An attribution model is a system that defines how you distribute credit across different touchpoints a customer has before making a purchase.

A touchpoint (or simply a touch) is any moment when a customer visits your website from an external source. For example, this could be from a paid ad, an email newsletter, an Instagram post, or a QR code. Any time a customer comes to your site for any reason, this is considered a touch.

When a customer places an order on your store the Origin system will then distribute attribution credit to one or more touches the customer had before the purchase.

Attribution models help you understand what actually worked in generating sales so you can spend your marketing money in the right places.

Different attribution models allow you to place emphasis on different touches (for example, first touch, last touch etc.).

When viewing your Origin reports you can change your attribution model to get a better understanding of which marketing channels are working best.

Attribution models

Origin supports multiple different attribution models to give you a complete understanding of your attribution data.

First touch

First touch attribution gives 100% credit to the first interaction your customer had with your store.

This is useful if you want to evaluate how your top-of-the-funnel marketing is working.

For example, you might create some awareness campaigns which don't prompt a user to buy immediately but create awareness of your brand and bring customers in your ecosystem. When comparing these campaigns you might use the "First touch" attribution model to compare which first impression was the most effective.

Last touch

Last touch attribution gives 100% credit to the last interaction your customer had with your store before purchasing.

This is useful for identifying which channels/campaigns drive immediate revenue.

You might use this model to compare direct call-to-action ads, e.g. "50% sale, buy now".

Last non-direct

Last non-direct attribution gives 100% credit to the customer's last touch that wasn't direct — skipping over any visits where the customer arrived directly (e.g. typing in your URL or returning from a bookmark).

This solves a common problem with Last touch: customers frequently make a final "direct" visit right before purchasing, which hides the marketing that actually brought them back. By ignoring those direct visits, this model credits the last marketing channel that influenced the sale.

For example, if a customer's journey was:

• visited site from Google search
• visited site from Instagram story
• visited site directly

Last touch would credit the direct visit, but Last non-direct would credit the Instagram story — the real marketing touch that drove the return.

Linear

Linear attribution gives equal credit to each touch in the users journey.

For example, if the user had 4 touches before a purchase:

• visited site from Facebook ad
• visited site from email newsletter
• visited site from Google search
• visited site from Instagram story

Then each of these touches would receive 25% credit for the order.

The linear attribution model gives you a more holistic understanding of the whole user journey and how they interact with your brand before making a purchase. It's ideal if you use retargeting or get customers from multiple different traffic sources.

In the above example if the total order amount was $100, for each of the listed touch points you would see $25 in attributed revenue. Of course, each touch point did not generate exactly $25 of revenue but this gives you a good idea of the split of "influence" each touchpoint gives.

Any click

Any click attribution gives 100% credit to every touch in the user's journey — each touchpoint is treated as fully responsible for the order.

Using the same 4-touch journey above (Facebook ad, email newsletter, Google search, Instagram story) on a $100 order, every touchpoint would show the full $100 in attributed revenue.

This model answers the question "which channels were involved in this purchase at all?" rather than splitting credit between them. Because each touch receives the full order value, attributed revenue across different channels will add up to more than the order total — this is by design, and is the key difference from Linear. It's useful when you want to see the complete set of channels that played any part in driving a sale, without diluting their contribution.

Origin attribution

Origin attribution is our own multi-touch model, designed to give the most balanced picture of which channels drive your sales. It gives 100% credit to the most recent touch in each channel. Direct and untagged visits are ignored, and where a customer interacts with the same channel more than once, only their most recent touch in that channel is credited.

For example, if a customer's journey was:

• visited site from Facebook ad
• visited site from Google search
• visited site from Facebook ad (again)
• visited site from email newsletter

Then Facebook (the second, more recent visit), Google, and email would each receive 100% credit — the earlier Facebook touch is dropped in favour of the later one.

This avoids over-counting a single channel a customer keeps returning through, while still recognising every distinct channel that contributed to the order. As with Any click, each credited channel receives the full order value, so totals across channels can exceed the order amount — but within any single channel, an order is only ever counted once. It's our recommended model for understanding the true channel mix behind your revenue.

Updated on: 03/06/2026