Comparing Revenue Attribution Methodologies: What to Look For
Unless you are already familiar with revenue attribution and are fairly technical/analytical, knowing what to look for when comparing competing offerings can be daunting. Many companies will show flashy dashboards and other eye candy, but talk very little about the underlying methodology. Sure, whatever methodology they are using may be better than double counting sales or the last touch method you are using currently, but is it accurate or actionable?
Below are four important points to look for when comparing Revenue Attribution methodologies:
1. Attribution shouldn’t be guess work
At one recent conference a competitor proudly announced that his company had seven different methods of attributing revenue. Wouldn’t that mean that at least six, if not all seven, are incorrect then? Attribution should be a function of what drives sales at the customer level – the model should identify the value of each treatment rather than some arbitrary method.
2. Attribution should be designed to handle any complexity or combination of treatments that occur
Many of the early methods that marketers adopted were simple and flexible, but generally lacked accuracy. In fact, many frequently lacked offline and brick and mortar data, or were unable to integrate disparate data sources, which limited the multi-campaign view.
Additionally, in the last ten years the marketing landscape has changed dramatically. Many new channels – social media, affiliate marketing, retargeting, comparative shopping sites, search, mobile – have grown in complexity and fight for credit with the other overlapping marketing treatments that customers experience.
Accurate marketing attribution measures the value of each marketing treatment, along with the decay of a treatment’s impact over time, for each channel. The methodology should be designed to handle any number or combination of treatments that occur, both online and offline.
3. Attribution should be based on the incremental effect of marketing
Not all sales can or should be attributed to marketing. If you turned off all of your marketing today, you would still see some portion of sales coming in that is Customer Driven. If you have a retail store, a portion of retail store sales may be attributed to marketing, but marketing shouldn’t receive credit for the person who happened to walk into the store and purchase while waiting for an appointment across the street.
For more established brands, brand equity plays a larger role than for newer brands with a smaller loyal customer base. All of the millions of dollars that were spent over many years of marketing play into that brand loyalty. If a customer has a habit of shopping in your store every Wednesday after lunch for 20 weeks in a row, and again shops in week 21– should the email that you happened to send that morning receive the credit, or should some or most of that credit go to Customer Driven?
The point of revenue attribution is to know what is working and to understand what levers to adjust to optimize marketing spend. Without giving appropriate credit to customer driven sales, too much credit would go to the marketing channels. This could lead to overspending in an area that may not drive the anticipated incremental sales.
4. Attribution should be used for targeting as well as giving historical credit
By definition, revenue attribution is retrospective. All of the sales have come in and the marketing budget has been spent. Attribution is simply dividing up the revenue to the appropriate channel.
…And then what? What do you do with that information? Marketers need attribution to be actionable. DataSong’s attribution technique enables targeting as well as simultaneous optimization of multiple marketing channels.