Attribution in marketing is the distribution of the value of the user’s actions on the site during different visits from different sources. In order to correctly determine the value of individual traffic sources as part of a marketing strategy, it is necessary to choose an attribution model. Here we explain their types and information they provide.

What is attribution?

As consumers, we know that decisions to buy are not always immediate. When choosing and comparing a product, we can visit the site several times from various sources: through search, social networks, paid ads, email or by clicking the previously saved link.

Suppose that a client learned about our company first ever in social networks. Then he/she went to the site for advertising in the search, saved the link, and a few days later made a purchase by clicking on it from his/her bookmarks. So, there were several visits, but which of these sources brought us the conversion? Depending on the chosen attribution model, the value of this conversion will be assigned to different sources.

What determines the principle of assigning attribution? First of all, your business model. In simple terms, this is how quickly your customers make a decision to purchase, which is directly related to cost, competition, and whether this item/service is a necessity in specific situations. Also, the choice of model can be influenced by the complexity of the ordering process, the number of traffic sources, and many other factors. That is, for different purposes and types of business, a different attribution model is suitable, and therefore there is no unambiguously right or wrong option. All of them are selected individually.

Why is attribution important?

For successful promotion, we need to know which traffic sources bring the most conversions. Attribution helps correctly assess the most important channels, advertising campaigns, ads, or even keywords in achieving your business goals, which is necessary for effective optimization of advertising investments.

Attribution is most important within paid traffic channels. If we are talking about organic and you see that by switching from free search, users perform the most valuable actions for your business, then, in order to increase the number of conversions, it would be logical to reallocate the budget with a bias towards SEO promotion, which is a rather long and complicated process.

There are also unscalable channels. For example, if email marketing is the most effective source, then this only indicates that email marketing is working well and you can think about implementing some improvements. You simply cannot significantly increase the number of conversions because an increase in the number of mailings can also have the opposite effect.

Google ads is just the same paid traffic channel that is easy to scale. Based on attribution, we can increase bids, budgets, and redistribute resources in every possible way to more effective campaigns, keywords, and ads, which is guaranteed to bring good results in a short period of time.

Attribution models

So, the principle the conversion value is assigned to different sources of the user’s transition is determined by the attribution model. There are several models in Google Analytics.

Last interaction. The entire conversion value is assigned to the last source clicked when making a conversion on the site.

Last indirect channel. This model follows the same principle as the previous one. The difference is that conversions by direct clicks are neglected here, as users quite often save a link to the site where they plan to place an order. That is, the conversion is credited to the user’s last indirect traffic source.

Last Google Ads Click. A conversion is counted for the last time a user clicked on an ad before making a purchase.

First interaction. A conversion value is assigned to the first source that brought the customer to the site.

Linear model. The value is evenly distributed across all sources the client used to go to the site.

Time decay attribution model. This model gives the largest share of value to the sources clicked on as close as possible to the moment of conversion. The share of value review period occurs every 7 days. For example, a source that a user clicked on 14 days before a conversion would have half as much value as one that they clicked on 7 days before.

Position-based model. The model counts 40% of the value of the first and last source of the user’s transition. The remaining 20% is evenly distributed to all other transitions.

Learn more about attribution models in Google Analytics:

All of the same models are used to evaluate attribution within a Google ad account but there is also data-driven attribution. In this model, the algorithm evaluates each user interaction with an ad, assigning them a different value based on their importance for the conversion. This model collects and analyzes data about your customers, making it the best option to use within an ad account. Its disadvantage is that it is available only if there is enough information about user behavior over a certain period of time. Only then will the system be able to make predictions and correctly evaluate which paths lead to conversion.

Learn more about data-driven attribution in video:

Comparison of attribution models

By choosing and focusing only on an attribution model, we can miss the channels, campaigns, ad groups, and keywords that the user also interacted with. The Attribution Comparison report in Google Analytics shows you the customer journey from different angles to determine the real value of all stages. For example, if you use last-click attribution, then you may mistakenly consider campaigns that received conversions only within this model as effective. But during the comparison, you can find that at different stages, users interacted with other ads, which means that it also affected the conversion.

It is the comparison of different models that gives us information about campaigns, ads, and keywords that may have been underestimated. This allows the advertiser to see a complete picture of the effectiveness of advertising, and thus confidently change bids and budgets, considering all the stages of interactions on the path to conversion.

For example, this report shows a comparison of two attribution models by first and last interaction. Depending on which sources the user used at different stages of the purchase journey, we see how the conversion value is assigned to different sources within different attribution models.

Associated conversions

It is a Google Analytics report that includes all sources of site visits the user used before the conversion. For example, the path to a purchase took several weeks, during which the user first went to the site through search ads, then through display ads, and then followed a link on social networks and made a conversion. In this case, the conversion will be credited to the social network, and clicks on contextual advertising will previously act as an associated conversion. The logic is that without earlier contact with the ad, the user might not have made the conversion. That is, the assisted conversions report helps us see a consistent user journey, and attribution helps us assign value to the most important stages of this journey.

For example, the report shows how many times each traffic source had an associated conversion, and how many conversions were assigned to this channel by the standard attribution model.

Analyzing the multi-channel funnel report, we can see different traffic sources in terms of how often they are the last or associated conversion channel, or they don’t bring conversions at all. This way we can identify which user behavior patterns most often lead to a purchase. In this case, this is a transition from organic search and then a direct link.

Also, the reports give us an insight into the average duration of the sales cycle on the site, that is, the day from the moment of the first interaction when users usually make a conversion. This example shows that most often users make a conversion on the day of their first visit. Second is the range from 12 to 30 days from the date of the first visit to the site.

All this data helps to assess the effectiveness of various sales channels with maximum accuracy, and thus most correctly optimize and scale the results.

A modern marketer needs to consider the complexity of a user’s path to conversion before deciding to freeze an advertising budget or pause campaigns, otherwise you can miss the most valuable sources of customer interaction. If you want to know how your customers make a purchase decision, which traffic channels are most effective and how many clicks they do before a conversion, please contact Well Web Marketing specialists. We help your business to be informed and make decisions based on real data.