Published: 10th June 2026
If Google is claiming the sale, Meta is claiming the demand, TikTok is claiming the discovery and Amazon is taking the conversion, your reporting is not giving you the full picture. This multi channel attribution guide is built for brands that sell across Amazon and DTC and need one answer to a simple commercial question: which channels are actually driving profitable growth?
Most attribution problems do not start with tracking. They start with channel silos. One team is looking at ROAS in Meta. Another is judging branded search efficiency in Google. Amazon is measured on ACOS. None of those views are wrong, but none of them is complete. If you scale spend using platform-reported numbers alone, you will almost always overvalue bottom-of-funnel demand capture and undervalue the channels creating intent in the first place.
Attribution is not a reporting exercise. It is a budget allocation system. If your model cannot help you decide where the next £10,000 should go, it is too academic to be useful.
For hybrid ecommerce brands, the challenge is sharper because the customer journey does not respect platform boundaries. A shopper might see a TikTok ad, search the brand on Google three days later, click a Meta retargeting ad, then buy on Amazon because they trust Prime delivery. Another might discover the product through Amazon Sponsored Brands, leave to compare on the DTC site, then convert through branded search. Both journeys generate revenue. Both can be misread.
That is why the goal is not perfect visibility. You are not going to get perfect visibility. The goal is directional truth you can act on with confidence.
Last-click attribution rewards the channel closest to conversion. That sounds sensible until you realise high-intent channels often harvest demand created elsewhere.
Google branded search is the clearest example. If demand has already been built through Meta, TikTok, YouTube or Amazon visibility, branded search may look like the hero while doing little more than closing traffic that was already leaning towards purchase. The same goes for Amazon retargeting and branded marketplace terms. Efficient on paper, yes. Fully responsible for the sale, rarely.
This creates a predictable budgeting error. Brands cut prospecting because it looks expensive, then push more into conversion capture because it looks efficient. For a while, blended numbers may hold. Then new customer volume softens, brand search plateaus and revenue growth stalls. The issue was never just creative or bidding. The issue was attribution logic.
There is no single model that works for every business. The right choice depends on your average order value, purchase cycle, repeat rate and how strongly Amazon and DTC influence each other.
First-click gives more credit to discovery. It helps when you need to understand which channels are opening the market, but it can overstate top-of-funnel impact if conversion rates are weak.
Last-click is useful for operational reporting on what closed the sale. It is not strong enough for strategic budget decisions on its own.
Linear attribution spreads credit across touchpoints. That can reduce bias, but it often flattens meaningful differences between a weak impression and a strong converting click.
Time-decay gives more weight to later interactions while still recognising earlier ones. This is often more realistic for brands with a considered purchase journey.
Position-based attribution usually favours the first and last touch with partial credit in the middle. For many ecommerce brands, this is a practical middle ground because it values both demand creation and demand capture.
Data-driven attribution can be powerful when you have enough clean conversion volume and sensible inputs. But it is not magic. If your tracking is fragmented, your channel taxonomy is inconsistent or Amazon data sits in a separate world, the model will still produce distorted answers. Smarter maths does not fix bad structure.
Start with the business question, not the platform. Most brands should separate three jobs in the journey: demand creation, demand capture and demand conversion. Once you do that, channel performance becomes easier to interpret.
Meta, TikTok and YouTube often play a stronger role in demand creation. Google Search often captures intent that already exists. Amazon can do both depending on the ad type, the keyword mix and whether the shopper starts on or returns to the marketplace. Looking at all of them through one identical efficiency lens is a mistake.
Next, define your source of truth. For most businesses, that should not be one ad platform dashboard. It should be a blended performance view combining media spend, site analytics, marketplace sales and overall revenue movement. Platform reporting still matters, but mainly as a diagnostic layer.
Then align your conversion windows to reality. If you sell a low-cost impulse product, a short window may be enough. If you sell premium products with more comparison behaviour, a seven-day click view can miss meaningful influence. Too short and you under-credit upper-funnel work. Too long and you start giving channels credit for noise.
Finally, segment by new versus returning customers wherever possible. Retargeting and branded traffic usually look stronger when returning demand is mixed in. If your goal is profitable scale, new customer acquisition needs its own lens.
A standard ecommerce attribution model already has blind spots. Add Amazon and the gaps widen.
The reason is simple. A significant share of customers will discover your brand off-Amazon and convert on Amazon, or the reverse. If you only measure DTC performance, upper-funnel media may look weaker than it really is because some of the sales are landing in the marketplace. If you only measure Amazon ads, branded demand from Google, Meta or TikTok may disappear from view.
This is where many brands waste budget. They optimise channels against isolated KPIs, then wonder why total growth is underwhelming. Google appears efficient but depends on demand built elsewhere. Meta looks costly but is feeding both DTC and Amazon. Amazon seems to be carrying conversion but may be benefiting from off-platform spend. Without a cross-channel view, every team can claim success while the business underperforms.
A stronger approach is to review blended revenue, brand search trends, Amazon sales movement and media spend together. You are looking for contribution patterns, not just direct claims. If Meta spend rises and branded search lifts, Amazon sales increase and blended CPA remains stable, that tells a more useful story than any single dashboard.
One of the biggest mistakes is judging channels by the same KPI when they do different jobs. Prospecting media should not be held to the same immediate ROAS target as high-intent search or retargeting.
Another is overreacting to platform-reported conversion drops without checking total revenue impact. Privacy changes, tracking loss and attribution window shifts can make a channel look weaker even when the business result is healthy.
A third is ignoring incrementality. Some campaigns report conversions that would have happened anyway. Branded search, remarketing and marketplace brand terms are the usual suspects. They matter, but not always at the spend level brands assign to them.
The final mistake is trying to make attribution too precise. Senior operators do not need a fantasy of certainty. They need enough confidence to move budget into the right channel mix and spot waste quickly.
A useful multi channel attribution guide should lead to better operating decisions. That means knowing which channels create new demand, which capture existing intent and which help convert shoppers who are already close.
For most scaling brands, the right setup combines platform data, analytics data and commercial judgement. You compare reported efficiency with blended outcomes. You watch what happens to branded search when prospecting spend changes. You track whether Amazon growth is being supported by off-platform investment. You stop rewarding channels simply because they are nearest to the till.
This is not about making attribution neat. It is about making budget decisions harder to get wrong.
For brands running Google, Meta, TikTok and Amazon together, the real advantage comes from integration. When one strategy governs discovery, capture and conversion, attribution becomes less about channel arguments and more about revenue control. That is the difference between reporting on activity and managing growth.
If your numbers look good platform by platform but the wider business is still carrying too much wasted spend, your attribution model is probably flattering the wrong channels. Fix that first, and better media decisions tend to follow.