Should you be investing in marketing measurement?
17 Aug 2023
The answer for most marketers is ‘probably’, and for anyone spending serious amounts of marketing money, then it’s a resounding ‘yes’.
But what is often not discussed is just how much you should budget for the measurement part, and exactly what you should expect to get back from it.
Magic formula
When looking at what you should be measuring, we found the formula here on the Harvard Business Review website. Although it is undoubtably correct, it also highlights the measurement challenge faced by most businesses who invest substantial amounts in marketing. Very few companies get near to measuring accurately ‘the incremental financial value gained as a result of marketing investment’.
So, let’s be bold and suggest some numbers for what you could spend on marketing measurement, assuming that you want to move beyond Google’s G4 and get something that is transparent and closer to ‘incremental financial value’.
A big driver of cost is the number and complexity of channels you use and the campaigns you send out. At the simple end of the spectrum are organisations purely using digital, with no call centre and all sales processed by an ecommerce system. Such ecosystems are relatively straightforward to evaluate.
Channel complexity
At the complex end are organisations using multiple indirect channels, such as TV combined with direct, and within direct, deploying both offline and online.
To measure just your direct channels, you can almost get away with only using customer journey-based attribution or MTA[1], although you will always be at risk of over claiming because you are ignoring both the baseline sales that would happen without any marketing, and the incremental effects on the brand of your marketing.
[1] Customer journey-based attribution or MTA works by evaluating the contribution of each individual step leading up to a sale, such an clicking through from an email or receiving a catalogue, and then attributing the value of the sale across the steps that led up to it. The value of a campaign is then derived from the value of each of the journey steps it created that preceded a sale.
To measure just your direct channels, you can almost get away with only using customer journey-based attribution or, although you will always be at risk of over claiming because you are ignoring both the baseline sales that would happen without any marketing, and the incremental effects on the brand of your marketingg the value of the sale across the steps that led up to it. The value of a campaign is then derived from the value of each of the journey steps it created that preceded a sale.
But when you are using indirect channels, with or without direct, you will need to use an approach like econometrics or MMM[1] to understand both the short and the longer-term effects of your marketing.
We find that there are great benefits in terms of granularity and accuracy to be obtained from using both MTA and MMM together. The findings from the MMM provide an umbrella view, and within that the MTA gets into the detail of how effective each direct campaign really is.
Marketing attribution
From our own research into typical industry charges for marketing attribution, both simple and complex, and for different levels of budget, we have compiled our own table. This is a very rough and ready guide, but we felt that some indication of likely costs was more helpful than none.
In our table the percentages relate to the overall marketing spend budget across all channels.
[1] Econometrics or MMM is a time-series based modelling approach that looks at all the possible influences impacting the volume of sales, and includes both marketing activities like TV or press, with external factors like seasonality, economic confidence, weather etc. It is good at estimating both the short-term effects of marketing and the longer term brand impacts.
Given the likely level of improved marketing ROI once accurate measurements are in place, these levels of investment should provide a very healthy ROI in themselves. However, for that to be realised a number of conditions need to be met:
- There needs to be cross functional support from the outset for the new measurement process being introduced, even though it may show different results from those currently being reported
- The measurement should be transparent in terms of how it is calculated, and independent in terms of who does it. Media owners such as Google or agencies wishing to protect spending budgets cannot be relied on to do this.
- The results need to be acted on, particularly across the budgeting process, as it is likely that spend will need to be moved from where it is now.
- It must be accepted that however much care is taken in the marketing measurement it is not an exact science, and particularly so when extrapolating from how spend is distributed now to how it might be in the future.
However, for anyone with a substantial marketing budget, the benefits from going down the route of marketing measurement should far outweigh the costs.
Please login to comment.
Comments