tracking · attribution · server-side
Meta says one number. Your sales say another.
iOS, ad blockers and rejected cookies strip 30 to 50% of your conversions before they reach Meta and Google. So the algorithm optimizes on broken data, cuts the audiences that actually sell, and burns budget on the ones that don't. We fix the data with server-side tracking, so the platforms see the truth and spend on what works.
Free audit in 48 hours. No access to your accounts required.
What is tracking and attribution?
Tracking is how an ad platform learns what happened after the click. It becomes advanced when it stops depending on the browser and starts leaving from the server. It carries the parameters that identify the person, and it does not count the same sale twice. Attribution is how the credit gets assigned across the ads, emails and pages the buyer touched. When either breaks, Meta and Google keep spending, but they optimize toward the wrong people, because the only feedback they get is wrong. Moon rebuilds that feedback loop server-side so it survives iOS, ad blockers and cookie loss.
What you get:
- The conversions your pixel is silently dropping, recovered
- Meta, Google and your store reporting numbers that converge
- Deduplicated events, so one sale is counted once, not twice
- Event quality high enough for the algorithm to optimize on
- Revenue tied back to the campaign that produced it
The problem
Your algorithm makes real-money decisions on data that's wrong.
A pixel loses 30 to 50% of events to iOS, ad blockers and rejected cookies. Meta over-counts with view-through, your store under-counts with last-click, and neither matches your real sales. Worse: when the pixel misses a sale, Meta thinks that audience failed, pulls budget from it, and scales your losers. You're not just blind. You're paying to be blind.
What that actually costs you, every month:
- Winning audiences get shut off, because the sales they produced never arrived in the platform.
- Losing audiences get scaled, because their tracked conversions look cleaner by comparison.
- Your cost per acquisition reads higher than it is, so you pause campaigns that were profitable.
- You cannot tell whether a bad month came from the ads, the page, or the measurement.
- Every optimization decision you make compounds the error, because the feedback never corrects itself.
None of this shows up as an error message. It shows up as a slow, permanent tax on the budget you already committed.
What we fix
Data the platforms can trust.
Server-side, deduped and tied to real revenue. Built by the engineer behind a production tracking platform.
01
Server-side tracking (sGTM + CAPI)
Recover the 30 to 50% of conversions your pixel loses. Data that survives iOS, blockers and cookie loss, with Event Match Quality above 7 so Meta can actually optimize.
02
One number you can trust
Events deduped and tied to real orders, so Meta, Google and your store finally tell the same story.
03
Consent Mode v2, done right
Compliant and actually recovering data, not the majority who set it up wrong and lose it anyway.
04
Attribution across channels
See which channel and campaign really drove the sale, so you scale winners instead of noise.
How it works
Audited, implemented, verified.
step_01
Audit
We find every event that's broken, missing or double-counted, and exactly where the money leaks.
step_02
Implement
Server-side GTM, CAPI, deduplication and consent, set up and connected as one system.
step_03
Verify
We tie every event to real revenue and push Event Match Quality up, so you trust the numbers.
What changes
What is different thirty days after this is fixed.
The algorithm starts learning again
Once conversions arrive complete, Meta and Google can tell a winning audience from a losing one. They stop cutting the segments that were quietly producing sales.
One number, not three
Ads, store and dashboard converge. You stop refereeing which report is right and go back to deciding what to do with this month's money.
Spend follows revenue, not guesses
With events tied to real money, you can scale on contribution instead of on whichever campaign happens to be tracked best.
It keeps working after the next privacy change
Server-side does not depend on the browser cooperating, so the next iOS release or blocker update does not reset your measurement to zero.
Proof
Tracking in production, not a theory.
We built WaTracker, our own product. It closes the WhatsApp blind spot: the conversation where the sale actually happens and from which nothing goes back to the ad. It ties the click to the lead and to the sale, including when someone lands straight in WhatsApp with no page in between. And it confirms the payment against the bank statement, not against a screenshot. That is 7,000+ leads per month and R$200K+ per month in attributed sales. More than R$50M has moved through the systems we built. Same engineering, pointed at your data.
Creators and founders we've worked with
Real people we've built and shipped work for. Not stock logos, not borrowed names.
Clients
Don't take our word for it.
“I thought the problem was my copy or my sales page, but the real problem was that the Pixel wasn't sending the right data. After Anderson set everything up for me, from the sales platform to the payment gateway, it made a huge difference. With the data flowing correctly, I'm scaling my products better. If your Pixel stopped working right, I highly recommend this.”
Jonatã Silveira
“I had no profit margin and needed to scale sales to pay the Facebook bills. Last month I did R$100k spending R$60k. This month I did R$100k spending R$50k. Sales went up and my revenue became stable. That was crucial to my business.”
Henrique Melo
Start here
Free tracking audit, back in 48 hours.
Send your site. We test what actually fires from the outside: which events exist, which are missing, what parameters they carry, and where the numbers break. No access to your GTM, your ad accounts or your analytics. Nothing for you to set up, nothing to approve internally.
What comes back:
- The events firing today, and the ones that should be and are not
- Parameters missing from the events you do have, and what that costs the match quality
- Where deduplication is absent, so sales are counted twice or not at all
- The gaps ranked by how much money each one is likely bleeding
- What to fix first, whether you hire us or hand it to your own team
The audit is free and yours to keep. Implementation is scoped per project, with price and timeline in writing before anything starts.
If the audit does not surface at least three concrete gaps in your tracking, we will tell you that plainly and there will be no follow-up.
Straight answers
Before you send it.
What makes tracking "advanced" rather than just a pixel someone installed?
An installed pixel is the floor: a browser tag firing an event. Advanced tracking is when the event leaves from the server (sGTM). It arrives through the conversions API carrying the parameters that identify the person. It is deduplicated, so one sale is not counted twice. It respects consent. And it agrees with the revenue your store recorded. The difference shows up in Event Match Quality, the event's match score. That number decides whether the algorithm can optimize at all. And your pixel does not move. Moon keeps what is already installed and builds server-side on top. You do not trade a setup that works for one still to be tested.
Do you need access to my GTM or ad accounts for the audit?
No. The audit runs from the outside: we load your site the way a visitor does and read what the tags actually send. Access only becomes relevant if you hire the implementation.
Why don't Meta and my store ever match?
They count differently: Meta uses view-through, your store uses last-click, and both lose data to privacy. We dedupe and feed clean server-side events so they converge and you can trust the number.
Does advanced tracking lower my cost per acquisition?
Being honest: tracking does not lower cost per acquisition. It makes the number you read become the number you actually have. CPA usually drops on screen right after an implementation. Almost always that is because conversions that were already happening got counted again. Not because media got cheaper. The real gain comes next. With the event arriving complete, Meta and Google stop cutting audiences that were selling quietly. And they stop scaling what only looked clean because it was better tracked. How much that moves in your case depends on how much you are losing today. That is exactly what Moon's free audit measures, before any promise. What nobody should promise you is a percentage. Anyone guaranteeing an X% drop in CPA without looking at your setup is guessing.
Does server-side tracking survive iOS and privacy changes?
It survives most of it, and it is worth saying what it does not solve. It solves the event that dies in the browser. Blocked by an ad blocker. Cut by Safari's ITP, which caps how long a cookie lives. Or lost because the script never loaded. Those start leaving from the server and they arrive. It does not solve refused consent. If someone declines in the banner, the data should not be sent, and server-side is not a way around that. Which is why Moon implements Consent Mode v2 alongside the migration, not after. What can be recovered within the rules gets recovered. What cannot is explicit from the start, instead of becoming a surprise at the next audit.
What if my tracking turns out to be fine?
Then we say so and you have saved yourself a project. That outcome is rare, but it happens, and we would rather tell you than sell you something you do not need.
Get your free tracking audit
Send your site URL and what you sell. We reply within 48 hours with the events you have, the ones you are missing, and what each gap is costing. No account access needed, and no call required to start.
Audit my trackingThe founder
Who builds this.
Moon was founded in June 2022 by Anderson Cavalcante, engineer and founder, in São Paulo. He has thirteen years in technology: he came from programming, went deep into marketing, data and tracking, then into AI engineering, and builds production systems that run on real data every day.
For one client, WhatsApp lead capture brings in 7,000+ leads a month. Out of it come R$200K+ a month in sales tied back to the campaign that produced each one. Another arrived in financial distress and now runs a R$2M-a-year operation on the sales tracking we built. At a third, our AI agents qualify and follow up on around 1,000 leads a week.