a Saturday-hours message for Eastgate.
Budget moved
Results in
3 experiments live. A test wins at 90% probability or better.
Fixes live
Approved by Jess MorganMon 10:12 am · 3 changes logged
Audit: 6 campaigns, 1,912 search terms, 14 ads · 6 min
Connected (read-only): Google Ads · Meta Ads · practice-management bookings
Every budget move stayed within ±20% a week.
This demo needs JavaScript. Here's the step-by-step version instead.Motion is reduced, so every step is shown at once. Press Play to step through them one at a time.
Simulated demo. Stillbrook Dental Group and the people named are fictional. Figures are illustrative.
It does the analysis a good marketing lead would do, every day, across every account.
Every campaign, ad group, keyword, audience and search term across Google Ads, Meta Ads, Microsoft Ads and LinkedIn.
Connects ad data to your website analytics and CRM, so you see cost per booked job or sale, not just cost per click.
Checks that conversions are counted correctly and flags broken tags and double counting. Good decisions need good data.
Adds negative keywords, fixes location and schedule settings and pauses what’s losing money, with your approval.
Tests one change at a time (creative, audience, offer, landing page or budget split), each with a clear hypothesis and a rule for when to stop.
Waits for enough data before declaring a winner, so you don’t chase noise.
Shifts spend toward the campaigns bringing in customers, inside the caps and rules you approve.
Writes headlines, ad copy and landing page variants in your brand voice, for your approval.
A short weekly note on what changed, what we tested, what we learned and what’s next, plus a monthly view of cost per lead, cost per acquisition and return on ad spend.
Who does what
Your marketer, agency or you decide
The Marketing Director does the daily work
A steady weekly rhythm, with every change logged and every big decision left to you.
We connect your ad accounts, analytics and CRM with read-only access first.
Typically week 1
You getA full picture, nothing changed
A full audit with each finding sized in dollars, plus the baseline we measure against: cost per lead, cost per acquisition and return on ad spend.
Typically weeks 1 to 2
You getEvery leak, sized in dollars
With your approval, we fix tracking, cut clear waste and tighten targeting.
Typically weeks 2 to 3
You getClean tracking, less waste
A running test plan. Each experiment has a hypothesis, a budget and a stopping rule.
Ongoing, in weekly cycles
You getAnswers you can trust
Weekly plain-English notes and a monthly review. Budget moves toward what works, inside your caps.
Ongoing
You getA weekly note, a monthly review
How fast tests produce answers depends on your spend and traffic. We’d rather wait than call a winner on noise.
The agent works inside limits you set. The big decisions stay with you.
Set inside the ad platforms wherever they support it. The agent can’t raise them without your approval.
New campaigns, new offers, budget increases and new messaging all wait for your OK.
Every ad follows your brand voice and a list of approved claims. No invented promises, discounts or guarantees.
Customer lists never go to an ad platform without your approval.
Every change is logged with the reason and the result, and any change can be rolled back.
Special ad categories (housing, employment) and regulated fields (healthcare, legal) follow platform policy and your compliance review.
If results dip or something looks off, the agent flags it right away and holds risky changes until you’ve seen it.
Pause the agent at any time. It makes no further changes until you say so.
Ad improvements show up as money you stop wasting and more customers from the same budget. This example keeps spend flat, puts the savings to work and counts only the gross profit on the extra jobs.
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.
Modeled value: $146,844 a year in the base case, up to $332,568 in the upside case.
7 to 14 hours back a month
Where the value comes from: Marketing team time, $4,284 to $8,568; More leads at the same spend, $142,560 to $324,000.
$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
With the agent
| 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
| 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 |
| 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 |
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.
What moves the number
No. It works for them. Your marketer or agency keeps the strategy, the brand and the final say on creative and budget. The Marketing Director does the daily work they rarely have time for: checking every search term and placement, spotting tired ads, running clean tests and writing up what happened. If you don’t have a marketer, it can run the accounts day to day inside the limits you set, and the big calls still come to you.
Not without your approval. We set spending caps inside the ad platforms wherever they support it, and only you can approve an increase.
Audit findings and quick fixes typically arrive in the first two weeks. Tests take longer. How long depends on your spend and traffic, and we won’t call a winner before the data supports it.
We start with Google Ads and Meta Ads (Facebook and Instagram), and also work with Microsoft Ads and LinkedIn. We connect your website analytics and CRM too, so results tie to real customers and real orders.
Not without your approval. Some advertisers upload customer lists to find similar buyers or to measure results. That’s your decision, made with whoever handles privacy for you, and the agent never does it on its own.
There’s no fixed minimum. At smaller budgets the biggest gains usually come from fixing tracking, cutting waste and tightening targeting, rather than from running lots of tests. We’ll tell you on the assessment call which applies to you.
Free AI assessment
On a free 30-minute call we’ll talk through your accounts and tell you what we’d check first. Share read-only access beforehand and we’ll come with specifics.
Read-only access is optional, and you can remove it at any time.
What you leave with
By estimated value and effort, specific to how your business runs.
A modeled estimate, with every assumption shown.
The first job to hand an agent, and what it would take.
Yours to keep, whether or not we work together.