When I was young, my dad kept telling me that if I wasn’t getting better, then I was going backwards. This seemed to apply particularly to cricket, where my skills weren’t always as good as I wanted them to be, so I battled hard to try to show him some improvement.
Despite it being tough love, he was right. The kids I played cricket with were getting better, and if I wasn’t improving, it meant I was falling behind.
In B2B marketing, we hear a lot of discussion about RoI (or RoMI if you insist). The measurement of RoI should be based on how much your campaigns improve sales. If your marketing isn’t growing sales, then your marketing is failing. Unfortunately, this is not as simple as it seems, and often RoI figures are somewhat fanciful.
Attribution Metrics: Dominant, but Wrong
In marketing we attribute sales to marketing activities, giving credit because the campaign “touched” someone who ultimately became a customer. Just because an activity engaged a customer, it doesn’t mean that the campaign caused an additional sale to occur.
Attribution is particularly bad as it tends to disproportionally reward bottom-of-the-funnel tactics. The closer your marketing activity is to a sale, the more likely it will be seen as contributing to that sale. No wonder companies like Google love attribution: it makes the performance of ads on their platform seem better than they really are!
Attribution at Trade Shows
Perhaps the simplest example is a trade show. You might meet one of your current customers and have a great chat about their needs. This might lead to a new project, and big sales. Perfect! The sale is attributed to the show, and the RoI looks brilliant.
But did the show really make the difference? If you hadn’t exhibited, it’s entirely possible that the sales team working on that account would have had the same conversation at a different time. The same project might have happened, in which case we spent money on a trade show and made no impact on sales.
Don’t misunderstand me! I am not saying this is the case for every conversation at a show. Nor am I saying that trade shows don’t pay for themselves. But if we’re honest with ourselves, there are opportunities attributed to trade shows that we all know would have happened whether or not we paid for the booth. RoI for trade shows is inevitably inflated because of this problem.
Attribution in Search Engine Marketing
A simple example is search advertising. Let’s assume your company makes widgets, and they can be bought online from your ecommerce website called widgetshop.com (yes, I did think of buying the domain for this article, but the price of $12,500 seemed a little too much!). We all know that not everyone types the URL into the top bar of the browser: a lot of us are not sure of the exact URL so we’ll type “widget shop” in and let Google do the hard work.
If you want to maximise the RoI of your campaign, simply run Google ads around the term “widget shop”. Your ad will appear above any organic results, so people are likely to click on the ad, go to the online store and spend money. Google will neatly count up the sales and attribute them to your search engine advertising campaign. Big RoI numbers, here we come!
But did we really generate any more sales? All the people searching for “widget shop” wanted to go to your ecommerce platform. They almost certainly would have clicked on the organic result, so it’s unlikely we really generated any additional sales.
In fact, things are worse. No one has changed their behaviour, but we’ve had to pay for the clicks, rather than getting them organically. Attribution makes it look like we have an amazing RoI, but in reality, the RoI is negative.
Incremental Sales
Of course, what we should do is measure whether we really increase sales, a metric called incrementality. The idea is that simply running tests will let you see if a particular marketing tactic truly grows sales or not.
If you want to learn more about incrementality, you must follow Avinash Kaushik. The guy is a real genius. But the problem is that he works primarily in B2C, and B2B is different.
Incrementality is Really Hard in B2B
Unfortunately, incremental sales growth is very hard to measure for many B2B companies. There are a number of good reasons for this:
Long sales cycles: B2B is typically characterized by sales cycles of months or years. If it takes a couple of years for a customer journey to complete, something that is not at all unusual if you are selling components that are designed into systems, then you need to run tests over a long period of time. Are you really going to stop attending that trade show you think delivers great results for three years to test whether not being there really does have a negative impact on sales? I don’t think so!
High volume of marketing touches: a couple of years ago, Forrester found it took an average of 27 interactions before a typical B2B sale was made. Pathmonk thinks it’s around 50 for B2B manufacturing. Whatever the number, it’s difficult to isolate the impact of one touchpoint from all the others in the customer journey.
Complex decision-making units (DMUs): B2B purchasing involves groups of people working in buying committees or DMUs. It’s rare that one person makes the decision alone in a B2B sale, and even rarer that you have visibility of the interactions that happen in the DMUs you are trying to influence. So, your test might positively influence one member of the DMU but not do enough to sway the whole group. You’ve got to tweak multiple tactics to influence enough people on the DMU to change the decision.
Low volume of sales: some B2B companies have high sales volumes, but others quite low. Very high value B2B products might simply not generate enough opportunities and wins to deliver sufficient data to know whether your change of tactics had an impact.
Difficulty of setting up tests: it’s fairly easy to test, say some paid social content, but what about trade shows? If there is one big event for your products, how do you A/B test? You can stop going to the event, but the different market environment from year to year is likely to have a bigger impact than whether or not you exhibit, masking any useful information you might get from the test.
Noise: there is so much going on in B2B that the noise generated can swamp your tests. Your competitors might increase marketing activities. Legislation might change. The economic environment might improve or crash. A competitor might launch a better product, or you might establish a market lead. Being able to see the results of a test in so much noise can be impossible.
You’ve Got to Have a Dream…
You might be tempted to give up at this point. It’s definitely the case that few B2B marketing teams can accurately measure incremental revenue created by their campaigns. But you can at least try.
Try experiments. Use intermediate metrics that really relate to achieving sales (something like setting sales meetings rather than clicks!). You are going to have to accept that you’re not going to get perfect data, and measuring incremental sales accurately will often be impossible. But the more you move away from simplistic attribution towards incrementality, the more likely you are to make better decisions and help your organization grow.
