B2B marketers generally have an easier job of measuring omnichannel campaigns. This presentation explained some of the techniques used by leading consumer companies that may give some ideas for omnichannel measurement for B2B.

One example might be looking at advert view ability data. A simple approach would be to measure on-target viewability (the ads that would have been seen by the target audience). A simple compound metric might be the viewable CPM (vCPM). A further step might measure the cost per unique visitor from the target audience or cost per on target contact acquired. Interestingly as the compound metric gets more complex, they often get closer to business goals.

The presenter described two types of metrics: The first, table stakes metrics, for example view ability, clicks, and other simple metrics. These are always useful but only tell part of the story. Compound metrics are much more useful, combining multiple data sources.

A charity case study was presented, where the charity was both prospecting for new doners and retargeting existing donors. Based on CPA and Return on Ad Spend, retargeting worked better than prospecting. If looked at purchase volume and cost per unique user, then prospecting showed better results. So compound metrics can not only deliver more information but may reveal a different story to table stakes metrics.

The presenter gave a number of different ways to measure: for example, measure link between footfall and online. There are also offline attribution: using identifiers such as email to tie sales to online activity. Retailer also track using loyalty cards.

Omnichannel measurement is difficult, but can take simple steps in the right direction. As a B2B marketer it’s interesting to see the range of offline tools available to large advertisers: from cookies and mobile device IDs to tracking the ads seen on a Smart TV using audio recognition, the capabilities were impressive.

Realistically most tech companies don’t have the resources or opportunities to conduct true omnichannel metrics, but the concept of moving from table-stakes to compound metrics is something that we should all be thinking about.