For ecommerce businesses seeking a way to understand your customers’ digital paths to conversion beyond the last click, it’s a challenging time.
Attribution was already a tricky piece of analysis, but with GDPR it’s become that bit harder – there are now additional barriers to obtaining Google and Facebook data in your log-files. And the dedicated tools that you can buy are expensive and complex to set up and use.
However, all is not lost. There’s a lot you can still do to push forward your deeper measurement plans. Here’s three things you can do:
1. Make the most out the analytics tools you already have
Many ecommerce players are already using tools like GMP (Google Marketing Platform – the new name for Double Click Manager and Google Analytics 360), Sizmek or Ensighten to manage their campaign data and are not using the full capabilities of these tools. If you have licences for these tools, then you already have attribution features that can be very powerful and yield valuable insights, without the need for extensive additional setup and implementation.
GMP, Sizmek and Ensighten offer data-driven models that can cover all digital media channels and be used as an alternative to dedicated attribution tools. They have various tools that allow you to measure marketing activity from a multi-touchpoint perspective, providing insights into specific patterns within your consumer paths along with a view into re-attributed performance of your media campaigns.
These tools aren’t quite as advanced as dedicated attribution platforms and will have certain limitations – such as the DDA (data-driven attribution) model being restricted to the last four touchpoints within Google Analytics, and the ability to stitch Facebook mobile or custom audience impressions into converting paths.
However, they do let you look at performance at different levels, including channel, campaign, ad group, placement and creative, from a multi-touch perspective in the short to medium term.
And 55’s studies of our clients’ reports have shown that custom reporting models can be set up that vary from the dedicated tools by less than 1% per channel.
2. Prioritise the easier insights first
Clients are often disappointed when they first turn on their attribution tools to find that they cannot see any actionable insights from the reports. And when they shift budgets, nothing seems to “move the dial”, in terms of recognisable changes in attributable conversion credits for a given channel.
We recommend that you optimise within each channel before you look for improvements between channels. Attribution tools must be maintained carefully by all internal and third-party agencies to generate reliable results, so that each campaign is set up correctly, with naming conventions, taxonomy and channel groupings all in place.
Looking at display advertising, for example, budget allocations between prospecting, retargeting and performance budgets can be improved using attribution analysis first. And before you start moving budgets into or out of the display channel, investment in video, banners, and other formats within each of these budgets should be optimised first too.
You will also need to make sufficiently bold shifts in budgets – i.e. large enough for the DDA algorithms to learn about your optimal customer paths – before you will build the confidence in the tool’s insights and before you can be sure that it is fully factoring in the changes as you intend
3. Use additional studies to augment your insights
We would always advocate that attribution studies are conducted in line with a range of qualifying insights that come from your knowledge of the peculiarities of your business performance and the effects of offline, PR and other factors on the performance of any given analytical study.
Some channels just cannot be included in your analysis with the same level of granularity as others – most notably paid social. Conversion and brand lift studies should be used to evaluate channels like this, so that you can incorporate variations in spend against the attribution insights you obtain in your DDA model.
Particularly heavy spenders in Facebook should consider setting up Facebook Attribution, which is a tool that focuses on optimizing budget within the Facebook and Instagram ecosystem and came out of beta in October 2018.
Also merging digital attribution with offline MMM (media mix modelling) will provide you the most comprehensive viewpoint in terms of allocating credit to digital channels within the wider context of all your media investments.
This is going to require a different type of test-and-learn framework than typical DDA models because the offline data cannot be attributed in the same ways, but I would be surprised if the questions that come from the most senior people in your business do not fall into this area – digital does not work in a bubble, even for ecommerce businesses, and the understanding the role of each channel needs to be combined with real-world insight to be credible.