Attribution models: don’t let imperfection stop you from attributing sales
In 18 years I haven’t seen anything or spoken with anyone who claims to have landed on THE model for digital attribution. That’s not wholly surprising in the world of business to business, with its unique sales journey. But that doesn’t mean B2B marketers shouldn’t try and develop a model to help them understand what various elements of the marketing mix really do work for them. In fact, quite the opposite: there’s never been a greater need for one.
Here are six areas every marketer should consider when evaluating channels:
Learning from consumer model
From a digital perspective, much focus for consumer marketers has been around tracking the offline purchase that has been influenced by online advertising, rather than the more straightforward digital purchase from a digital ad campaign.
The same should hold true for B2B marketers. There is a very easy way to evaluate performance in a single channel world, but we don’t live in a purely digital world. You need to understand the impact of online spend in an offline world, even without levels of footfall common to consumer marketers.
How can this be done? Make sure you are asking for (and tracking) feedback from the sales teams. Anecdotal evidence isn’t always something you should dismiss (for example, are pre-sales teams finding it easier to set up appointments?) Research is also your friend. Using market research (from independent parties) to evaluate brand uplift is key. But make sure it is independent – never ask the supplier to mark their own homework.
Search is part of the bigger picture
Search is an important tactic, but much of the impact of marketing happens before, in the run up to a prospect searching for your services or your name. That means the channels that help build up awareness of your brand, the problems that you aim to solve, the products you offer and your expertise and USPs, are vital.
All of that activity should lead to the user finding you via search, so investment is important, but when evaluating the impact of search remember that it’s very much the final touch point.
Drilling down into search spend
Evaluating any tactic shouldn’t be done at the very top level. Interrogating data further always provides the best possible opportunity of seeing what specifically is working. Search is a great example. If you have allocated that fixed amount towards search, understanding how that breaks down further helps you understand how other tactics support.
Any marketer would prefer more of their search spend to be allocated to very specific keywords (your brand name, the specific product or service you provide) rather than the longer tail of search terms (which may be relevant but also may not). Seeing an increase in these specific search terms is an indicator that some other marketing tactic is increasing awareness. So suddenly that investment is not just demonstrably working – it’s making the separate search budget work harder.
How martech can help
Martech kicks in when there is a some form of a relationship with the prospect or customer. At the very least, you need a visit to the website to start understanding user behaviour and engagement.
Adoption of any platform that helps you understand prospects better can be incredibly powerful, as long as you are aware of the limitations. Take into account those prospects you want to engage with but are as yet unknown. Basing an entire strategy around the known prospects and customers leaves you in danger of simply repeating patterns and not expanding them to incorporate the entire target market.
One of the most successful tactics any marketer can use is re-targeting (or re-marketing as we prefer to call it – a tweak in the approach used by consumer marketers to drive abandoned shopping carts). However, it can also lead you to over-emphasise how powerful it can be.
Seeing that re-marketing is driving fantastic repeat visits to your website and driving initial sales inquiries can be misleading – this isn’t an easy tactic to expand quickly. Simply driving significantly more traffic to the site won’t lead to a greater opportunity to re-market users with the same strong results. The quality must remain as high. Ensure sure you are still attracting the same type of (relevant) users.
Broader comms integration
PR has certainly not had the easiest of rides over the past decade. The challenge of adapting to digital, the explosion of social media and the rise of the influencer are just the start. And many of the old metrics for demonstrating success (reach and AEV, for example) don’t make sense in an effective attribution model.
However, few marketers would deny the importance of PR as part of a comms strategy. How do we get authoritative people to talk about what we do so potential customers know about us and trust us? Although those pieces rarely lead to an immediate site visit, much less an inquiry or sale, their effect is something you can monitor over time.
As well as tracking interactions with coverage online – how many social shares there are and links back to your channels; what’s the DA of the platform; how many engagements there are – you should also see YoY gains in relevant website traffic. Traffic that stays on the site and reads the type of content you want them to read; the kind of content that makes them more likely to become your customer.
While there might not be a single model which can tell you exactly how to allocate every marketing dollar, by tracking all activity over time, intelligently and not in isolation, patterns will emerge to help you refine activity. And if someone does develop the elusive attribution model in the meantime, I’d suggest getting a patent PDF. I’d quite like to be proven wrong.