In part one of this blog series, we discuss some of the challenges to attaining useful measures of Marketing ROI (MROI), and introduce some concepts to overcome the traditional challenges to ROI measurement. These concepts will be expended upon in subsequent blogs.
To measure ROI, you must have a consistent and realistic measure of return, and the only consistent and real measurements of return are financial targets. Therefore, you must begin by stating your desired financial targets.
You should also be able to state up front what your desired ROI is. Too frequently, marketers carry out activities and then wonder what those activities were worth. This is backwards—marketers should define their target ROI at the outset.
If you know your financial targets and your ROI targets, it is simple to calculate your target marketing budget. Campaigns should then be devised to meet the financial target with the appropriately-sized budget.
This is not the typical approach to calculating Marketing ROI (MROI), to the detriment of the function. Marketers tend to measure what they can rather than what they should. Therefore, they tend to report on second-order metrics (impressions, views, likes, downloads, attendees) that are difficult to square with financial value. They also tend to measure at the wrong granularity.
Ultimately what matters in marketing is whether the marketing efforts are generating financial value to the company. Any metric that is decoupled from financial value has an undesirable level of indirection. This is not to say that all metrics must be financial, but they must all be demonstrably connected to a measurable financial outcome.
The more ROI can be grounded in financial targets, the better the team can communicate the value of marketing within and outside the team, the more credibility it will have in justifying its budget requests and explaining its results, and the easier it is to communicate the impacts of budget changes. This is achievable, but it may require looking at the deceptively simple ROI calculation in a new way.
The high-level challenge of measuring MROI
Marketing ROI (MROI) is known to be difficult to measure. The concept is simple. Indeed, the formula is just a twist on a standard ROI calculation:
MROI = (Business value from Marketing - Marketing Investment)/Marketing Investment
The challenge that most marketers encounter is in consistently defining and capturing the data needed to fill in the variables in the equation. Let’s start with the easier of the two variables: Marketing Investment. What should that include? Let’s imagine a digital campaign— should you include the media buy? It seems pretty clear you should. What about the creative agency you used to make the content for that specific campaign? Yes. How about a prorated share of the content writer you have on retainer? She contributed to it, after all, but how much time was it exactly, and what were her rates again? What about a prorated share of the company FTEs? Corporate overhead? Meals for the meetings you had with the agency? A share of the SEO optimization platform costs? It’s tough to know precisely where to draw the line, but you can see that each of these costs changes the denominator and that’s going to change the ROI.
The other key variable in the numerator is Business Value from Marketing. Revenue seems like a clear indicator of business value. But is that one-time revenue or customer lifetime value (LTV) that you should care about? How about gross margin? Bookings value is good too, but you’ll need to map that to revenue formulaically. Pipeline? Opportunities? Impressions? Brand perception? Each step further up the marketing funnel, the more intangible a discrete measurement of marketing value becomes. Yet, we all know we have to carry out these activities high up the marketing funnel in order to influence the revenue that is realized at the end. Marketers want to provide data to quantify their impact, but they have to acknowledge that there’s a large sensitivity in their results that will yield a broad range of potential outcomes—and that data doesn’t hold up well under scrutiny from peers outside the marketing team.
Add to this core challenge the further complexities introduced by trying to measure traditional and digital channels consistently, tracing multi-channel campaigns, and attempting to compare your performance to industry peers, and the challenges just multiply.
Historical context of Marketing ROI
Here’s a quote from a recent article about shifts in marketing investment, published on the website Search Engine Land (our bold):
“While marketers should ideally address the full funnel, most no longer have the budget. Accordingly, they’re emphasizing performance campaigns because those are easier to track. Any spending that can’t be justified in terms of clear ROI is being cut in many places.”
This neatly encapsulates a major problem in the marketing function. Investing in campaigns because they can be measured is the wrong order of operations. We should invest in campaigns that meet the needs of the business and are aligned with our goals.
Marketers often refer to the John Wanamaker axiom, “Half the money I spend on advertising is wasted; I just don’t know which half.” We’ve heard this dropped into meetings in response to a question about the ROI of marketing investments. And it often works— everyone has a chuckle and the conversation moves on. Wanamaker’s comment is a good characterization of one challenge of measuring MROI, but it isn’t a reason to not try. Also, Wanmaker died in 1922. We can’t use it as an excuse forever, and we can’t just shrug our shoulders and fly blind.
On the other hand, the most dangerously misinterpreted quote in marketing (sometimes attributed to Drucker, sometimes to Deming, sometimes to Ridgeway - we settled on Deming) is, “What gets measured gets managed.” The reason it’s dangerous is that this is only half of his quote, and consequently it is misinterpreted as being an instruction. People think we should manage by what we can measure, and by implication, not manage by that which is difficult to measure. To make that point more explicitly, there is significant business risk in excluding from decision-making those criteria which are difficult to quantify with precision. The full text of Deming’s quote, sadly not usually provided in its entirety, makes a completely different point:
“What gets measured gets managed - even when it’s pointless to measure and manage it, and even if it harms the purpose of the organization to do so.”
Here’s an example of that. Earlier in Dan’s career, he delivered technology that could automate a large number of customer service calls quickly, meaning they didn’t need to be handled by a costly customer service agent. Many of his customers, contact center leaders, would insist the only metric they cared about was reducing Average Call Handle Time (ACHT), the average duration of a phone call into the contact center. He would discuss this with them and suggest there must be other things that mattered, such as successful call outcomes, customer experience, overall cost savings and so on. But some customers insisted it was only about ACHT. When this happened, he would ask why they didn’t answer all their calls and then immediately hang up on their customers without talking to them. They’d achieve fantastic ACHT that way. The reason they didn’t do that was because they knew that would be a terrible business decision. The problem is, they were being managed by what could easily be measured. What they really wanted to do was to reduce ACHT while maintaining those other benefits like first-call resolution or customer satisfaction that they found much more difficult to measure reliably.
This is therefore a cautionary tale: beware of organizing around potentially irrelevant metrics. As Igor Ansoff noted, “Managers start off trying to manage what they want, and finish up wanting what they can measure.”
The marketing function contains a fantastic example of this: we started off wanting to measure and tune our marketing investments and mix, and we finished up with the pseudo-science of marketing attribution models.
Are these our only options then? Wanamaker’s world, in which we fly blind and accept the inherent inefficiency of our activities, or Deming’s, in which we labor under the false precision of managing merely by what we can measure while under-valuing things we know to be critical, but which are difficult to measure? No. We can do much better. Our path toward practical, valuable MROI measurement is based on four key elements:
Focus on big-picture, first-order ROI calculation
Carefully select meaningful metrics
Adopt a new, consistent approach to MROI measurement across your whole marketing plan
Baseline and benchmark your performance
In our next blog, we will delve into these in more depth, outlining our vision for consistent, reproducible MROI measurement.
Note: some of the quotes for this article were culled from an excellent article on what to measure in business, which can be found here
Dan Faulkner is the CTO of Plannuh. Dan has degrees in speech and language processing and marketing. He got his marketing degree after running research for text-to-speech synthesis research at SpeechWorks (now Nuance) and must have been looking for something easy to do. You can follow him on Twitter and LinkedIn.