What to spend where? Making sense of marketing attribution

“Half of my marketing works,” the old saying goes, “I’m just not sure which half.”

Therein lies the problem that marketing attribution seeks to address. But successful attribution is trickier than the quote above suggests. The fragmentation of channels across an ever-growing digital landscape means that marketing efforts are rarely split in half (or thirds, or quarters); gone are the days of picking between newspapers and radio or TV and billboards. 

Over the past decade, in additional to traditional offline marketing methods, you’ve had to spend money on your website, blogging, email marketing, social media… the list goes on and on… and you’d love to know which channels are working best. You can focus your efforts on the winners, rethink your strategy with the losers; having an understanding of your marketing attribution helps you know how to best invent.

The good news is that there are tools to help with this. The bad news is that it’s not always easy to know the best way to use these tools. Sure, you can count click-throughs; see who’s using your online coupon, who’s acting on your email marketing, and so forth. But some customers, for lack of a better word, meander. The journey from first contact to conversion is sometimes convoluted.

The most common technique for understanding attribution is to simply count the customer’s “last touch” before purchase. This seems intuitive on one level… but when you consider that some customers’ last touch is also their first, while others’ last touch might be their tenth. One might be an impulse buy; the other a thoroughly researched, deeply considered purchase. Their journeys and their motives are likely very different; and the impact of your marketing efforts on them are likely just as different. Consider in particular the customer who purchased after many “touches”… it was likely a combination of marketing channels that steered him towards his decision. Was the last one really that important? Was it a tipping point, or had the decision been made many clicks prior? 

You see the difficulty. Your question remains: where to spend the money? What’s really driving conversion? Other techniques include “first touch” (attributing conversion to the customer’s initial reaction to your marketing efforts) and “even steven” (which spreads the credit out over all of your customer’s touches). These have their merits but problems remain. Last touch is problematic in the same way that first touch is; it ignores the customer’s journey and takes the significance away from the (potentially many) steps from first touch to conversion.

You can begin to consider a more advanced attribution model after looking at very basic trends—say, the percentages of customers who buy on their first visit to your site vs. those who buy on return trips. Those single-touch customers will likely show what marketing efforts are working (assuming there’s a clear winner and that their conversions are not split across a wide array of channels). This is a valid starting place for knowing how and where to invest your marketing efforts. Keep in mind, though, that this leaves out the percentage of customers who’re converted on return visits… if these longer journeys are responsible for a significant percentage (or a majority) of your conversions, you’ll need to dig deeper and find an attribution model that’s more complex.

For the best results, your data analytics must apply weighted credit to each channel, placement, creative, and campaign. Data must be broken out by every social action, keyword search, impression… the list goes on, but the point is simple enough: to get the most complete picture, you need to leverage all of the data at your disposal to get an accurate measure of marketing attribution.

Do you have any other tips for understanding marketing attribution? Let us know in the comments.

Related Stories

Leave a comment


This will only be used to quickly provide signup information and will not allow us to post to your account or appear on your timeline.

5 Jul 2014, 5:35 a.m.

This is indeed true! For instance, we generally look at content attribution, by the pages person making the purchase visited in his purchase session. For online campaigns we are using utm cookies. For other forms of marketing where we need to provide the business number, we use MyOperator numbers to get tracking of where the person picked the number from. But all these always overlap. So I guess there is no 1 right answer. It can only be 'less right' or 'more right'.