Multiple-touchpoint marketing strategies often incorporate marketing mediums that don't naturally fit into "touchpoint" attribution models. These mediums can generate considerably profitable campaigns, at what often seems like very little incremental marketing spend. To accurately measure cohort profitability, your attribution model must consider both the revenue and cost implications of these support channels.
Support channel definition
Your marketing landscape will exhibit two types of marketing activities: Primary channels and support channels. The definitions of each are constrained by how the channel generates revenue, rather than how traffic is generated.
Primary marketing channels: revenue from these channels skews towards directly attributable mediums. Examples include SEM, direct traffic, or unattributable mediums where you're able to reattribute revenue through matchback attribution.
Support marketing channels: revenue generated through these channels is frequently backed by tools that claim their own attribution measurements. A few examples of support marketing channels include email, onsite exit popups such as those from services such as Yieldify, onsite CRM reengagement activities such as targeted promotions based on browsing behavior, and voucher or coupon affiliate sites.
Support channels are particularly tricky to work with because it is difficult to prove their incrementality. This is because revenue from these channels may already be captured through more traditional click-based attribution models, suggesting they're not worth the time, effort, or marketing spend.
Reattribute support channel revenue
Support channels and reengagement activities are similar to the supporting beams of a wooden pier. Without them, the top level wood beams might theoretically float and sustain a person walking a short distance. However, this "primary marketing channel only" structure would be relatively inefficient. Underlying support beams, or support channels, would make the structure more sound and remove structural gaps to make the journey more efficient.
Singling out the value from a single beam (single campaign in a support channel), among thousands, can be a nearly impossible task. I don't recommend it. Instead, the value created from each of these mediums can be reattributed to the originating sources of revenue.
Conveniently, the impacts of your support channels will already be felt, in aggregate, by your primary channels. This hinges on the assumption that all of your traffic is exposed to the same support channel systems, but is likely to be the case for most of us.
4 quick use cases
- Emails: you send them to all of your list members, most frequently without judgement of what the source of your email was. Though the email can trigger a sale, the original visit was captured elsewhere. Many attribution models will already include revenue from emails as attributed to the original visit source.
- Exit popups: if you have them set up, its likely that all of your traffic will come into contact with these as a support channel. Revenue will already be distributed between your primary marketing channels.
- Onsite CRM & offsite abandonment retargeting: sites that show promotions based on browsing behavior typically do so with a blind eye against incoming source of traffic. As with exit popups, incremental revenue will already be distributed between your primary marketing channels.
- Voucher affiliate sites: they rarely generate incoming leads from casual browsers, but thrive on the claim of strengthening checkout passthrough rates. Unless you actively restrict which of your traffic is able to use vouchers, we can assume that voucher use is evenly distributed and therefore revenue from voucher affiliates is likewise evenly distributed.
Reattribute support channel cost
As you can see, in pretty much all of these scenarios, traffic coming in contact with a support channel is already likely to be distributing its respective revenue benefits. This means that costs of support channels can be distributed evenly as well.
The way to do this runs on the same principle as previously described in how to blend data for reattribution, the only difference being that you distribute costs instead of revenue. Specifically, distribute proportional support channel costs into the proportional cost composition of your overall marketing budget split.
A Caveat
Sometimes certain traffic won't feel the impacts of support channels because you've constrained the traffic from doing so. Simple solution - just exclude the cost (and revenue) from those campaigns.