Marketing in a Minute – Episode 21 | Ahmed Moaaz

A customer watched a Reel, clicked an ad, then bought from an email. Which one deserves the credit? Understanding attribution models stops you from shutting down channels that work quietly.

What is the credit-assignment problem?

A customer bought from you today. Before buying, they saw your ad on Instagram; two days later they searched your name on Google and entered from the search result; then an email reached them; and finally they clicked a retargeting ad and completed the order. This episode’s question: which channel deserves credit for this order? That is the subject of attribution.

The problem is that every channel in that story will declare in its report that it owns the sale. Add up the platforms’ reports and you will find the sum of “attributed orders” larger than your actual orders – sometimes by multiples.

The common models for distributing credit

• Last click: all credit goes to the last channel the customer touched before buying. The simplest and most used model, but it is unfair to the channels that first introduced you to the customer.
• First click: all credit goes to whoever started the relationship. Fair to discovery, but it ignores everything that later convinced the customer to buy.
• Linear: credit split equally across all touches. Formally fair, but not all touches truly carry equal influence.
• Time-decay: touches closer to the purchase take a larger share because they usually influence the decision more strongly.
• Position-based: 40% to the first touch and 40% to the last, with 20% divided among those between, honoring both who introduced you and who closed the deal.

There is no absolutely “correct” model; each tells a different story about the same data. What matters is understanding the story of the model you use before building a budget decision on it.

Why do ad platforms exaggerate their numbers?

Not necessarily bad intent – it is the nature of measurement itself:

• View-through attribution: some platforms claim the sale if your ad merely passed before the customer’s eyes without a click, as long as they bought within a certain time window.
• Long attribution windows: a seven-day-click, one-day-view window means any order within it counts for the platform, even if it came for an entirely different reason.
• Multiple devices: they saw the ad on their phone and bought on their computer; the platform can barely follow the full journey.
• Each platform sees only itself: Meta does not know what Google did, so it claims the whole sale – and Google does the same.

How do you measure in a more grounded way?

• UTM tags on every link: every ad and email carries clear source tags, so your store reports tell you where every order actually came from, instead of trusting platform reports.
• Dedicated discount codes: a code per campaign or influencer. Whoever used a channel’s code genuinely interacted with it before buying – strong ground-level evidence of the touch, though it does not by itself prove the channel decisively caused the purchase.
• Ask the customer: a “how did you hear about us?” question after purchase reveals channels no digital tracking shows, like personal recommendations.
• Compare with the big picture: the MER number from a previous episode is your final judge. If the platforms say they produced 90,000 EGP in sales and your store’s total is 100,000, pause: the figure could be legitimate if most of your sales really come from their paid channels, but more often the same orders were counted by more than one platform at once. Compare with your store records before believing.

A practical example

A store spent 10,000 EGP on Meta and 6,000 on Google in a month. Meta’s report says: 120 orders worth 48,000 EGP. Google’s report says: 90 orders worth 36,000 EGP. The total: 210 orders worth 84,000 EGP. The store’s actual records: only 110 orders worth 44,000 EGP from paid channels – the two platforms together claimed nearly double the truth. With UTM tags, it turned out 70 orders had their last click from Meta and 40 from Google. A decision built on platform reports would have led to doubling the budget on phantom results, while the grounded numbers give a picture you can rely on.

What is the practical takeaway for a small or medium business?

You do not need expensive attribution tools. A three-piece system is enough: UTM tags on everything, a “how did you find us?” question after purchase, and a monthly comparison between platform claims and your actual store records. And when in doubt between two channels fighting over credit, try switching one off for two weeks and watch your total sales; the effect of switching off is more truthful than any report.

Common attribution mistakes

• Believing last click alone: it is unfair to discovery channels, so you stop them – then two months later sales dry up because nobody is introducing you to new customers anymore.
• Adding platform numbers as one truth: as we saw, the sum always exceeds your reality because the same sale gets counted twice and three times.
• Changing budgets daily based on platform reports: attribution numbers fluctuate naturally day to day; a budget decision needs at least two to four weeks of data.
• Ignoring non-digital channels: a friend’s recommendation or a family-gathering conversation appears on no dashboard, yet it may be your strongest channel. Asking the customer is your only way to see it.

Start with the simplest possible version this week: create UTM links for your currently active campaigns (many free tools generate them in seconds), and add the “how did you hear about us?” question to the post-purchase page or confirmation message. After one month of clean data, open your store reports and compare them with the platforms’ claims. The gap you see will change how you read every ad report from then on – and you will finally know which pound of your budget works and which only pretends to work while adding nothing real to your actual sales.

Summary

Credit assignment is a matter of measurement and fairness together: every channel will praise itself, and your job is to hear everyone, then judge by your store’s numbers, not theirs. Use UTM tags, discount codes and the customer question; know the limits of every attribution model you read; and always compare with the big picture. Whoever believes every report will pay for the same sale three times without knowing it.

Keep following Marketing in a Minute: one marketing concept per episode. Have a question about this number? Send it on WhatsApp.