Last updated on May 10th, 2018 at 04:25 pm
It’s been a while since we attended the Google Premier Partners Summit in New York, which means we had a time to review and plan our 2018 data strategies. And if there’s one thing that was repeated through all the sessions – it was definitely the attribution.
What is channel attribution?
During the Summit session, attribution was explained in a very simple way: if you left something under your seat in row 10 and ask for it to be passed to the front, your valuable will be passed from hand to hand between 9 people before reaching the person closest to you. They effectively return the item to you, but would they have been able to do that on their own? No, the same way as the last click in a path to conversion is not solely responsible for generating that conversion. You must analyse all steps and engagement points in user’s journey to understand the full picture – what attracted them in the first place, what helped them in their competitive research and finally what convinced them they should buy, sign up or book with you. Giving relative credit to all those engagement points is what we call attribution – and its crucial to base all your future marketing planning on this approach.
Why attribution matters?
Look at your Google Analytics data today. What % of completed conversions or transactions is credited to Direct channel? How likely is it that someone went directly to your site, converted and left, with no prior research or comparison? In the digital world brand loyalty matters a lot less, as users focus more on convenience and value. And if you want to see how much this trend impact your online business, check this one simple report:
In Google Analytics, under Conversion reports, under Multi-Channel Funnels, view Path Length report. At the top, select the key conversion type your analysing, like Transaction or specific Goal and you are ready to view the results:
From this example, we can see that over 58% of all revenue came from paths with 2 or more interactions – meaning that the user engaged with your brand twice or more before converting. If your analysis is based on last-click model only, all that revenue is attributed to the last touch-point, which clearly isn’t accurate. And if you still need convincing, quickly switch to the Top Conversion Paths report, to see just how complicated some of the user behaviour can be.
Planning your marketing activity, and more importantly budgets, brings in huge risks from business perspective. On optimisation level, you are most likely missing out on cost-efficient opportunities to connect with users earlier in their research and convert them faster, minimising the risk of losing those leads to competitors.
On budget level, you might be dismissing channels which engage users at the very start, introducing them to your brand and product. This could be video, social or display, as it’s often the visual channels which initial capture the attention. Just because the user eventually clicked on organic result and converted, doesn’t mean that they suffered from amnesia and forgot about seeing and engaging with your other ads and channels.
What you can do to prepare for attribution in 2018?
Like with every good strategy – start with analysis. Within Google Analytics you have access to a whole suite of reports which can help you understand the complicated conversion ecosystem better and see many of the digital channels in new light. From Assisted Conversions to Model Comparison Tool – don’t be afraid to dive into the data and ask the basic questions: how many engagements does it take for my website to convert average visitor? Which channels introduce, which assist, and which complete the conversion? How to best structure your creative messaging for 2018 to take advantage of those learnings and deliver message which are more relevant to the research stage the user is at?
From here, decide how you will define success in 2018 and how to assign more accurate metrics to different stages of customer engagement. Even if you’re not mentally prepared to move away from the last-click attribution model, include other models and assisted conversions in your reporting and analysis, to ensure you’re not missing any important data. And start preparing yourself to say good-bye to the outdated methodology, which most likely already lost you some profit. If you accept the fact that last-click attribution is completely inaccurate – you will find it a lot easier to make a leap forward and catch up with you customers. They’ve been waiting for you in the multi-touch point universe for a while…