Google and Meta both claim the same lead. Who is right?

Google's report shows 48 conversions. Meta's shows 61. The CRM recorded 73 leads in total.

Nobody made a mistake, nobody lied, and adding the first two gives 109 — a number that does not exist.

Why both claim it

A common journey:

1. The person sees a brand video on Instagram on Tuesday
2. Thinks about it for two days
3. Searches the company name on Google on Thursday
4. Clicks the search ad and fills in the form

Meta claims it: there was contact inside the view-through window. Google claims it: it was the last click before conversion.

Both are right within their own rules. The error belongs to neither platform — it belongs to whoever reads both reports as parts of a single total.

The specific trap of brand search

This is the biggest distortion in cross-channel analysis.

When someone searches your company name on Google, the conversion is cheap, immediate and nearly guaranteed. The brand campaign shows a ridiculously good cost per conversion — and looks like the most efficient channel in the account.

But that person already wanted your company. The interest was created somewhere else: an Instagram ad, a referral, a YouTube video. Brand search harvested demand another channel planted.

That leads to two wrong decisions, in opposite directions:

• Cutting social because "search converts better" — and watching brand search dry up a month later
• Switching off the brand campaign because "they would come anyway" — and letting a competitor buy your name

The way out is to treat brand search as a thermometer, not a channel. Search volume for your brand is an indicator of how well the top of the funnel is working.

What to do instead of adding

1. Use the CRM as the source of the total

The lead count is the CRM's. The sales count is the CRM's. The platforms report spend, and spend is never double-claimed.

total sales           CRM
total spend           sum of platforms
real cost per sale    total spend ÷ CRM sales

That number has no conflict because it does not depend on attribution.

2. Store both first touch and last touch

First touch — how the person met the brand — and last touch — what they clicked before converting — tell different stories, and both matter.

In practice: store the first visit's UTM in a first-party cookie and the converting visit's UTM in another field. Two fields in the CRM, and the argument about "who brought them" stops being theoretical.

3. Compare the trend, not the absolute

If Meta consistently claims 1.6x what the CRM records, that factor is stable and useful. When it changes sharply, something changed — tracking, campaign mix, window.

The trend of the claim is more informative than its value.

4. Test by switching off

The only experiment that genuinely answers. Pause a channel for two to four weeks and watch the company total, not that channel's number.

If the total falls in proportion to what the channel claimed, it was bringing new sales. If it barely moves, it was harvesting.

Start with the smallest channel, in a month without an aggressive target, and with the client informed.

What to tell the client

They will ask why the numbers do not match, and the answer has to fit in one paragraph:

The platforms count differently and each claims the contact it had with the person. If someone sees the ad on Instagram and then searches on Google, both record the same sale. That is why we use the CRM as the total — 73 sales this month — and the platforms only to understand what worked inside each one. Adding the two reports would give 109, a number that does not exist.

Giving that explanation once, at the start, avoids months of argument.

The summary

Both are right and adding is wrong. The CRM is the total; the platforms are diagnostics. Store first and last touch. Treat brand search as a top-of-funnel thermometer, not an efficient channel. And when you need a real answer about a channel's contribution, switch it off and watch the total.

For the mechanics of the windows that cause the duplication, see attribution windows. For the budget decision that follows, see splitting budget across channels.