Which marketing is actually making me money?
Understanding marketing attribution and which marketing actions have the biggest impact on the bottom line is crucial not only to understanding the efficacy of a campaign but also to growing your business.
Every year, millions of pounds are wasted on marketing that fails to deliver, even worse millions are spent on marketing that fails to deliver but with no understanding of why or how it has done so.
What part of your marketing first made someone aware of your company? What informed their purchase decision? What finally convinced them to actually buy your product or service? And which one is most important? Knowing the answers to this will help you understand where you should be spending your valuable marketing budget.
Welcome to the world of marketing attribution.
Marketing attribution has become more and more prevalent alongside the growth of digital marketing. Previously, selected viewing households, focus groups and customer surveys were how organisations understood whether their TV advertising or billboards were effective. These methods all have inherent weaknesses that often revolve around small sample sizes & people’s opinion of their habits and actions being different from what they actually do. Today there is a wealth of digital data at our fingertips that provide accurate, real-time insights on consumer activity. This data is an incredibly valuable resource to marketers but only when used correctly.
Much has been written about false metrics in digital marketing; “My post got loads of clicks so that must be delivering value for my organisation” etc. Whilst engagement is always desirable with marketing it means nothing if none of those people buy anything from you or encourage others to do so. So, understanding which metrics matter is vital to iterating your marketing but as far as attribution is concerned this only part of the story. You can know that you ran an ad and 25 people who clicked on that ad converted to sales when directed to your website. Great, you now know how many impressions – clicks – conversions that campaign delivered. But what if the ad was only part of the story (hint: it definitely is), what if after seeing your ad they then read a few of your blogs to learn more, then used a live-chat on your site to ask one of your team some questions before then purchasing your product?
What has been the most important step (or touchpoint) on this customers journey to purchase? The ad that caught their attention? The blog that informed them? The customer service agent that helped them decide to buy?
All of them are important but not in equal ways.
Marketing attribution models
There are a number of different attribution models the place emphasis on different touchpoints. “First-touch” would allocate all the importance to the ad that first caught their attention, “Last-touch” to the live chat that convinced them to buy. These one-touch models have a place but are limited if you really want to understand the effectiveness of your all touchpoints on the customer journey.
There is a “Linear model” that will equally allocate credit to every touchpoint irrelevant of which ones are more effective and a “Time-decay” model that allocates diminishing credit the further from conversion a touch point is; which can massively under-play the importance of earlier touchpoints.
There are further models, such as the “U” and “W” model but like the aforementioned approaches they can undervalue certain touchpoints on the journey.
The most effective approach
The approach that is the most effective is a “Custom” or “Algorithmic” model. This model is often ignored as it can be complex to set up but it has by far and away the most effective results as it is unique to each business. This model involves scoring each touchpoint for what it is actually worth to the conversion. This model becomes more effective with the more data you feed into it, as the millions of customer journey permutations can be analysed side by side (like a multivariate test) to properly understand the impact each stage has.
Take our previous example of a relatively straightforward journey of ad – blog – live chat – conversion. By comparing 100 customers who followed this journey to 100 customers who followed an email – blog – live chat – conversion journey we can understand the weighted effectiveness of email v ad on this particular journey. Obviously both the ad (changes of content or where it is displayed) and the email (subject line, content, sending time) have numerous variables as well. When you start to look at all the variable permutations it becomes clear why many marketers turn towards algorithms for this process, feeding in all the variable data, to determine which touchpoints are the most valuable on the customer journey.
Once you have established this you know the ROI from all of your marketing activities, how each of them impact your customer acquisition cost and which ones you should invest further into and which you should scale back. This will never be a perfect system as there are many offline actions (trad. Advertising; magazines, PR, radio) that are difficult to track and your model will only become more effective as more data is fed into it. But despite these drawbacks these custom/algorithmic models are, currently, the most effective way for marketers to answer that holy grail of what is my return on spend for each and every touch point.