Back to Blog

Measuring ROI in marketing

measurement roi Apr 07, 2021

By Andrea Tartaglia – Co-Founder/Director @ The Marketing Leaders Ltd.  

 

Measuring the impact of marketing investments remains the subject of much debate. It is a complex topic and there is no simple answer to the question: ‘how do you measure the effectiveness of marketing activities?’ 

 ROI (Return On Investment) is a good place to start.

The thinking beyond it is very straightforward and for that reason it is a metric I like a lot.

The basic principle is that an activity is deemed successful if it produces revenues in excess of its cost.

In even simple terms: the goal is to make more than you spend.

Simple. Beautiful!

ROI is useful in the context of marketing for a number of reasons, the main ones being the following: 

  • It can be used to justify your marketing spend: knowing that the planned marketing activity produces a positive outcome in terms of sales can help marketers to gain support from senior management and other departments (sales for example)
  • It is a useful metric to decide where to allocate marketing funds: in a world of limited resources, the funds will be assigned to activities projected to produce sales in excess of their cost
  • It can be used it to prioritize your marketing activities: building from the previous point and assuming there is not enough budget to run all the activities that have a positive impact on revenues, precedence must be given to the activities with the highest estimated ROI
  • It is a measure of the effectiveness (and efficiency) of the marketing spend: after the activities have run their course and produced results, calculating the ROI determines if they actually succeeded in reaching the expected sales goals

Personally, I get a bit defensive when I read that ROI can be used to justify marketing spend. I prefer looking at this from a different point of view, which has a subtle difference but one that is very important to me. 

There is no need to ‘justify’ marketing activities. Of course, they are useful and are meant to produce a positive outcome. At the same time, there is a need for marketers to feel (and be) accountable for their work. In this sense looking at ROI ensures that marketers keep a commercial frame of mind, caring for being creative, innovative, creating brand impact and also for affecting the bottom line.    

It makes marketing more tangible in a commercial context and confirms the marketing’s role in driving growth (which has always been the goal of marketing despite ‘growth marketing’ becoming a buzz word only relatively recently) – rant over.

The ROI formula

Having said all this, how do you calculate the ROI of a marketing activity? You apply the ROI formula:

ROI = (sales generated by the marketing activity – cost of the marketing activity) / cost of the marketing activity

Any activity with a ROI > 1 has a positive outcome on the P&L.

Any activity with a ROI < 1 is not effective as it costs more than what it produces.

 

As I mentioned, it sounds simple however it is deceptively simple. The reality is much more complex and calculating the ROI of marketing is not as straightforward as I made it sound so far.

The main reason is that there are many factors contributing to generating sales. Cause and effect are difficult to correlate in most commercial realities. The interdependencies are too many to consider, and marketing is never an isolated input producing a well-defined output.

In this context, the ‘simple’ ROI formula stands only with a good dose of assumptions, which may or may not undermine its validity.

Example: Marvel’s Avengers

I tried to unpick all this once with the help of my agency. We looked at all available data across several months and used econometric models to understand, measure and isolate causality around the growth of a specific brand (in my case Marvel’s Avengers).

 

 The results were very illuminating. The exercise highlighted a number of surprising discoveries but also brought to the forefront a number of limitations.

On the positive side, the modelling showed us the activities that had a more direct impact of sales growth and we were surprised (to a point) to find out that the strongest correlations were not the ones we originally imagined.

We thought that the activity with the strongest positive correlation to sales would be the successful roll-out of movies in cinemas; while in reality the most direct positive impact came from the in-store execution of merchandising programs (which was in turn only indirectly affected by the release of movies into cinemas, regardless of their performance – within a certain range).

With this new information at hand, we could revisit our strategies to focus on the marketing activities with the most direct impact on sales.

However, the exercise took quite a long time to be completed and only gave us a snapshot of a particular timeframe, offering limited certainty on how repeatable the correlations observed were.

As useful as it was, it was not something we could replicate for all brands or for the same brand over and over. The effort did not justify the result. Or in other words (ironically) the ROI of the exercise was questionable.

Which is a nice sideway to another reason things are not as simple as they seem.

Is any positive ROI a good ROI?

It depends…

It depends on the commercial context in which the ROI is measured and evaluated. It is important to clarify before starting any activity what the level of ROI is considered acceptable and measure/evaluate activities accordingly.

Very rarely simply a positive ROI is going to be deemed enough. Most likely there will be a minimum benchmark or threshold to reach - say a ROI of 5 or 10, which means every dollar spent must generate at least 5 or 10 dollars in return.

Example: ROI in the licensing business

This was particularly important in my years managing a consumer product licensing business, considering the complexity and peculiarities of the business model.

 

Here is why: in the licensing business, the licensor grants the rights for a property to a licensee in exchange of a licensing fee, normally a royalty or % off the retail sales of the products featuring the property sold by the licensee.

I worked for a licensor with a proactive approach to marketing, willing to invest in marketing activities that could generate a higher demand for the licensed products.

This approach presented an interesting challenge when calculating the ROI of the marketing activities carried out by the licensor as the licensor is not getting the full return on the incremental sales but only a portion (the royalty %) which pushed us to establish a very high ROI so that only activities generating a very significant uplift in sales were considered.

And more complexity…

Without discouraging anyone from using ROI as a good metric for assessing the effectiveness of marketing investments, there are a few other elements of complexity worth bringing up briefly:

  • As mentioned, marketing does not work in isolation and there are significant interdependencies with the work of other departments and teams. This needs to be recognized when looking at the ROI of marketing investments. And as always, a collaborative approach helps with a common understanding on what effect the marketing initiatives really have
  • There are other metrics that can be used, and it is important not to create confusion between these and the ROI; however, all metrics contribute to assess the success of the marketing investments. It is very likely any marketing activity is associated to some KPIs that must be met. The difference between KPIs and ROI (which could be considered the ultimate KPI) can be the subject of another post…
  • The timeframe considered when calculating the ROI is also a critical factor. Marketing activities tend to have long lasting effects and their impact might not be completely know at the time when the ROI is being calculated. In a way. You might assume the ROI ends up being a conservative estimate of the true impact of marketing investments
  • Also linked to timing, there is the complexity related to the fact that the investments and their results might be accounted for in different fiscal years, creating a disconnect between the ongoing running of marketing activities and how they fit in the context of budgeting and managing the yearly P&L (putting a strain in the relationship between marketing and finance)

And here is the big one!

Finally, the last element of complexity I want to mention is the fact that not all marketing activities are meant to have a direct impact on driving revenues, which makes assessing their effectiveness through ROI a difficult, if not pointless, exercise.

This is really where the KPIs I mentioned above come into place and are better indicators of the success (or lack of) of such marketing initiatives.

Let me be very clear of this point: I believe all marketing must have a positive impact on sales (or revenues, or profitability) but sometimes that impact is not direct and is dependent on other activities more immediately connected to generating revenues.

Here are a couple of examples to clarify this point:

  • You can assume brand building and ‘top of funnel’ activities are important marketing actions with a long term, somewhat indirect effect on sales
  • Conversely, trade marketing (think POS and/or off shelf positioning and communication) can be easily related to uplifts in sales

 

In conclusion

I mentioned it a few times, but it is worth repeating: ROI is a simple and effective metric to use to assess the effectiveness of marketing investments. 

However, behind its simplicity, there are ‘hidden’ complexities that require looking at other metrics too to fully understand the impact (hopefully positive), projected or realized, of marketing investments.

As mentioned above, complexity comes from:

  • the interdependency of marketing activities from all other activities of the Brand/company
  • the timing of the marketing activity being assessed, and timeframe used for the assessment (especially when activity and results cross over multiple fiscal years)
  • the purpose of the marketing activities, which might not be to generate sales growth (at least directly)

My general approach is to be ‘data informed’ (as opposed to ‘data driven’) and in this context, ROI gives me enough to go by. I can always complement it with other metrics if and when I need as deeper understanding. 

In the end it is very rare to be in the position I was with my last big Star Wars marketing push for consumer products, when I could run a market test before scaling up the campaign and demonstrate that the campaign was delivering a sales uplift between 30 and 60%. Not bad!

 

More articles about ROI and Marketing Measurements are available at the Knowledge House, which is reserved to The Marketing Leaders' FULL Members.