The problems with attribution

Multitouch attribution is pitched as a nirvana for most B2B marketing teams. Yet there are a few problems with allowing these systems to reach their full potential. In fact, some of these can make an attribution model downright unreliable.

Accounting for unconverted leads

This is, of course, Salesforce specific, but most attribution models deal only in contacts when it comes to opportunity influence. Leads that should belong on the account, but have yet to be converted rarely are considered in this model. The issue here is that you may have relevant, recent activity that is left unconsidered when viewing your opportunity influence.

All Contacts vs Only Contact Roles

An early limitation of attribution tools (and the campaign influence reporting of Salesforce) was that they only looked at contacts that were marked with a contact role on the opportunity. This was problematic because by default, contact roles would only get associated if consciously added by the opp creator. At best you’d get one contact role, being the person who was converted from a lead into an opp, but you couldn’t hope for much more than that. This is the specific limitation that makes the built in Salesforce influence reporting so useless.

Nowadays most of the tools allow you to consider any contacts on the account for attribution weighting, but this creates a new problem – you may be pulling in activity and contacts that aren’t relevant to this opp, and diluting the real influencing touchpoints.

Attribution models are blind to certain touchpoints

Attribution tools come in a couple flavors, and often the differences between the tools come down to how well they track certain touchpoints over others. But bottom line, no out of the box tool is equally effective at tracking all touchpoints, despite their claims. Bizible, for instance, is most focused on tracking digital touchpoints, and it does it well – better than anyone else I’d argue. But their offline tracking leaves a lot to be desired, and requires manual configuration for every touchpoint type.

Bottom line, the value proposition of attribution models is to weigh all meaningful interactions related to the sale – but it turns out it’s tricky to actually pull in all those interactions for consideration.

What about add on and upgrade opportunities?

The tools available today are almost exclusively focused on prospecting scenarios – that is, the initial opportunity for an account. What they’re poor at handling is future buying cycles at an account. Existing customer upsell/cross sell, renewals, and more. They tend to lump all touchpoints into all opps, meaning you’re counting influence for activities that have nothing to do with the current opportunity. This is exacerbated when you have multiple buying groups or business units within the same account – your influence wires will get crossed.

Artificial limits from the attribution providers

Frustratingly, attribution vendors often market themselves as a tool to consider every interaction, yet in reality they place hard limits on the amount of touchpoints they even allow you visibility to. This means you’re looking at only a subset of the interactions that occurred, once again affecting the weighting of your attribution model.

This makes one wonder if attribution is even worth it. These are all solvable scenarios, either with the right mix of vendors, and likely some custom work on top of it. But even with a perfectly humming system, you still have challenges, namely in conveying understanding and buy in from the rest of the organization. Which is arguably even harder to solve. Not to say it isn’t worth solving – but it’s important to recognize attribution is not simply solved by activating a third party solution.