Skip to main content
TrackingConsulting
Now booking · 2–4 day turnaround

PMax not converting · Conversion tracking

Recover 40–80% of your PMax not converting ad revenue.

Done-for-you 1st-party server-side tracking on Meta, Google, TikTok and Microsoft. Built in 2–4 days. You own the entire stack. No monthly SaaS fees.

✓ 30-minute call · No commitment · You leave with a written remediation plan even if we’re not the right fit.

Built in 2–4 daysOne-time fee95% accuracy guaranteed400-day cookie lifetime1st-party server-side

Built on — and certified for — the platforms that matter

Stape

Pro Partner

Meta

Tech Provider

Google

Tag Manager

TikTok

Events API

Microsoft

UET Partner

LinkedIn

CAPI Partner

Pinterest

CAPI Partner

Built on the same partner stack used by enterprise DTC brands.

Ariful Islam — Founder, Tracking Consulting

Meet your consultant

Ariful Islam

Founder · 8+ years in paid media & tracking

Verified earnings

$200K+

earned on Upwork

Top Rated Plus
4.9/5 · 300+

Verify the receipts

You work directly with Arif — no SDR, no account manager.

By the numbers

No fake testimonials. Just the math.

70+

Platforms supported

Shopify · Woo · Magento · BigCommerce · 65 more

8+ / 10

Meta EMQ guaranteed

in writing — keep working free until hit

2–4d

Build to live

accurate data flowing within 7 days

$0

Monthly fees, ever

one-time fixed price · you own the stack

400d

Cookie lifetime

max RFC 6265bis allows · 57× Safari ITP

9

Ad channels covered

Meta · Google · TikTok · MS · LI · Pin · Snap · Reddit · X

95%+

Tracking accuracy

verified vs Stripe · GA4 · ad-platform reports

30 min

Free founder call

no SDR · no pitch deck · written remediation plan

Numbers from realistic engagements. Your specific recovery depends on your iOS share, ad-blocker rate, and current setup — we’ll quantify it on the call.

Real clients · real videos · real outcomes

Don’t take our word for it. Watch the receipts.

Founders, marketers, and agency owners describing the before-and-after of their PMax not converting tracking rebuild — in their own words, on camera.

The problem

Why PMax has nothing to optimize toward

PMax not converting-specific issues
Leak #01

Nothing in the account is set as a Primary goal

Only Primary conversion goals feed Smart Bidding. If Purchase or Qualified Lead sits under Secondary — or the campaign inherited account-default goals nobody reviewed — PMax reports on one thing and bids on another.

Leak #02

Every conversion is worth the same to the algorithm

When the tag sends value 0, or a hardcoded $50 on every order, Max Conversion Value has nothing to rank clicks by. PMax spreads budget evenly across cheap and expensive customers because it can't tell them apart.

Leak #03

PMax claims conversions your order table doesn't contain

View-through credit on Display and YouTube inventory, brand search absorbed into the campaign, and a conversion action set to count "Every" instead of "One" all inflate the number. The campaign banks all of it as its own.

Leak #04

Merchant Center and your purchase event use different IDs

If item_id in the purchase event doesn't match the feed's id or offer_id, including at variant level, conversions never attach to products. Listing groups and asset groups get no product-level feedback at all.

Who this is for

Honest about who we’re a fit for.

We don’t take every project. Here’s how to tell if we’re right for you.

Good fit

You’re a great fit if…

  • You run PMax not converting (or are migrating to it)
  • You spend $5K+/month on paid ads (Meta, Google, TikTok, Microsoft, LinkedIn)
  • Your reported ROAS doesn’t match what your bank account is saying
  • You want to own your tracking stack — no monthly SaaS fees forever
  • You have 2–4 days to do this once and never think about it again
Not a fit

Skip us if…

  • You spend less than $500/month on paid ads (browser pixel is fine)
  • You want a SaaS dashboard with monthly fees forever (Triple Whale, Hyros)
  • You need attribution modeling — we fix the data foundation, not multi-touch reports
  • You want a 6-month enterprise consulting engagement (we ship in 4 days)
  • You expect us to manage your ad accounts (we don’t — that’s a different service)

If you’re still unsure, book the call anyway — we’ll tell you straight whether we can help.

Our position

If your reported ROAS disagrees with your bank account, you don’t have an attribution problem. You have a data foundation problem. No SaaS dashboard fixes that — only a rebuilt 1st-party server-side stack does.

Ariful Islam — Founder
The outcome

What changes once PMax is bidding on something real

One Primary goal that means what you think it means

Server-side

The campaign optimizes toward the conversion that represents money — not a newsletter signup, not a scroll-depth event someone imported from GA4 three years ago and left as Primary.

Value-based bidding with values that are actually true

8+ EMQ

Real transaction or contract amounts, in the right currency, on every conversion. tROAS and Max Conversion Value stop being aspirational settings and start ranking clicks by what they're worth.

PMax conversions you can find in your order table

Deduped

We match campaign-reported conversions to orders by transaction ID and date, then name each gap: view-through, brand overlap, duplicate counting, or genuine data loss. Four different problems, four different fixes.

Audience signal built from customers, not from visitors

Backend-verified

A Customer Match list from your actual purchase history — refreshed on a schedule, not uploaded once — attached as asset-group signal, which is the input PMax has the least of by default.

The full transformation report

Before vs after — built in 2–4 days, accurate data within 7 days

We deliver the 1st-party server-side stack in 2–4 business days. Each metric below measured 7 days pre-launch vs 7 days post-launch using your own ad-platform reports + Stripe + GA4. No 30-day waiting game.

Attribution accuracy

  • Reported ROAS vs Stripe match
    Before60% (35% drift)
    After95%+ match
  • iOS conversion match rate
    Before~58%
    After~92%
  • Cross-domain click-ID retention
    Before~40%
    After~95%
  • Server-attributed conversions
    Before0%
    After100%

Audience quality

  • Meta Event Match Quality
    Before4.5 / 10
    After8+ / 10
  • Cookie lifetime (Safari)
    Before7 days (ITP cap)
    After400 days (max allowed)
  • Lookalike seed-list size
    BeforeShrinking as signal decays
    AfterRebuilt from server-side matches
  • Identity match params sent
    Before1–2 (email maybe)
    After9 (email · phone · IP · UA · click_id · etc.)
  • Retargeting list health
    BeforeDecaying
    AfterCompounding

Bidding performance

  • Smart Bidding stability
    BeforeRetraining weekly
    AfterLocked-in
  • CPA trend
    BeforeDrifts up as signal degrades
    AfterBidding learns from complete data
  • ROAS variance week-over-week
    Before±35%
    After±5%
  • Conversion volume Meta sees
    Before65%
    After95%+

Compliance & ownership

  • GDPR / CCPA audit-ready
    BeforeWeak
    AfterAudit-clean
  • Consent Mode v2 wiring
    BeforeNot configured
    AfterFully wired + verified
  • 3rd-party cookie deprecation ready
    BeforeNo
    AfterYes
  • Data ownership
    BeforeVendor-locked
    After100% yours

Server-side

money events from your backend

8+ / 10

Meta EMQ guaranteed

95%+

tracking accuracy

A typical engagement

The shape of a typical PMax not converting rebuild

We’re not going to invent a fake testimonial here. Instead — here’s the realistic pattern of what happens, anonymized but accurate to engagements we ship.

Day 0 — They arrive

A 7-figure PMax not converting store spending ~$25K/month on Meta. Reported ROAS shows 4.1× in Ads Manager. Their bookkeeper’s number says 2.6×. The founder doesn’t know which is real, so they keep doubling down on what looks like a winning campaign — except cash flow tells a different story.

Day 1 — Audit

We screen-share. Within 20 minutes we’ve found the leaks: native pixel firing twice on iOS, CAPI dedup using the wrong event_id (Meta is double-counting), GA4 showing 38% (not set), and the Customer Events API never enabled. The 1.5× ROAS gap explains itself.

Day 2–3 — Build

Server-side GTM container goes live on data.theirstore.com. Meta CAPI rebuilt with full advanced matching. Google Enhanced Conversions wired. TikTok Events API turned on. All event_ids deterministic. iOS recovered.

Day 4 — Launch

Live-fire test purchases through every funnel step. Meta Event Match Quality goes from 4.5 to 8.7 — verified live in Events Manager. They get a Notion runbook + Loom walkthrough so their dev team can extend it.

Day 5–7 — Accuracy

Conversion data starts flowing cleanly. Day 7: reported ROAS reads 2.8× (matches Stripe within 5%). Lookalike audiences grow because Meta finally has clean seed data. Their CAC dashboard makes sense for the first time in a year.

Day 30 — The math compounds

Smart Bidding has 30 days of clean signal. CPA drops 18%. They scale Meta budget 40% because they finally trust the data. Net new revenue from the recovered tracking: roughly $14K/mo above what they were getting before.

Numbers above are realistic averages from engagements we’ve shipped. Your specific recovery depends on your iOS share, ad-blocker rate, and current setup — we’ll quantify it on the call.

The cookie window unlock

400-day cookie lifetime — the longest your browser allows

Default PMax not converting setups put cookies on a 7–30 day timer. Safari ITP caps client-set cookies at 7 days. That kills your retargeting windows, shrinks lookalike audiences, and starves Smart Bidding of historical signal.

We set 1st-party cookies via HTTP headers on your own subdomain — exempt from ITP’s 7-day cap. Lifetime: 400 days. The maximum Chrome, Firefox and Safari currently permit.

  • 57× longer retargeting window vs Safari default

    Reach users who saw your ad 6 months ago instead of last week.

  • Bigger lookalike seed pools

    More retained users in your custom audience = better lookalike quality on Meta, Google, TikTok.

  • Smart Bidding has 13× more signal history

    Algorithms optimize on 400 days of behavior, not 7 days. CPA stabilizes faster.

  • Critical for high-AOV / long-consideration purchases

    If your average buyer takes 30+ days to convert, default cookies miss them. 400-day cookies don’t.

Cookie lifetime — head to head

How long cookies survive in real-world traffic

Safari ITP (default)7 days
Chrome / Firefox (typical)30 days
Meta default retargeting window90 days
Google Ads default cookie540 days(but capped by browser)
Our 1st-party server-side400 days

400 days is the hard maximum enforced by RFC 6265bis (Chrome 104+, Safari 17+, Firefox 110+). We set it via HTTP Set-Cookie: Max-Age=34560000 on your subdomain. Survives Safari ITP’s 7-day client-side cap.

Under the hood

What’s actually happening on your server-side stack

Six capabilities your default PMax not converting setup doesn’t have. Built on the same Stape Pro stack used by enterprise DTC brands.

Cookie window

Cookie Keeper

400-day 1st-party cookies

Set via HTTP headers on your subdomain — the maximum browsers allow. Survives Safari ITP’s 7-day cap that kills retargeting.

Ad-block resistance

Custom Loader

Custom-loaded GTM scripts

Your gtm.js and analytics scripts load through a custom path on YOUR domain. uBlock, AdBlock Plus, Brave can’t identify them as trackers.

Attribution

Click ID Restorer

Click ID restoration

fbclid / gclid / ttclid / wbraid recovered from the URL and persisted server-side. Multi-redirect funnels stop dropping attribution. Safari user IDs restored.

Profit signal

POAS Data Feed

POAS — Profit on Ad Spend

Send actual profit data (not revenue) to Meta / Google. Smart Bidding optimizes for the metric that matters: net margin, not gross revenue.

Identity

User ID

Cross-session User ID

Stable userID generated server-side from IP + UA + SSL fingerprint. Stitches anonymous → known → returning visitors across 400 days.

Audience signal

GEO Headers

GeoIP enrichment

Country, region, city, postal code passed to Meta / Google CAPI from your server — without exposing the raw user IP to the browser.

Compliance

Anonymizer

PII hashing by default

Email, phone, name, address are SHA-256 hashed on your server before they leave for Meta or Google. Zero raw PII in transit.

Quality filtering

Bot Detection

Bot traffic detection

Bot/scraper traffic identified at the gateway and blocked from polluting your audiences. Smart Bidding optimizes against humans, not bots.

Reliability

Dedup logic

Browser ↔ server dedup

Deterministic event_id and event_time so Meta / Google never double-count. Your reported ROAS stops contradicting Stripe.

Reputation

Dedicated IP

Dedicated outbound IP

Your sGTM gets a dedicated static IP for outgoing CAPI requests. Better deliverability to Meta, Google, TikTok — no shared-IP noise.

Residency

Multi-region

EU / US region routing

GDPR-conscious stores route their CAPI traffic through EU regions (Stape EU or self-hosted GCR europe-west). No US data hop.

Data warehouse

BigQuery export

BigQuery + Looker dashboard

Server logs + GA4 raw events streamed to BigQuery. Looker Studio dashboard branded for your team. Premium tier.

We’re official Stape Pro partners — every Power-Up above is enabled and configured for you. Or we self-host the equivalent on your Cloud Run / GKE if your traffic justifies the cost crossover. You pick.

Live data flow

What your stack is doing — right now

Every conversion fires through your subdomain, gets enriched with hashed identity params, deduped against the browser pixel, then forwarded to Meta CAPI, Google EC, TikTok Events API and Microsoft UET in parallel.

You see this stream live in your sGTM debug console — not a SaaS dashboard you rent. It’s your data, on your server, on your subdomain.

  • p99 latency< 80ms
  • Avg. event payload9 identity params
  • Browser → Server dedupDeterministic event_id
  • Failure handlingRetry · queue · dead-letter
sgtm-debug · data.fix.com
streaming
10:42:18[INFO ]evt=PageView | id=ev_a3f2.. | ua=ios17 | ip_hash=ok
10:42:18[OK ]→ meta.capi status=200 emq_pred=8.6
10:42:18[OK ]→ google.ec status=200 match=enhanced
10:42:24[INFO ]evt=AddToCart | id=ev_b1c8.. | val=$84.00
10:42:24[DEDUP]browser↔server merged on event_id=b1c8..
10:42:24[OK ]→ meta.capi status=200 emq_pred=8.7
10:42:24[OK ]→ tiktok.eventsapi status=200
10:42:31[INFO ]evt=Purchase | id=ev_c9d4.. | val=$184.00
10:42:31[POAS ]profit=$58.42 sent (margin 31.7%)
10:42:31[OK ]→ meta.capi status=200 emq=8.8 ✓
10:42:31[OK ]→ google.ec status=200 ✓
10:42:31[OK ]→ microsoft.uet status=200 ✓
10:42:33[BOT ]blocked: ua=AhrefsBot · 1 evt suppressed
uptime: 99.98%all channels healthy
Every paid-media channel · 1st-party server-side

CAPI / Events API ready for every ad platform you run

One server-side gateway. Every major ad platform fed real conversion data, deduped, with full identity match.

Meta

Conversions API

Server-side

Google

Enhanced Conv.

Server-side

TikTok

Events API

Server-side

Microsoft

UET Server

Server-side

LinkedIn

CAPI

Server-side

Pinterest

CAPI

Server-side

Snapchat

CAPI

Server-side

Reddit

CAPI

Server-side

X / Twitter

Events API

Server-side
What you get

What we rebuild before PMax gets another dollar

Item 01

A conversion goal audit, action by action

Every conversion action listed with its category, counting setting, attribution window, Primary/Secondary status, and whether the campaign uses account-default or campaign-specific goals.

Item 02

One money event, fired from your backend

purchase or qualified_lead sent server-side with transaction_id, value and currency, so PMax counts the thing that actually happened rather than a thank-you page someone can refresh.

Item 03

Enhanced Conversions on every PMax conversion action

PMax has no keyword-level data to work with, so hashed customer data carries more of the matching load here than in Search. Wired for both web conversions and Leads.

Item 04

item_id aligned to the Merchant Center feed

Purchase and view_item item IDs mapped to the feed's id and offer_id at variant level, then confirmed in the diagnostics panel rather than assumed.

Item 05

A Customer Match list on a refresh schedule

High-value customer segments pulled from your CRM or order history, hashed, uploaded, and attached as audience signal — with a job that keeps them current instead of a one-time CSV.

Item 06

A backend reconciliation report

PMax-reported conversions against your own orders for the same window, side by side, with each discrepancy explained. This is the deliverable clients keep coming back to.

The honest math

Three ways to fix your tracking. One of them makes sense.

Option 1

Hire a senior tracking engineer

$15K–$25Kin dev cost

or 2–4 weeks of senior FTE time

  • 2–4 weeks before anything ships
  • Misses iOS / Safari / dedup edge cases
  • No EMQ guarantee — it’s “done” when they say so
  • Pixel debugging is not their day job

Option 2

Triple Whale / Hyros / Northbeam

$300–$2,000/month forever

$18K–$120K over 5 years

  • You rent a dashboard — they own your data
  • Cancel and you lose attribution overnight
  • Per-event / per-visitor pricing surprises
  • Still doesn’t fix your underlying tracking
Recommended

Option 3 — us

Done-for-you 1st-party server-side

$400–$1,800one-time

pay once · own the stack forever

  • Live in 2–4 days · accurate data within 7
  • iOS / Safari / dedup all handled by default
  • Meta EMQ 8+ guaranteed in writing
  • You own everything — server, container, runbook

We’re upfront: if you have a senior in-house dev with capacity and 4 weeks, hiring is fine. If you spend $1M+/yr on ads and want a SaaS dashboard layered on top, Triple Whale works. For most paid-ad advertisers between, we’re the math that makes sense.

From the founder’s desk
Ariful Islam — Founder
Available now

Ariful Islam

Founder & lead engineer

  • Experience8+ years in tracking
  • Engagements300+ DTC brands shipped
  • Specialty1st-party sGTM · CAPI
  • PartnerStape Pro · Meta Tech

“I’ll personally walk through your tracking on the call. No SDR. No sales pitch.”

I’ve spent 8 years rebuilding tracking for DTC brands, SaaS founders, agencies, and one thing I’ve learned: most tracking problems aren’t technical — they’re trust problems. You don’t know who to believe. Your current vendor swears it’s fine. Meta’s reports look great. Stripe says otherwise.

On our call, I’ll personally show you exactly what’s leaking, exactly what it would cost to recover, and whether you should hire us at all. If your current setup is genuinely fine, I’ll tell you. If a $400 browser-side fix is enough, I’ll say so. No upsell pressure.

Book the call. Worst case, you walk away with a free 47-point audit checklist. Best case, we deliver your 1st-party server-side stack in 2–4 days and your data is fully accurate within 7 days.

Arif— Founder

How it works

Built in 2–4 days. Accurate data within 7.

We deliver the entire 1st-party server-side stack in 2–4 business days. You start collecting accurate, deduped conversion data the moment it goes live — and within 7 days every signal is flowing cleanly to Meta, Google, TikTok and the rest.

1Day 1

Kickoff + audit

30-min call. We map your current pixels, GTM, dataLayer, consent setup, and ad-platform integrations. You get a written report of every leak.

2Day 2–3

Build the stack

Server-side GTM container deployed on your subdomain. Meta CAPI · Google EC · TikTok Events API · Microsoft UET wired. Identity match + dedup configured.

3Day 4

QA + go-live

Live-fire test purchases across every funnel step. Meta EMQ verified at 8+. Tag Assistant + Pixel Helper passes. Side-by-side report delivered.

4Day 5–7

Verify accuracy

Your data flows clean. Within 7 days every signal — Meta, Google, TikTok, Microsoft — is fully accurate. Notion runbook + Loom walkthrough handoff.

Diagnose it honestly

Week one silence is normal. Conversions your orders can't find aren't.

PMax has two failure modes that look identical on the dashboard and are nothing alike. One is a campaign that hasn't been given enough conversions to learn from yet. The other is a campaign confidently bidding on a signal that isn't true.

Normal — not a fault

  • A brand-new PMax campaign reporting nothing in week one

    Google's own guidance sits around 30 conversions in 30 days per campaign before Smart Bidding has much to work with. Early silence is the ramp. Rebuilding tags mid-ramp just adds a variable you can't isolate later.

  • PMax conversions will not match GA4 conversions

    Google Ads credits the click's campaign inside its own window and can include view-through; GA4 applies its own model to sessions. These two numbers never agree on any account, and a rebuild won't make them.

  • Spend lands on Display and YouTube you didn't pick

    PMax allocates across channels itself. Cheap impressions with few clicks showing up in the asset report is the channel mix doing its job — unless the same campaign is also absorbing your brand search.

  • Conversions are dated to the click, not the sale

    Google reports a conversion back on the day of the click, so yesterday's numbers keep filling in for days afterward. A report that backfills is correct. A report that never backfills is worth a look.

Actually broken

  • PMax reports purchases your order table has never seen

    After you allow for view-through and cross-device, if reported conversions still can't be matched to any order by ID and date, the event is firing on something that isn't a sale — a refreshable thank-you page, or test orders.

  • Conversion count exactly doubles on a subset of days

    Two tags counting the same event, usually a hardcoded gtag left in the theme alongside the GTM tag, both set to count "Every". One order, two conversions, and Smart Bidding believes both of them.

  • Every conversion carries the same value, or zero

    Value-based bidding with a constant value is bidding with no value at all. Max Conversion Value can't rank a $40 customer against a $4,000 one, so it spreads budget evenly and a tROAS target becomes decoration.

  • Product reporting is empty while purchases record

    Purchases are landing but no listing group or individual product shows any conversions. That's item_id in the purchase event failing to match the Merchant Center feed's id, usually at variant level.

Diagnostic sequence

Four console checks that separate learning from a broken signal

All of this is readable inside Google Ads and one spreadsheet, and it tells you whether to wait or to rebuild.

  1. 1

    Open Goals and read what is actually Primary

    Google Ads, then Goals, then Summary. Write down every Primary conversion action and whether the PMax campaign uses account-default goals or its own. Anything Primary that isn't money is noise you're paying to optimize toward.

  2. 2

    Check the counting setting on your money action

    In the conversion action's settings, "Every" is right for ecommerce and wrong for lead gen. A lead action set to Every counts one person submitting the form three times as three conversions, and PMax bids accordingly.

  3. 3

    Match seven days of PMax conversions to your orders

    Download the conversion segment by day, pull the same dates from Shopify, Stripe or your order table, and compare counts first and transaction IDs second. Write the gap down before you theorize about it.

  4. 4

    Split that gap by cause before blaming any tag

    Segment the Ads report by conversion action, device and network, then read the view-through and cross-device columns. Model differences explain part of it. Whatever is left over is the technical part, and only that part needs fixing.

  5. 5

    Check brand terms in the PMax insights report

    PMax, then Insights, then search categories. If your own brand name is driving a large share of the conversions, the campaign is banking demand you already had rather than creating any.

  6. 6

    Confirm one item_id resolves in Merchant Center

    Take a single purchase event out of GTM Preview, copy its item_id exactly, and search Merchant Center for that string. Variant-level mismatches are the common failure and they never announce themselves.

Source-of-truth validation

How we prove PMax is counting sales you actually made

The number that matters isn't in Google Ads. It's in your order table, and every claim we make gets checked against it row by row.

Source of truthShopify or Stripe order records · your CRM closed-won
  • Place a real test purchase and confirm exactly one conversion arrives with the correct value, currency and transaction_id.
  • Match a full week of PMax-reported conversions to order IDs and record the match rate before and after the rebuild.
  • Label every unmatched conversion as view-through, cross-device, brand overlap, duplicate counting, or genuine data loss.
  • Compare total reported conversion value to your own revenue for the same window, on a cash basis, and explain the delta.
  • Check Merchant Center diagnostics for unmatched item IDs once the feed mapping is live, at variant level not parent level.
  • Re-run the same reconciliation 30 days later so a silent regression surfaces in weeks instead of at the quarterly review.

Straight talk

When PMax doesn't need a rebuild

Half the accounts that ask us about dead PMax campaigns have a volume problem or a patience problem, not a tracking problem.

  • The campaign launched under two weeks ago with almost no conversions yet. That's learning. Changing the tags now makes it un-diagnosable.
  • A week of PMax conversions already matches your orders row for row. The measurement is fine; the problem is the offer, the feed or the creative.
  • You're under roughly 30 conversions a month. Consolidating campaigns will move more than re-architecting the data pipeline will.
  • You run lead gen with no qualification stage recorded anywhere. Build that first — there is nothing meaningful to send back to Google yet.

Got questions?

The honest answers.

No buzzwords, no upsell. If we don’t know, we say so.

01Can I fix this myself?

Part of it, today, for free. Go set exactly one Primary goal, check whether your conversion action counts Every or One, and turn on Enhanced Conversions. That's console work and it fixes a real share of dead PMax accounts. What's harder is the part that decides whether PMax ever works: a server-side money event carrying true values, item IDs that match the feed, and a reconciliation you actually trust. If you have a developer who can fire an event from the order backend, you can get most of the way alone.

02PMax shows conversions but my backend doesn't. Which one is lying?

Usually neither — they're counting different things. Google credits view-through impressions on Display and YouTube placements, counts cross-device journeys your database can't see, and may be absorbing brand searches that would have converted anyway. Your order table counts orders. The useful move isn't picking a side, it's matching them row by row so you know how much of the difference is model versus how much is broken tags.

03Should I just pause PMax until the tracking is fixed?

Defensible, with a caveat. Spending into a broken signal teaches the algorithm the wrong lesson every day, so pausing isn't crazy. But restarting puts the campaign back into learning, so don't pause on Monday and relaunch on Wednesday. If you're going to pause, pause until the rebuild is verified, then relaunch once with the signal correct.

04How many conversions does PMax actually need?

Google's own guidance sits around 30 conversions in 30 days per campaign for Smart Bidding to have something to learn from. If you're well under that, consolidating campaigns and bidding on an earlier pipeline stage will do more than any tagging work. I'd rather tell you that than sell you a rebuild that can't help.

05Does this apply to lead-gen PMax, not just ecommerce?

Yes, and the failure is worse there. Lead-gen PMax typically bids on form fills, which is the one signal that's easy to fake and easy to flood. The money event for lead gen is the qualified stage, returned to Google by offline upload. Without that loop, PMax optimizes toward whoever fills forms fastest, which is rarely who buys.

06You also run paid media. Would you take over the campaigns?

We do, but only on data we've rebuilt and verified — never on inherited numbers. Plenty of clients have us fix the measurement and keep their existing team or agency running the account, and that's a perfectly good outcome. The rebuild stands on its own.

07When do we NOT need this?

If PMax is hitting a ROAS or CPA you're happy with, and spot-checking a week of reported conversions against your orders lines up cleanly, leave it alone. Also if total account spend is small — at low volume, campaign consolidation and offer work beat re-architecting the data pipeline, and I'd tell you so on the call.

Still have questions? Book a 30-min call — bring them all.

The call

What happens on your free 30-min call

No pitch deck, no SDR, no upsell pressure. Founder-led. Bring your ad-platform screens and your PMax not converting setup — we’ll work through it together.

Min 1–5

We get the lay of your store

What you sell, what you spend on ads, who handles dev. Quick context-gathering — no questionnaire upfront.

Min 5–15

We audit your tracking live

Screen-share. We open your Meta Events Manager, GTM, and ad reports. You see exactly which conversions are leaking.

Min 15–25

We map a fix

What needs rebuilding, what stays. The exact PMax not converting setup we'd ship and what it would recover for you.

Min 25–30

Decide. Or don’t.

If we're a fit, we send a 1-page proposal in writing. If not, you walk away with a free remediation plan you can hand to anyone.

No pitch deckFounder-led, not SDRYou leave with a written planZero commitment
Free · 30 minutes · No commitment

Let's find out how many of those PMax conversions are real

Send a week of PMax conversion data and the matching rows from your order table. I'll tell you which gaps are Google's model, which are double counting, and which are actually broken.

Free · 30-min consultation · No commitment · Replies in under 2 hours · Available worldwide via WhatsApp

95% guarantee · fixed-fee