Conquering attribution: Make every interaction countby
The customer journey is highly complex - and while most organisations think that they are tackling attribution in order to understand the impact of marketing on those journeys, the process which they have in place is not one which takes into account today’s intertwining of numerous digital touchpoints. To make matters worse Marketing teams are siloed, with various departments working in isolation from each other, resulting in uncoordinated campaigns with the true value just not being measured – or organisations finding it very hard to do so.
Attribution is becoming an ever greater marketing mountain to climb – so how do you simplify digital attribution and make the mountain easier to conquer than it feels from the foot of it?
A joined-up approach
Many organisations are simply not aware that current attribution models are disjointed and inaccurate. Various teams are working in isolation – all designing and running campaigns in an uncoordinated manner. As a result, when looking from a customer perspective to see which campaigns are driving the most conversions, each team is usually only associating value to its own particular channel. There is also often a further divide between offline and online marketing activities. With the customer potentially involved in multiple touchpoints, there is no visibility of the customer journey – and no way to attribute value accordingly.
So where is the starting block for organisations to achieve this joined up attribution across their digital channels? Firstly, it all comes down to data – getting data from the various systems to understand that digital customer journey – from marketing emails to Pay Per Click (PPC) keywords. Once the data has been identified it must be brought together and integrated, therefore identifying the multiple digital touchpoints per customer.
But data integration is just the start. How do organisations then attribute a score to each touchpoint? Is one touchpoint of more value than another – should the touchpoint closer to the point of sale be ranked highest? Organisations need to use internal discretion as to how each touchpoint will be scored based on their own business and subsequently build a model which values all touchpoints leading up to the end result, whether it is registration, sale or purchase.
Score attribution needn’t be a complex task. Basic attribution models which organisations have been implementing are exponential or uniform: with uniform, all touchpoints get the same score; and with exponential the touch point closest to the point of sale gets the highest score and the touchpoint furthest away gets the lowest.
Score attribution needn’t be a complex task.
Building on these models, companies can then become more sophisticated, developing attribution models that incorporate the value and cost of the marketing activity, the level of engagement of an individual customer after viewing the campaign (e.g. whether items were placed in the basket or the number of clicks) and the brand awareness and loyalty driven by the campaign activity. Scoring each digital touchpoint on the basis of a number of variables that reflect each customer’s visit is more complex but also a logical progression from the uniform and exponential attribution models.
At a time of growing scrutiny on the marketing budget, this kind of attribution is becoming more and more essential. With accurate attribution models, companies can improve the quality of the marketing spend and make better decisions about channel investments – for example, bidding on different keywords depending on whether a customer is on tablet or desktop device.
Given the sheer scale of opportunity offered by analytics, it is incredibly important to balance objectives with realistic goals - don’t try to run before you can walk, let alone crawl. The prioritisation process is not only about identifying key business objectives but also ensuring those goals are achievable given the data and analytics skill sets currently in place. Once up and running, a company can quickly evolve through the crawl, walk and run methodology to embrace ever more complex analytics toolsets and drive even more value; but they need to start taking baby steps to make any progress.