How ad platforms count and report conversions differently

If you tally paid media, you already cognize the feeling. Google Ads says you drove 400 conversions past month. Meta claims different 250. Microsoft adds 60 more.

Add them up, and you’ve apparently sold to 710 people. But your finance team’s study says only 480 income deed the bank.

So who’s lying?

Nobody.

Most group presume the numbers are wrong.

Ad platforms almost ever study much conversions than your business records. Platforms study higher numbers because they count conversions differently. Once you understand those counting methods, the evident contradiction becomes overmuch easier to construe and use.

Start pinch the incentive

Here’s the uncomfortable truth that explains almost everything else: it’s successful a platform’s commercialized liking to study much conversions.

The much conversions a level shows you, the amended it looks. The amended it looks, the much you judge it’s working. The much you judge it’s working, the much you spend. That’s the business model. Always retrieve what makes the level money.

This is logical economics. Given a prime betwixt counting conservatively and counting generously, each level has a structural logic to count generously, and they each do.

It’s counting, not lying

Here’s the reframe that keeps this from turning into different “the platforms are lying to you” rant.

The number of existent conversions is fixed. There are only truthful galore existent conversions successful a fixed period, nary matter really galore get reported crossed Google, Meta, and Microsoft combined. Three platforms tin each declare the aforesaid sale, and often do, but the customer still bought only once.

Focus little connected reconciling each number and much connected knowing really each level counts conversions. Don’t get excessively caught up chasing a perfect, unified number crossed each platform.

Understand the differences, activity pinch what you’ve got, and judge that “good capable to move the dial” is usually the correct modular for the business.

Dig deeper: Why attribution and effect are nary longer the aforesaid point successful PPC

The structural reasons the numbers don’t statement up

If you want to explicate the gaps to a CFO, aliases to yourself, these are the actual reasons platforms diverge from each different and from your ain systems.

Attribution windows

This is 1 of the biggest. Meta defaults to a seven-day click model (plus a one-day view). Google Ads pinch data-driven attribution looks backmost up to 90 days.

Before you alteration thing else, those 2 platforms are already counting different conversions simply because they’re looking astatine different clip frames.

What counts arsenic an ‘engagement’

Platforms don’t work together connected what earns credit. On Meta, a carousel swipe, a video view, aliases a station stock tin beryllium treated arsenic an engagement and person attribution credit.

On Google Ads and Microsoft Ads, you mostly person to click the ad. The customer travel is the same. The attribution rules aren’t.

View-through conversions, particularly YouTube

This is simply a awesome root of inflation. Display, programmatic, affiliate, and YouTube channels often count conversions from group who saw an advertisement alternatively than clicked it.

YouTube view-throughs, successful particular, tin overinflate results because a position is invisible to your analytics, ecommerce platform, and CRM. Those systems can’t cognize personification saw an ad, only that they clicked aliases arrived.

There’s thing incorrect pinch optimizing toward and reporting connected YouTube conversions. Just don’t people your retargeting homework pinch view-throughs. They should beryllium modeled and, ideally, validated against incrementality.

Dig deeper: Why amended signals thrust paid hunt performance

The in-platform attribution model

Even wrong a azygous attribution window, the measurement in installments gets distributed changes the picture. Google’s default data-driven attribution (DDA) spreads fractional in installments crossed each relationship successful the Google Ads situation complete 90 days, based connected its instrumentality learning model.

Meta typically uses a last-touch, one-touch model. Different distribution logic produces different reported numbers for the nonstop aforesaid underlying journey.

Platform silos versus analytics platforms

Each level tin only spot what happens wrong its ain walls.

• Google Ads tracks Google Ads.
• Meta tracks Meta.
• An analytics platform, CRM, aliases ecommerce strategy sees the full travel crossed email, paid social, organic, affiliate, and nonstop channels, past assigns in installments utilizing its ain logic, often connected a last-touch basis.

That’s why those systems study different numbers from the level dashboards. It’s besides why each level tin laic declare to the aforesaid conversion.

Modeled conversions

Privacy changes collapsed the aged measurement of tracking, truthful each level built systems to capable the gap, each pinch its ain methodology based connected the information it has. Google uses enhanced conversions and Consent Mode.

Meta uses information matching, looking for personally identifiable accusation (PII) indicators to find whether personification who interacted pinch an advertisement later converted. The request is legitimate, but it’s besides a existent root of discrepancy and, successful places, a achromatic box.

Cross-device tracking

Google and Meta some exemplary journeys that span aggregate devices belonging to the aforesaid person. That modeling is different genuine logic the numbers diverge from each different and from your ain systems.

Dig deeper: Your ROAS looks awesome — but is it really driving growth?

---

The costs of misreading level data

Conversion information leads to insight. Insight leads to decisions. If the penetration is built connected misread data, the decisions will beryllium worse each time.

The group who get burned aren’t the ones who understand the differences. They’re the ones who can’t explicate why 1 level says 1 point and different says thing else. That’s wherever you travel up and manus a stakeholder penetration that’s simply wrong.

There’s besides an accounting trap worthy naming. The infinitesimal you dainty level numbers arsenic the golden modular for accounting, the full point falls over.

Conversion actions are tracked utilizing different methodologies. They aren’t counting the money successful your bank. Use them for optimization wrong capacity trading alternatively than for accounting. Those are 2 different jobs.

This is besides wherever CMO and CFO skepticism comes in, and it’s usually misjudged. Leaders are often suspicious of platform-reported data.

Assuming a cleanable setup — 1 information layer, accordant GTM triggers, and a robust search model — there’s nary logic to judge the numbers are wrong. They’re counted otherwise and counted generously. Understanding that favoritism is what matters.

The pragmatic rule to onshore on

Here’s the norm of thumb that makes each of this usable alternatively than paralyzing:

If each the numbers are moving successful the correct guidance — overreported Google Ads, overreported Meta, overreported Microsoft — there’s a very bully chance the business is moving successful the correct direction, too.

You don’t request a perfect, reconciled fig to cognize whether your trading is working. You request the platforms to beryllium broadly trending successful the correct guidance and the business information to corroborate it.

What does bully look like?

There’s nary problem utilizing level metrics to optimize campaigns, provender the algorithms, and study up to your CMO, CEO, aliases CFO.

The non-negotiable is knowing the counting methodologies down each platform. That’s the statement betwixt useful penetration and confidently incorrect insight.

Mature advertisers spell a measurement further. They move beyond earthy level counts done incrementality testing, trading operation modeling, and first-party information that attributes capacity to existent customers and existent purchases.

The azygous astir valuable move is feeding the correct business information backmost into the platforms. Instead of obsessing complete which level claims much conversions, attraction connected life value, CAC, merchandise margin, returns, and lead quality. Those are the business signals that thrust business results.

Comparing which level reports much conversions is often a fruitless exercise. The separator comes from feeding genuine business signals to the algorithms, truthful they make genuine business outcomes.

Dig deeper: Why your B2B PPC metrics whitethorn beryllium lying to you

The 1 point to do tomorrow

Ask your paid media squad a azygous question: Do you understand the different accounting methodologies betwixt the platforms?

If they can’t explicate why Google says 1 point and Meta says another, that’s the spread to adjacent first. Everything other follows from it.

Platform numbers aren’t wrong. They’re counted otherwise and generously. Use them to optimize, not to account.

Feed your existent business information backmost to the algorithms, and fto those signals thrust optimization.