AI growth manager for
your Google Ads
GoodLads turns campaign performance into hypotheses to improve ROAS or CPA - applied in one click, never without your approval. Also in Claude Code and Codex
Built for accounts spending €10K to €250K a month
The work you hired us for.
Five jobs every account owner has - what you'd say, what GoodLads does about it, and what that looks like on screen.
GoodLads rolls every campaign's spend, conversions and ROAS into one account overview - a treemap sized by what each campaign spends and colored by how it does against the account average.
GoodLads reads your search terms, keyword quality, geography and audiences every day, and turns what it finds into a ranked feed of ideas - each naming the campaign and the spend at stake.
Every idea arrives as the safest lever that still works: a proper experiment where the channel allows one, a paused new ad where it doesn't, and a clearly marked live edit only when nothing else can do the job.
Every applied idea reports back to your board with lift versus control, a p-value and a plain verdict - you see whether it's working without opening Google Ads.
A winner becomes permanent on the live campaign in one click. A loser is retired just as cleanly - and either way, the learning feeds the next round of ideas.
Every hypothesis, tracked to a verdict.
Your hypotheses live on a kanban board - proposed, scheduled, live, completed. Each one states the observation behind it, the change it proposes, and the metric move it predicts - so you always know what's being tested and what it returned.
Proposed
3Fresh from your last 30 days
- Lead with free returns in the headlineCTR slid 14% in 30 days - a free-returns RSA in a 50/50 experiment should lift ROAS from 2.8 to 3.1 in 2 weeksROAS
- Short, mobile-first headlines68% of clicks are mobile and long headlines truncate - shorter headlines should lift ROAS from 3.1 to 3.3 in 3 weeksROAS
- Cheaper conversions at the same daily spend3 weeks under the CPA target - stepping tCPA from €24 to €22 should hold volume and cut CPA to €22 in 2 weeksCPA
Scheduled
2Approved, queued to go live
- Seasonal copy for the summer saleLast July's peak lifted CTR 22% - switching to seasonal copy should lift ROAS from 3.2 to 3.5 over the 4 sale weeksROAS
- Sitelinks with shipping costs and delivery timesTop exit reason is shipping surprise - adding cost and delivery sitelinks should cut CPA from €31 to €28 in 3 weeksCPA
Live
2Running as Google Ads experiments
- Social-proof headlines on non-brand SearchReview count won on brand - social proof should lift ROAS from 2.9 to 3.2 in 2 weeks. 50/50 experiment, day 9 of 14ROAS
- First step down on the tCPA target2 weeks of overdelivery - moving tCPA from €26 to €24 should cut CPA to €24 in 2 weeks. One click to roll backCPA
Completed
3Measured, decided, learned from
- Benefit-led headlines beat the price-led controlHypothesis: ROAS from 2.9 to 3.2 in 2 weeks. Measured +14%, so the winner took the budgetROAS+14% ROAS · Made full-on
- Lower tCPA held conversion volumeHypothesis: CPA from €27 to €24 in 3 weeks at flat volume. Measured -11%, next step now proposedCPA-11% CPA · Made full-on
- Sitelink test on the brand campaignHypothesis: ROAS from 4.0 to 4.2 in 2 weeks. No lift measured - learning kept for the next roundROASNo lift · Retired
We climb the ladder before we touch what's live.
The best accounts win by testing hypotheses relentlessly. There's a right way to do that - a ladder we always climb, one rung at a time, and we only step onto a live campaign when there's no higher rung available.
Edit a live campaign
Editing a running ad resets what it has learned. We only do this when there's no better rung to climb - a live bid target on an App campaign, or a negative keyword that has to block a term to save spend - and it's always marked in orange with a two-step confirm.
Create a new ad
A fresh ad runs alongside the original, so you compare fairly and keep whichever earns its place.
Run an experiment
A proper Google Ads experiment with a statistical test. The winner takes the budget - and you have proof.
Three ways to work with us.
Pro
$100 / month
14 days free
Cancel any time, pay nothing
- One Google Ads account, every campaign and keyword sized by spend and scored against your own average
- New experiments out of your last 30 days, each naming the campaign and the spend at stake
- One click builds the experiment in Google Ads, and nothing changes without your approval
- Every test tracked to a verdict: lift against control, then promote the winner or retire it
- Use it from Claude, Claude Code, ChatGPT or Codex over MCP, at no extra cost
Agency
$1,000 / month
14 days free
Cancel any time, pay nothing
- Up to 10 Google Ads accounts, one flat price - no cut of their spend
- Everything in Pro, on one board across every client
- Seats for your whole team, no per-user fee
- We set up your client accounts, groups and seats with you
Strategic Help
$2,000 / month
Starts with a call
Billing starts on day one
- Up to 4 calls a month with the founders of GoodLads
- We build your paid ads strategy and execute it with you
- Our support between the calls, straight to the two of us
- Everything in Agency, across all your accounts, no cap
The team you get on a call
No support queue and no account manager. Agency and Enterprise buy time with the two people who build the product.
Questions, answered.
The things advertisers ask before they connect their first account.
What does GoodLads do?
GoodLads connects to your Google Ads accounts, reads the last 30 days of performance for every campaign, and writes three hypotheses per campaign aimed at your ROAS or CPA. Each one names the signal it found in your own data, the change it proposes, and the metric move it predicts. Apply the ones you like in one click - as a Google Ads experiment where the campaign can run one, a new paused ad where it cannot, or a clearly marked live change for the few levers that have no safer form - and follow each one on a kanban board from proposed to a measured verdict.
Will GoodLads change my campaigns automatically?
No, and there is no autopilot to switch on. GoodLads reads your account and proposes; nothing reaches Google Ads until you press apply on a specific hypothesis. What it does ship is built to be undone - a new ad arrives paused, an experiment can be ended, and the one or two levers that touch a serving campaign say so on the button before you click.
What access does GoodLads need to my Google Ads account?
You connect through Google's own consent screen. Google publishes no read-only scope for Google Ads, so that screen asks for the full one - GoodLads holds itself to reading in its own code, and writes only on the click you make on a card. Connect a manager account and you get its whole hierarchy listed, then choose which accounts GoodLads watches. Untrack an account whenever you like, or remove GoodLads from your Google account's permissions and every token dies with it.
Which campaign types does GoodLads cover?
It reads every campaign in the account. Search is where the full set of levers works: a Google Ads experiment copies a Search campaign, and keywords, search terms and responsive search ads are Search constructs. App campaigns cannot run an experiment at all, so they get the one lever that fits - the campaign bidding target, applied live and marked as such. Performance Max gets one too: a Google Ads asset experiment, with the headlines and descriptions the campaign uses today in the control arm and new ones in the challenger arm. Google splits the traffic between them and reports which did better, so it is a graded test rather than a guess - and because it runs inside the asset group that is already serving, the card tells you plainly that half your traffic sees the new copy from the moment you deploy. Fewer, applicable ideas beat a full list you cannot act on.
Does GoodLads ever edit a live ad?
Rarely, and only when there is no better way. Editing a live ad resets what it has learned, so GoodLads defaults to a Google Ads experiment, and to a new paused ad when an experiment is not possible. A few levers - a negative keyword, an App campaign's bid target - have no experiment form, so they apply to the live campaign. Those buttons are orange, ask for a second confirm, and still fire only on your click.
What happens after I apply a hypothesis?
It moves across your kanban board: from proposed to scheduled, live while the experiment runs, and completed once there is a verdict on your ROAS or CPA. Winners can be made full-on in one click; the rest are retired, and the learning feeds the next round of hypotheses.
Who is GoodLads for?
Advertisers and agencies managing Google Ads accounts that spend roughly €10K to €250K a month, who want expert-level hypotheses to improve ROAS or CPA without building or hiring a full in-house team.
How is GoodLads different from Optmyzr, Opteo or Adalysis?
Those tools hand you many levers and rules to operate yourself. GoodLads gives you the three hypotheses that matter this month, grounded in your own results and the business behind the account, plus a safe one-click way to test them and a kanban board that tracks every one to a verdict.
Can I use GoodLads from Claude or ChatGPT?
Yes, on every plan, at no extra cost. GoodLads runs an MCP server, so you can point Claude, Claude Code, ChatGPT or Codex at your account and ask about your campaigns, keywords and search terms directly. Your AI reads your data and can put what it finds on your board as a proposal; it cannot touch your Google Ads account, and deploying anything is still your click in GoodLads. Your model, your tokens, so there is nothing to meter.
How much does GoodLads cost?
Three flat plans, no spend-based math. Pro is $100 a month for one Google Ads account. Agency is $1,000 a month for up to 10 Google Ads accounts, whatever they spend, with seats for your whole team. Enterprise is $2,000 a month, across all your accounts with no cap, and adds the two of us: we build your paid ads strategy, execute it with you, and give you up to 4 calls a month. Pro and Agency start with 14 days free. Prices are in US dollars; if you pay in another currency, Stripe converts at the day's rate when you check out.
Is there a free trial?
Yes. Pro and Agency both start with 14 days free. You enter a card so the subscription converts by itself when the trial ends, and you can cancel from your profile at any point before that and pay nothing. Enterprise is our own time rather than the product alone, so it starts with a call and bills from day one.
How do I change or cancel my plan?
From your profile inside GoodLads. Changing plan, updating the card and cancelling all go through Stripe's billing portal, so a cancellation never waits on us. Yearly billing lives there too: a year on Pro or Agency costs six months, so $600 and $6,000 respectively.
Put good lads on your Google Ads.
Connect an account and read your report in about a minute: what the account is doing, where the money is going to waste, and three experiments worth running. No card, and nothing in the account changes.
Audit my account free