Addressing Incrementality in the Affiliate Channel
From Wiki
In deciding how to apportion spend and resource to each channel, advertisers need a clear view not just of the volume and ROI each produces but the effect of these channels on each other.
The place of the publisher channel in an advertiser’s online marketing mix is broad but complex. The average publisher will promote an advertiser in a way that mixes a multitude of different promotional methods, breaching the boundaries of traditional publisher categorisation. In deciding how to apportion spend and resource to each channel, advertisers need a clear view not just of the volume and ROI each produces but the effect of these channels on each other.
In the publisher sphere, as elsewhere, advertisers raise questions bearing on the extent to which the sales they receive through this channel are incremental. Commonly-asked questions include, ‘Would the customer have bought already?’, ‘How do I know I am not paying for sales that I would have got anyway?’ and ‘Are publishers capitalising on promotions that are financed by our business to give their own campaigns a boost?’
Whilst this article will not attempt to address definitively all the issues surrounding this difficult topic, it will try to give advertisers a perspective from which to assess how incremental their publisher sales are.
What is ‘incremental’?
Whilst certain types of publishers have come under greater scrutiny than others when questions of incrementality arise, these methods reflect – and possibly drive – contemporary ways of shopping online. The rise of voucher coding and cashback has paralleled the rise of the smart, deal-conscious consumer. There first needs to be therefore recognition on the part of advertisers that questions of incrementality should be framed by the reality of online shopping habits. Just as it is easier to shop online so competition for the same customers over the same products becomes fiercer; and as the economy weakens, consumers become more eager to shop around to find a bargain.
Achieving total incrementality is as unrealistic an expectation for the publisher channel as it is for any other marketing channel: a certain proportion of customers will have transacted on an advertiser’s site regardless. Similarly, in a competitive environment it is not realistic to expect that incremental sales equate only to those from new customers. Just because a customer has bought before does not necessarily mean that they will do so again.
Attribution and De-duplication
The question of incrementality is inextricably linked to the question of attribution. Customers’ paths to purchase are increasingly complex, a large number of touchpoints might be involved, and ascertaining the influence that those touchpoints have on the customer’s eventual decision to buy is not something that can be fully measured. Advertisers will need to be able to answer three questions which, whilst ostensibly simple, conceal a number of important complications. First, what these touchpoints are. Online channels are easy to measure, but offline media (TV or print advertising that contributes to a customer’s knowledge of a brand later intent to buy) is more difficult. Hence why sufficient depth to attribute fully can never be obtained.
Secondly, how to track and measure them. Advertisers may choose a third-party tracking provider de-duplicate channels against one another, but the way this is done should be reasoned and transparent. Rather than looking at each channel separately advertisers need seek to understand which multi-channel paths are most effective at producing conversions. For example, whilst some may believe publishers are often being incorrectly credited for sales that are not incremental, a publisher may counter that an advertiser’s brand paid search more often credited for a sale that the publisher should be acknowledged for. Brand paid search is very rarely incremental. Advertisers often use de-duplication as a way to ascertain incrementality amongst publishers, but the logic by which de-duplication solutions are implemented (what rules determine the activity credited with the sale and the relative value of these in the customer journey) should be considered carefully based on an understanding of the how customers interact with different touchpoints prior to the sale. Thus advertisers should work from attribution to de-duplication rather than vice versa. In other words, de-duplication is not an attribution strategy.
Thirdly, what value to attach to each of these touchpoints. This is the crux of the issue of attribution which will inform assessments of incrementality. Whilst advertisers may sometimes believe they are paying overly high acquisition costs to publishers with insufficient demonstrations of incrementality, it should be remembered that resulting customers can be remarketed to and re-engaged by the advertiser at a later date independently of the publisher. It might be argued that those customers may nonetheless choose to return via a publisher, but the consumer’s choice of where to shop will be their own, and it is surely better that a customer return to the advertiser via a publisher than to a competitor. Changing the approach to incrementality
This article does not claim to answer the more precise concerns that are sometimes raised by individual advertisers. These will often be unique to the market and competitive environment each operates in. However, the above represents some way in which advertisers can approach the question of incrementality more broadly as a basis for examination of concerns particular to them.