Marketing attribution explained for small businesses

What attribution actually means

Picture a customer named Dana. She sees your Instagram ad on Monday, searches your name on Google on Wednesday and clicks a search ad, opens your newsletter on Friday, then types your web address and buys. Four touches, one sale. Which one earned it?

Marketing attribution is the method you use to answer that. It is a set of rules, called an attribution model, that hands out credit for each sale across the touches that came before it. Those credits roll up into the reports you use to judge each channel, so the model you choose decides which channels look like winners. That matters because most owners move money toward whatever looks like a winner. If your model ignores the ad that first introduced Dana, you may cut the very campaign that keeps new customers coming in.

The common attribution models

You will see these names in ad platforms and analytics tools:

• Last click: the final touch before the purchase gets all the credit. In Dana's case, the newsletter or the direct visit wins and Instagram gets nothing.
• First click: the first touch gets all the credit. Instagram wins. Good for spotting what introduces new people, bad at everything else.
• Linear: every touch gets an equal share. Each of Dana's four touches gets 25%.
• Position-based: the first and last touches get most of the credit, commonly 40% each, and the middle touches share the rest.
• Data-driven: software compares paths that ended in a sale with paths that didn't and assigns credit based on which touches seem to make a difference. It needs a fair amount of data to work well.

What Google and Meta use by default

The choices have narrowed. In April 2023 Google announced it was retiring the first click, linear, time decay and position-based models in Google Ads and Google Analytics 4, noting that less than 3% of Google Ads web conversions were attributed using them. Data-driven, last click and external attribution stayed. Google's help page on attribution models now says data-driven attribution is the default for most conversion actions.

Meta uses its own approach. Its reporting documentation lists the default window as 7 days after a click and 1 day after a view. Each platform also only sees its own ads, which is why their totals rarely agree. Our article on why ad platforms claim more sales than you made explains that gap.

Lookback windows matter as much as the model

A lookback window, also called an attribution window, is how far back in time a tool searches for touches before a sale. A touch outside the window gets no credit at all. Google Analytics 4, according to Google's documentation, uses a default lookback of 90 days for most key events and 30 days for first visits and first app opens.

Change the window and the credit moves, even with the same model. We go deeper in attribution windows explained.

How a small business should use attribution

You don't need a data team to use attribution well. You need consistency and a reality check.

Matrix ROI, a reporting app from the team at Smith App Studio, lines up each platform's claimed conversions beside the payments your own records verify, so the attribution story always sits next to the money. You can start a 14-day trial with no card.

• Pick one model for decisions and stick with it for at least a few months, so trends mean something.
• Tag every link you control with campaign parameters so your analytics can tell sources apart. See the UTM parameters guide.
• Compare attributed revenue with what actually landed in your bank or store. If attributed totals are far higher, your model is generous.
• Treat attribution as a clue about what works, not proof. To know whether an ad caused sales, you need a test that holds some people back from seeing it.