Stop Trusting Ad Platform Reports: Use 3rd Party Attribution

Google says 400 purchases, Meta says 350, and your Shopify backend says 280 orders. Who’s right? The gap between dashboard Return On Ad Spend (ROAS) and true business performance is one of the most common conversations we have as a marketing agency.

The appeal of the Ad platform reporting is obvious, it’s convenient, free and the (ROAS) stats in Ad platforms can often look great. However, that’s the problem, there’s an underlying structural bias that incentivises advertisers to gauge performance this way.

Let’s be honest about the limitations of Ad platform reporting and how they position your returns on advertising investment, and look at how you can achieve a less biased view through 3rd-party attribution platforms and a well-managed CRM to allocate your budgets for better ROI, ultimately.

Why Ad Platform Results Can’t Be Trusted

There is a structural problem in that every ad platform is both the ad seller and the measurement provider in their own dashboards. And they can’t be trusted to ‘mark their own homework’. Each platform attributes credit generously to itself. There is no neutral referee. This is to be expected from a platform whose core commercial interest is retaining advertisers’ spend there.

To put this in investment terms, you wouldn’t only take investment advice from the company you were considering buying shares in; they might just have a little bias towards their future prospects. You’d want impartial advice. This is why we benchmark platform figures against GA4 and a client’s own CRM or sales data as standard, rather than reporting platform numbers at face value.

While free products like Google Analytics are genuinely helpful and offer excellent tools for measurement and marketing, there is always a cost to free products. Google products will tend towards awarding Google Ads the most credit for results across your digital journeys; the same is true for Meta, TikTok, etc.

Attribution Settings

When you advertise on Ad platforms without a 3rd party view, you’re operating within a ‘walled garden’. Each Ad platform measures conversions using their own pixel/platform based attribution system, not a neutral third party. How you configure your attribution in each platform matters.

If you use TikTok’s ‘Engaged View-Through’ attribution, TikTok will claim credit for conversions within 24h of someone watching 6 seconds of a video.

Meta’s default 7-day click + 1-day view window means a conversion that happened within 24 hours after someone glanced at an ad gets claimed. Until January 2026, this included 28-day view-through windows, which inflated success metrics even more with anyone having viewed an Ad, whether they actually interacted with it or not, over the last 28 days and then converted on your site. Meta states:

46% of purchase conversions on Reels now happen within 2 seconds of video attention

I take that claim with a healthy dose of scepticism; you’re free to make your own mind up.

How Clicks Are Counted

Meta documentation confirmed that likes, shares, and saves were being counted as ‘clicks’ in attribution windows, not just clicks on links to visit a website as many marketers expect a ‘click’ to indicate. We were always aware of this as an agency, so we always only reported on link clicks as ‘clicks’, but many advertisers were surprised to see their click-through rates (CTR) plummet since March 2026.

Modelled Conversions

iOS14 introduced new privacy restrictions on Apple devices which further reduce trackable signals, making platform-reported numbers even less reliable. Cookie consent restrictions mean that Google Ads uses modelled conversions and Meta Ads too. Filling gaps in conversion data with statistical modelling will never be perfect and can overestimate.

GDPR & Ireland

EU consent restrictions, including cookie consent, reduce trackable signals further, making Irish and European advertisers especially exposed to over-reporting as platforms rely more heavily on modelled conversions in the EU.

What Platform Dashboards Are Counting

Taking a focused look at Google Ads, Meta Ads, and TikTok Ad attribution quirks, let’s review what each platform dashboard is actually reporting on.

Google Ads

Google removed their first-click, linear, time-decay, and position-based attribution models in 2023. As of September 2025, Data-Driven Attribution (DDA) is the default model, which gives partial credit to each marketing channel that a user engaged with on their journey to a conversion with your business. DDA is more sophisticated than last-click (the last channel the user engaged with gets the conversion credit)

DDA is undoubtedly a big improvement on what we had before but is still Google-centric in its reporting. Conversion modelling can add 15–20% to reported numbers when Google tries to fill the data gaps it can no longer actually track due to privacy.

Meta Ads

iOS14 reduced reporting windows to 7-day click / 1-day view. January 2026 API changes removed 7-day and 28-day view-through windows entirely. Some advertisers saw 30–40% of conversions disappear overnight, not because performance dropped, but because the measurement window shrank.

Meta’s new Incremental Attribution metric (April 2025) is a step in the right direction as it attempts to measure whether a conversion was actually caused by the ad. But it’s opt-in and not yet widely adopted.

TikTok Ads

TikTok is the youngest platform in this list with less mature measurement infrastructure. Engaged-view attribution and in-app conversion tracking make it especially difficult to isolate true incremental value vs. organic or cross-platform influence.

Google Analytics 4

While not perfect and it’s Data Driven Attribution (DDA) is a big step in the right direction to providing a more honest view, GA4 typically underreports vs. Ad platforms (18–35% for paid campaigns when cookies are blocked), but it provides a more neutral reference point than any single Ad platform dashboard, and it’s free so it’s no wonder many businesses are taking their guidance from here.

The Double (And Triple) Counting Conversions Problem

Looking at the example of a customer clicking a Google Shopping ad on Monday, then sees a Meta retargeting ad on Wednesday, and buys on Thursday by searching for your brand to navigate to your site. Both Ad platforms claim the Sale conversion. Your CRM only records one sale. This can become a triple-counted conversion when three paid media marketing channels are involved in the journey.

The practical consequence is that budget decisions made on platform data alone shift spend toward channels that are best at claiming credit, not necessarily the ones driving the most incremental revenue.

Google Analytics 4 DDA goes some way towards combating this by offering data-driven attribution, which gives partial credit to multiple marketing channels involved in the user journey towards a conversion instead of solely ‘last click’ attribution, which credits only the last channel involved before the conversion occurred, but it’s still biased towards Google Channels.

3rd-Party Attribution Platforms Deliver a Less Biased Picture

Depending on your business objectives, be they lead generation, ecommerce or both, there are 3rd-party attribution platforms that can help you achieve a neutral multi-touch attribution overview without bias towards a particular Ad platform. This is the next level up from relying on in-platform reporting.

As 3rd party platforms sit outside any individual platform’s ecosystem, they allow your business to build a unified view of your customer journey using a combination of first-party data and server-side tracking deployed tracking pixels.

Using server-side tracking matters if you want the best possible data, and Meta Conversions API (CAPI) adoption has “plateaued” at mid-market brands. Advertisers without Conversions API in place can lose a significant share of conversion visibility, which means the algorithm optimises on incomplete data. Closing that gap is one of the first things worth prioritising

In simpler terms, 3rd party platforms will be fairer in where credit is given to Ad platforms for your desired outcome and help you better allocate budget accordingly. Combined with a good quality CRM which is well managed internally within the business, these are essential technical elements for any business that wants to get serious about growing and scaling their business.

The Main 3rd Party Attribution Platforms & Who They’re For

The right tool depends entirely on your revenue, channel mix and whether you’re lead-gen or ecommerce. Most brands don’t need the enterprise end of this.

Northbeam

Strongest for enterprise-level ecommerce brands (think €5M–€500M+ annually in ecommerce revenues). It includes multi-touch attribution, brand-channel credit, and deep modelling capabilities. Correcting attribution discrepancies with Northbeam has been shown to improve ROAS by +34% in documented DTC cases.

Triple Whale

Probably the most popular, as trusted by 60,000+ brands and better for smaller DTC brands (circa €1M–€100M revenue annually). Easier to use, faster dashboards, 7 attribution models. Geo holdout tests (excluding half your audience from seeing ads) suggest Triple Whale synthetic incrementality shows 70–85% of true incremental lift vs. reported numbers.

Rockerbox

Rockerbox is a well-regarded mid-market option for ecommerce businesses, particularly for multi-channel retail brands. Good for blending online and offline data sources (people buying in store).

Polar Analytics

Shopify-native integration with a built-in Snowflake data warehouse. Their hybrid server-side pixel retains journeys that Safari’s 7-day cookie clearing would otherwise lose.

Ruler Analytics

Particularly relevant for lead generation and B2B businesses where conversion events are form fills, calls, or CRM entries rather than purchases.

The Limitations of 3rd Party Attribution

Even best-in-class attribution platforms are smoothed models (averaged), not precise measurements. They still can’t perfectly isolate the causal impact, especially on cross-device journeys without identity resolution (where your assigned user ID from your CRM identifying the user is not available). They’re better than platform dashboards, but they’re still not the full picture. Northbeam has no identity resolution layer; Triple Whale and others use probabilistic matching, which is useful, but does not deliver the confidence other methods can.

That’s where zooming out from ad platforms and even 3rd party attribution platforms can deliver a more accurate picture. This is where Marketing Mix Modelling comes in, which is the next level up in marketing measurement and looks at ‘incrementality’ as a KPI. ‘Incrementality’ is a term you’ll see repeatedly used throughout this piece and this is essentially measuring the causal effect of your advertising (did this Ad cause this Conversion), by segmenting your target audience into groups, those that see your Ads and those that don’t (holdout groups) to understand which conversions would have happened anyway and those that were influenced by your investment in Ads.

In 20+ geo-based holdout tests across Shopify brands in 2025–2026, true incremental lift achieved from Meta landed at just 40–60% of what Meta reported. Even Triple Whale’s synthetic incrementality showed 70–85% of reported revenue, better than Meta, but still not the truth.

Another key limitation for those seeking a true overview of all marketing activities, third-party tools are digital-channel-only. They can’t account for TV, radio, out-of-home, sponsorship, or the effect of macroeconomic conditions on your sales. If you run any offline activity, they’re blind to it.

Third-party attribution is a significant upgrade over trusting platform dashboards, but it is not the full picture. For that, there’s the scientific approach of achieving your overview through Marketing Mix Modelling which will be the topic of another post.

Get Started

1. Audit the gap: compare your platform dashboards against GA4 and your CRM. The discrepancy you find is the starting evidence of over-attribution and the business case for change.
2. Implement server-side tracking: Meta Conversions API and Google Enhanced Conversions are the essential foundation before any third-party tool will give you reliable data. This is infrastructure, not an optional upgrade.
3. Run an incrementality test: Meta’s Conversion Lift study or a geo holdout gives you a real, causal data point on one channel’s true incremental impact. Use it to calibrate how much to trust your current numbers.
4. Evaluate a third-party tool: for eCommerce or performance marketing brands spending meaningfully on paid media across multiple channels, an independent attribution platform typically delivers a return through improved allocation decisions.

Better Data Leads To Better Decisions

Platforms are incentivised to show you good news; it is just how ad-funded businesses work. Independent measurement removes that incentive from the equation. Third-party attribution tools and a well-managed CRM are the right first step beyond platform reporting; they give a far more honest picture of your customer journey.

If you’d like to review the measurement setup behind your campaigns, our performance marketing team at Friday is always happy to have that conversation.