Multi Touch Attribution Examples: 5 Customer Journeys and How Each Model Credits Them

Multi touch attribution splits credit for an order across the clicks that led to it, instead of handing everything to one touchpoint. The model you pick decides how that split works, and different models can tell you very different stories about the same sale.

Definitions only get you so far. So below are five hypothetical ecommerce journeys, each run through five multi-touch attribution models, with the dollar math shown. By the end you should know which model to open for which decision.

What are the multi-touch attribution models in these examples?

These examples use five models, each answering a slightly different question. If you want the background first, start with our explainer on what multi-touch attribution is.

• First click: 100% of the order to the first click in the journey.
• Last click: 100% to the final click before purchase.
• Linear: equal credit to every touchpoint.
• Paid linear: equal credit to paid touchpoints only. Email and organic visits get nothing.
• Lighthouse: AdBeacon's model that works like linear but credits only the platform being optimized. It's built for in-platform decisions, not omnichannel reporting, so its numbers aren't meant to be added up across platforms.

One more note. Google removed first click, linear, time decay and position-based models from Google Ads and GA4 in 2023, keeping data-driven and last click. That's why comparisons like the ones below now need an independent tool, a point we covered when Google dropped four attribution models. All examples assume click-only tracking, so only verified clicks count as touchpoints.

Example 1: The impulse buy

When a journey has a single click, every model gives the same answer. That's the baseline.

A shopper sees a Meta ad for a $60 phone case, clicks it, and buys in the same session.

What it means: for short, single-click journeys, model choice barely matters. If most of your orders look like this, spend less time debating models and more time on creative and offer.

Example 2: The research-heavy purchase

When shoppers compare before buying, first click and last click start to disagree sharply.

A shopper clicks a Meta ad, later clicks a Google non-brand search ad, comes back through an organic search, then clicks a Google brand search ad and buys a $120 order. Four touchpoints.

What it means: last click makes Meta look like it did nothing, even though it started the journey. If you cut Meta based on last click, you may find Google brand search volume drops a few weeks later. Our post on the case against staying on last click covers that pattern in more detail.

Example 3: The email-closed sale

When email closes the sale, paid linear and last click give opposite answers, and both are useful.

A shopper clicks a Meta prospecting ad, signs up for a discount, then clicks a Klaviyo email two days later and buys a $90 order.

Klaviyo will likely claim this order too. Its message conversion tracking uses a last touch model by default, with a 5-day window for email clicks on new accounts. Meta may also claim it. That's one $90 order showing up as $180 across two dashboards.

What it means: use paid linear when you're judging ad spend, because email is a cost you'd incur anyway. Use linear or last click when you're judging whether your flows are pulling their weight. Connecting your Klaviyo data to ad attribution lets you see both views in one place.

Example 4: The returning customer

Returning customers often end on a direct visit, which can quietly swallow credit under last click.

A past customer clicks a Google brand search ad, then clicks a Klaviyo email, then comes back by typing your URL and buys a $150 order.

AdBeacon also offers linear excluding direct, which would split this order $75 to Google and $75 to email.

What it means: direct traffic is usually the result of earlier marketing, not a channel you can buy more of. When you report on returning customers, exclude direct so the credit lands somewhere you can act on.

Example 5: The high-ticket purchase with a long lag

For expensive products, the journey can run longer than any ad platform's window, and every model depends on how far back you can see.

A shopper is considering a $600 product. Day 1, they click a TikTok ad. Day 12, a Meta retargeting ad. Day 30, a Google non-brand search ad. Day 40, an email. Day 45, a Google brand search ad, then they buy.

Check the math: linear splits $600 across five clicks at $120 each, and Google has two of them. Paid linear splits $600 across four paid clicks at $150 each.

Now the window problem. Meta's standard click-through setting counts events within 1 or 7 days after a link click. The Meta click came 33 days before purchase, so Meta's own reporting wouldn't count it under either click setting. A 30-day lookback in an attribution tool would also drop the TikTok click. AdBeacon supports lookback windows up to 180 days, which keeps the first touch in view.

What it means: for high-ticket products, check your typical days-to-purchase before trusting any model. Foresight Sports, a higher-ticket brand, saw orders rise 119.9% and ROAS rise 96.1% over 90 days with an agency partner, as covered in the Foresight Sports case study.

Which multi-touch attribution model fits which decision?

Match the model to the decision, and compare at least two before you move money.

For more on these tradeoffs, see our attribution models comparison for ecommerce. To see how AdBeacon lays model comparisons next to platform-reported numbers, look at the attribution reporting overview.

What to do next

1. Pull 20 recent orders and map their click paths. Count how many have one click versus three or more.
2. Check your median days to purchase. If it's longer than 7 days, platform-reported numbers are missing early clicks.
3. Compare first click and last click by channel for the last 30 days. Flag any channel whose revenue swings by half or more.
4. Report paid media on paid linear, and owned channels on linear or last click.
5. Use Lighthouse only for in-platform scaling decisions, never as a cross-channel total.

Multi touch attribution doesn't hand you one correct answer. It hands you several views of the same order, and the differences between them show you where a channel is starting journeys, closing them, or borrowing credit. Pick the model that fits the decision, and check it against a second one before you move budget. If you'd like to run these five comparisons on your own store's orders, see your multi-touch paths with our team.

Frequently Asked Questions

What is an example of multi touch attribution?

A shopper clicks a Meta ad, then a Google ad, then an email, and buys a $90 order. Under linear multi touch attribution, each of the three clicks gets $30 of credit.

What is the difference between linear and paid linear attribution?

Linear splits credit equally across every touchpoint, including email and organic visits. Paid linear splits it only across paid ad clicks, which gives a cleaner read on ad spend.

Why can't I use first click or linear in Google Ads anymore?

Google removed first click, linear, time decay and position-based models from Google Ads and GA4 in 2023. Data-driven and last click remain, so comparing rules-based models now requires an independent attribution tool.

Which attribution model is best for high-ticket products?

Start with linear or first click and a long lookback window, because high-ticket journeys often run past 7-day platform windows. Then compare with last click to see what closes.

Should I add up Lighthouse numbers across platforms?

No. Lighthouse credits only the platform being optimized, so it's designed for in-platform decisions rather than omnichannel totals.

Sources

• Google Ads Help: First click, linear, time decay, and position-based attribution models are going away
• Klaviyo Help Center: Understanding message conversion tracking
• Meta Business Help Center: About attribution models and attribution settings