DEV Community

Cover image for How to Unify Google Ads, Meta and CRM Data
TrackStack
TrackStack

Posted on Originally published at trackstack.tech

How to Unify Google Ads, Meta and CRM Data

Google Ads says it drove 220 conversions, Meta claims 180, and your CRM shows 260 new customers in total. Each report is internally consistent, yet together they make budget decisions a guessing game. This guide is for marketers, founders and analysts running paid campaigns with a CRM behind them, and it shows how to connect the three data sources into one trustworthy report. You will learn which identifiers act as join keys, how to capture and store them, how to send CRM outcomes back to both ad platforms, which tooling fits which budget, and how to calculate blended CAC and ROAS without double-counting.

Quick Answer

To unify Google Ads, Meta and CRM data, capture click IDs (GCLID, fbclid) and UTMs on every lead form, store them on the CRM contact, then upload CRM outcomes back to each platform and export daily ad spend into one table. Join everything on the CRM record, and judge performance by CRM revenue.

Why the three reports never agree

Each platform grades its own homework. Google Ads credits a conversion to the last Google click it can see. Meta credits the same person to its own ad, often through a different attribution window. The CRM knows who actually paid but usually has no idea which ad brought them in. Add the fact that a lead may click both ads before converting, and the same sale appears in two dashboards at once.

The gap grew in 2026 as privacy rules and platform changes reduced visibility. Meta, for example, removed some view-through windows from its reporting in January 2026, and browsers keep limiting cookies. The GA4 attribution guide covers how Google's analytics layer handles this, but GA4 still cannot see what happens inside your sales pipeline.

The fix is not another dashboard. It is a shared identifier that travels from the ad click to the CRM record and back, so that every conversion is counted once, against the revenue it really produced.

The join keys that make unification possible

Two systems can only be merged when they share a value. In paid media, the useful values are click identifiers, anonymous browser identifiers, and hashed personal data. Think of them as a layered fallback: use the most precise key available, then the next.

Key Where it comes from Best used for
GCLID Google Ads auto-tagging (URL parameter) Offline conversion import to Google Ads
GBRAID / WBRAID Google Ads on iOS app-to-web and web-to-app journeys Import for privacy-restricted traffic
fbclid / fbc Meta ad click; fbc is the cookie built from fbclid Conversions API matching
fbp Meta browser cookie Fallback matching for Meta
Hashed email / phone Form field, hashed with SHA-256 Enhanced conversions, Meta matching
UTM parameters Your own link tagging Cross-channel grouping in CRM and BI

Two details cause most silent failures. GCLID values are case-sensitive, so lowercasing them in a form tool or spreadsheet breaks every later match. And auto-tagging must be switched on in Google Ads, otherwise no GCLID is appended at all.

Step 1: Capture click IDs and UTMs at the form

Everything downstream depends on this step. When a visitor lands from an ad, the URL carries parameters such as gclid, fbclid and your UTMs. A short script, or your tag manager, should read them on landing, store them in first-party storage, and copy them into hidden fields on every form the visitor submits.

Without that, the lead arrives in the CRM with a name and an email but no origin. A walkthrough on tracking form submissions without code explains the form-event side, and the custom dimensions guide shows how to carry the same parameters into GA4 reports.

A practical checklist for the form layer:

  • Add hidden fields for gclid, gbraid, wbraid, fbclid, fbc, fbp and the five UTM parameters.
  • Persist values for at least the length of the sales cycle, not just the session, so a returning visitor keeps the original source.
  • Keep first-touch and last-touch UTMs in separate fields instead of overwriting one with the other.
  • Test with a real ad click in a private window, then open the CRM record and confirm every value arrived unchanged.

A common failure in setups like this is a landing page that works perfectly, while a form embedded from another domain never receives the parameters. Check that path explicitly.

Step 2: Store everything on the CRM contact

Create dedicated custom properties for each identifier, plus fields for lead source, campaign, first-touch date and the lifecycle stage timestamps. Resist the temptation to reuse a generic notes field, because later automations need clean, filterable properties.

Decide early what a conversion means in your business. A form fill is cheap. A qualified lead, a booked demo or a paid invoice is what the ad platforms should optimize for. Record a timestamp for each stage, since both Google and Meta ask for the time of the event when you upload it. Matching on hashed email or phone as a second key is worth planning now, because a share of leads will arrive without any click ID.

Duplicates deserve attention too. One person often submits several forms, books a call, and later buys under a colleague's address. Choose a normalized email as the primary identity, merge duplicate contacts on a schedule, and keep the original click IDs from the first record when you merge. Otherwise the revenue lands on a contact that carries no source, and the channel that started the journey loses credit.

Step 3: Send CRM outcomes back to the platforms

This is the step that turns reporting into optimization. Smart bidding and Meta's delivery system learn from the conversions you report. If they only see form fills, they will find more people who fill forms, not more people who buy.

For Google Ads, upload offline conversions using the stored GCLID, or GBRAID and WBRAID where relevant. Where click IDs are missing, enhanced conversions for leads let you supply hashed email or phone instead. Google is also steering developers toward its Data Manager API for these uploads, so check which route your CRM connector uses. Conversions must be imported within the platform's allowed window after the click, so schedule the upload daily rather than monthly. The Google Ads conversion tracking setup covers the conversion actions you need before importing.

For Meta, use the Conversions API with CRM events. Send the event name, event time, fbc and fbp where you have them, and hashed email and phone. Include an event ID so Meta can deduplicate against any pixel event for the same action. Meta's Events Manager shows an Event Match Quality score, which is the quickest way to see whether your identifiers are doing their job. Low scores usually mean too few identifiers per event.

After the first upload, verify it. Google Ads reports how many imported rows matched a click, and Meta's Events Manager lists received events with their diagnostics. A match rate far below the share of leads that carried a click ID points to a formatting or timing problem. Fix it before scaling the process, because bidding models learn from whatever you send, including mistakes.

Many CRMs offer native integrations for one or both platforms. Use them if they pass the parameters you need. If not, an automation tool or a small script can read qualified leads daily and post them to each API.

Step 4: Pull ad spend into one place

Click IDs connect leads to ads, but spend and impressions must still come from the platforms. The practical question is how to move those numbers into a single table, and the answer depends on volume and technical skill.

Method Cost model Best for Limits
Manual CSV or Sheets add-on Free to low One or two channels, weekly review Manual effort, error-prone
Connector tools (Supermetrics, Coupler.io, Windsor.ai) Subscription by sources and usage Teams that report in Sheets or Looker Studio Cost grows with sources
Automation platforms (Zapier, Make) Per-task or per-operation Event-level CRM syncing Not built for bulk spend history
Warehouse (BigQuery) Storage and query usage Analysts who want full control Needs SQL; Meta has no native transfer
Marketing analytics platforms Custom or tiered pricing Larger budgets, many channels Higher cost, setup time

For most small and mid-sized teams the sweet spot is a connector feeding a spreadsheet or BigQuery. BigQuery's Data Transfer Service has a Google Ads connector at no charge for the transfer itself, though it runs daily and you still pay for storage and queries. Meta data needs a connector or a custom pipeline. Pricing for these tools changes often, so check current plans before committing.

If you only need a lightweight start, the guide to connecting Google Sheets to your CRM is enough for a first version, and the attribution tools comparison helps when you outgrow it.

Step 5: Build the unified report

Structure the final table with one row per CRM lead or deal, carrying source, campaign, click IDs, stage timestamps and revenue. Aggregate spend daily by platform and campaign. Then join the two on date, source and campaign using the UTMs you standardized in step 1.

Three metrics belong at the top of the report. Blended CAC is total marketing spend divided by all new customers, which no platform can distort. Channel CAC divides each channel's spend by the customers the CRM attributes to it. ROAS compares CRM revenue, not platform-reported value, to spend. Pair them with the model you chose in the attribution models explainer, and state the model on the report so nobody compares unlike numbers.

Add lifetime value once you have enough history. A channel with a higher CAC can still win if its customers renew or buy again, so compare CAC with revenue per customer over six or twelve months. Review the report on a fixed rhythm, weekly for budgets and monthly for strategy, and keep a short changelog of tracking changes so a sudden jump can be traced to its cause.

Here is an illustrative example with assumed numbers: $6,000 on Google Ads, $4,000 on Meta, and $120 average first order.

Channel Platform-reported conversions CRM-matched customers Platform CAC CRM CAC CRM ROAS
Google Ads 220 150 $27.27 $40.00 3.0
Meta 180 70 $22.22 $57.14 2.1
Unmatched / organic - 40 - - -
Total 400 260 - $38.46 blended -

In the platforms, Meta looks cheaper. In the CRM, it is the more expensive channel, and 180 platform-reported conversions never matched a CRM customer, because of overlap between the two platforms or leads that never bought. Without the join, the budget would likely have shifted in the wrong direction.

Common mistakes to avoid

  • Lowercasing or truncating click IDs somewhere between the form and the CRM.
  • Importing conversions weeks late, after the platform's window has closed.
  • Uploading every lead as a conversion, which trains bidding on low-quality signals.
  • Sending hashed data without normalizing it first: trim whitespace and lowercase emails before hashing.
  • Skipping consent. Only send personal data you have a legal basis to share, and respect consent signals under GDPR and similar rules.
  • Treating platform-reported ROAS and CRM ROAS as interchangeable.

Fix these in order. A clean capture layer solves more problems than any reporting tool can.

FAQ

Can I unify the data without a developer?

Often yes. Hidden form fields, a tag manager, native CRM integrations and a connector tool cover most small setups. You need a developer mainly for custom form embeds, server-side events or a warehouse pipeline.

What if some leads have no click ID?

Use hashed email or phone as the fallback key. Both Google's enhanced conversions for leads and Meta's Conversions API can match on them, though match rates are lower than with click IDs.

Do I need BigQuery?

No. A connector feeding Sheets or Looker Studio is enough while volume is modest. BigQuery becomes useful when you have large datasets, need SQL joins, or want to keep long-term history.

How often should CRM conversions be uploaded?

Daily is a sound default. Faster feedback helps bidding algorithms learn, and it keeps you inside each platform's allowed upload window after the click.

Why do platform conversions exceed CRM customers?

Platforms count conversions, not unique customers, and each takes credit under its own rules. One buyer who clicked both ads can appear in both reports, and some conversions never become paying customers.

Is this compliant with privacy rules?

It can be, if you collect consent where required, hash personal data before sending it, and share only what the platforms need. Confirm the details with a qualified legal advisor for your region.

Conclusion

Unification is mostly plumbing: capture identifiers once, store them on the CRM record, send outcomes back, and combine spend in one table. Start with step 1 this week, because every later improvement depends on it.

TL;DR

  • Google Ads, Meta and the CRM disagree because each platform credits conversions by its own rules, while the CRM knows revenue but not the source.
  • The fix is a shared identifier: capture GCLID, fbclid and UTMs on every form, store them on the CRM contact, and use hashed email or phone as the fallback key.
  • Send qualified-lead and paid outcomes back to Google Ads (offline conversions) and Meta (Conversions API), ideally daily, then verify match rates.
  • Pull daily spend into one table with a connector, spreadsheet or BigQuery, join on the CRM record, and judge channels by blended CAC and CRM-based ROAS, not platform-reported numbers.

Originally published on Trackstack Editorial.

Top comments (0)