What is Attribution Modeling?
Attribution Modeling is essentially assigning credit to each touchpoint in the consumer journey. This helps to show what role each channel (PPC, Social, Organic) is taking in a consumers path to purchase.
Why do you need Attribution Modeling?
One of the biggest frustrations our clients face is not being able to tell how effective their overall marketing strategy is. Sure, they can tell how many leads or sales they are getting from their PPC, Organic, Social and Direct traffic separately, but what about overall?
The typical consumer journey has many touchpoints across devices spreading over weeks and sometimes even months. How can we measure the effect of all touchpoints on the eventual conversion? Attribution Modeling.
Why is it so useful?
Attribution modeling helps to show exactly how a customer interacts with your brand online. Using this information can help you optimize your messaging and content to assist with speaking directly to that customer which in turn can increase conversions. For example, if your data shows that most people are exposed to your PPC ads first, you might want to change the messaging to be less sales-y. Your communication would be focused on informing them about what you do and building a relationship. You could then treat PPC as a way to introduce yourself to new customers.
In addition to being able to shape your marketing efforts, attribution modeling is a viable way to accurately represent the ROI of each channel. Each individual channel gets its’ respective credit for each conversion. This allows you to see how much revenue each channel is responsible for.
If one channel is completing the majority of total conversions it could influence your marketing planning for next year. If you want to continue to boost that channel, you could invest more heavily into it. Or, on the other hand, you could want to invest more heavily in your other channels in order to increase overall conversions and have them convert at a similar level as your dominant channel.
Examples of Attribution Modeling
Although there are many types of attribution models (and you can create your own custom models) Google analytics has 5 standard types of attribution modeling. The total number of conversions does not change with attribution modeling, simply the credit is divided more evenly. For example, if you have 35 conversions total (15 Organic, 5 Social 15 PPC) your total conversions will still be 35, however, the distribution may be different – 15.5 Organic, 3.5 Social, 16 PPC.
Last Click Attribution
Last click attribution is the standard attribution model utilized by Google Analytics & Adwords. It gives full credit to the final click before conversion. If the last click before conversion was on a PPC ad, that channel, PPC, would receive 1.0 conversion.
First Click Attribution
First click attribution is another very popular attribution model. It gives full credit to the first click in a chain of touchpoints. Therefore, if a user clicks on an organic result first, but ends up converting from clicking on a PPC ad a few days later, the organic result would receive 1.0 conversion.
Linear attribution assigns an equal value for each and every touchpoint in a consumers journey. For example, if a user clicks an organic listing, clicks a PPC ad, and also clicks to the website from Social media, each channel will receive 0.33 of a conversion regardless of which channel led to the final conversion.
Time Decay Attribution
Time decay attribution assigns a different value to each touchpoint in the consumers journey based on when it took place relative to the conversion. The channel that led to the final conversion will receive the highest value (as it was the most recent point of contact) and the touchpoint that took place the longest ago will receive the smallest. For example, Organic leads to the conversion after a user clicked on a PPC ad 1 weeks ago and a Social post 2 days ago until finally converting today from organic results. A potential breakdown of the attribution could be as follows: PPC: 0.1, Social: 0.3, Organic: 0.6. The further in the past the contact was made the less value is attributed to that channel.
Position Based Attribution
Position based attribution assigns 40% credit to the first and last touchpoints, then divides the remaining 20% between the middle points of contact. For example, if the first click is Social, second is organic, and third is PPC, the conversion distribution would be Social: 0.4 Organic: 0.2 PPC: 0.4.
Which Attribution Model is Best?
There is no right answer to which attribution model is best. The fact is it really depends. It depends on your business, your marketing efforts, and what you would like to be tracking.
In general, first and last click attributions are not very effective at showing the effect of each channel on the final conversion, however they are very good at measuring 1-to-1 click conversions. Linear, Time Decay, and Position-Based all relate to your business goals.
My personal favourite is Time Decay as I feel it provides a clear representation and value for each touchpoint.