Conversion attribution: what, how and why?

“Everything can and must be measurable” is a motto that, as far as we’re concerned, would look perfect on a plaque. At Traffic Today , by ‘making things measurable’ we mean providing insight into all kinds of website statistics. However, making things measurable is one thing. The real skill often lies in correctly interpreting this data. And that can be achieved using conversion attribution.

Tools for analysing your statistics, such as Google Analytics, can provide you with a vast amount of data. This data is extremely valuable, particularly when interpreted correctly. Assessing conversions from different traffic sources is a key part of this. In this article, we’ll help you get started with assessing conversions by taking a close look at the various conversion attribution models.

What is conversion attribution?

Conversion attribution is the process by which a specific conversion (checkout, registration form) is attributed to a specific stage in the customer journey that was necessary to achieve the conversion. Conversion attribution therefore looks not only at the final interaction, but also at the earlier steps that were necessary to achieve the conversion.

Suppose a user purchased a pair of white trainers after seeing an Instagram advert, but had previously visited your online shop after reading a blog about the top 10 shoe trends of 2024 and subsequently signed up for the newsletter. In that case, all these steps have contributed to the conversion click, not just the final ad click. We call this conversion attribution.

Practical example of conversion attribution

Let’s look at a practical example to clarify this further. It is important to note that this example is based on the ‘last-click model’, which means that the final touchpoint before conversion takes place is credited with 100 per cent of the value points for the sale.

Suppose you run an online shop selling televisions. You use various marketing channels to promote your product. For example, you advertise extensively on Google AdWords, Google Shopping and Facebook, and you also do some affiliate marketing. You’ve put together a great marketing mix for yourself, ensuring that you have a specific message visible at every stage of the customer journey.

A potential customer first visits your online shop via your Facebook advert. They’re interested in what you’re offering and, a day later, Google your company name. Via an AdWords advert, they return to your online shop and decide to buy a television. A great sale!

Logically, you might now conclude that the Facebook advert did its job well. After all, without this advert, the customer would not have found their way to your online shop in the first place. However, Google Analytics attributes the revenue to Google AdWords because this is the most recent traffic source used. This happens another fifty times this month, and at the end of the month you decide to switch off Facebook because Google Analytics indicates that it isn’t delivering enough for you. You probably realise where we’re going with this…

How do you apply conversion attribution?

It starts with thinking logically and looking beyond the obvious. If you bear that in mind and follow the practical tips below, conversion attribution can provide you with highly valuable information.

1. Path length

Take a close look at the path length in Google Analytics. This indicates how many interactions took place before a conversion was achieved.

Suppose someone first visits your site via AdWords, then later via organic (unpaid search engine results) and then places an order; this constitutes a path length of 2. The longer the path, the more important it is to accurately assess a particular traffic source. This is because there have been more points of interaction and more traffic sources involved.

You can find the path length in Analytics under Conversions > Multi-channel funnels > Path length.

2. Conversion support

In Analytics, be sure to take a close look at the ‘assisted conversions’ tab. These are all conversions from specific traffic sources where the conversion has been a contributing factor. Google uses the following definition:

“A supporting interaction is any touchpoint that lies within the conversion path but is not the final interaction.”

For example, a particular traffic source may deliver significant value at the start of the funnel but not as the final click. You can find the assisted conversions in Analytics under Conversions > Multi-channel funnels > Assisted Conversions.

3. Delay

It is also important to look at the ‘delay’ report. This report indicates the number of days between the first click and the actual conversion. This metric is particularly valuable in sectors where this delay is relatively high. Think, for example, of insurance companies or travel agencies. Based on this data, you can set up a robust remarketing strategy that allows you to bid aggressively at specific moments.

4. Standard attribution models

Google itself already offers a number of models in Analytics that allow you to use conversion attribution in various ways:

• Last interaction
  100% attribution to the last traffic source.
• Last non-direct click
  100% attribution to the last traffic source, excluding direct traffic. This model is, incidentally, the standard model in Google Analytics for conversion and transaction value attributions.
• Last AdWords click
  100% attribution to the most recent AdWords advert that was clicked on.
• First interaction
  100% attribution to the traffic source that generated the first click to the website.
• Linear
  Proportional attribution to all traffic sources involved in the purchase funnel.
• Time decay
  Attribution based on a specific period. This model assigns the highest value points to the touchpoints closest to the moment of conversion.
• Position-based
  This allows you to create your own hybrid attribution model. For example, you could attribute 40 per cent to the first click source and 20 per cent to the last click source.

5. Comparing attribution models

You can easily compare the attribution models mentioned above in Analytics. You’ll find this feature under Conversions > Attribution > Model Comparison Tool.

Conclusion

It is therefore not (always) correct to rely solely on the standard reports in Analytics (and thus on ‘last-click non-direct’) to generate valuable insights. Every traffic source, campaign and conversion is different, and a thorough analysis is essential for a proper assessment. We hope this article has set you on the right track when it comes to conversion attribution. After all, Google Analytics offers you, the user, a vast wealth of information. However, it’s up to you to make the most of it.

Need help? We’d be happy to assist you with AdWords campaigns, conversion optimisation and much more. Get in touch for a no-obligation chat and discover what Traffic Today can do for your online performance.