What’s the ROI of Analytics? (Part 1)

It may seem like a strange question but it is one that still tends to come up. Last week on a discussion thread, for example, someone was asking about how to demonstrate the return on investment (ROI) in analytics. That’s a nice irony, I thought, considering analytics spends most of its time measuring the ROI of marketing investment. Isn’t that ROI enough in itself? Apparently not, if the question is still being asked. So how do you measure the ROI of analytics?

As with all ROI calculations, we need to understand what the investment levels are and then what the returns are. In the case of analytics the investments are probably easier to work out, but it’s necessary to calculate the “total cost of ownership” of the analytics capability of the business. These typically break down into three categories:

  • Technology
  • Services
  • People

Technology costs are fairly self-evident and well understood. Most digital analytics technologies operate on some kind of subscription model, but other pricing models include licence costs with maintenance or renewal fees in subsequent years. The key aspect for understanding ROI is not just to look at the annual or first-year costs but to look at them over a period of time, perhaps the expected average over the next three years.

Most analytical technologies require some level of services around them, either from the technology vendor or from other service providers such as consultancies. The U.S. and the European markets have all seen recent consolidation of consulting businesses into larger entities, perhaps with the associated impact on prices, but there remain sufficient smaller, niche, or boutique players to provide a competitive market.

I think it’s important to allocate sufficient investment in services/consulting around an analytics technology, particularly in the early days of its adoption. They are not necessarily the simplest of technologies and there may be a learning curve involved. It’s false economy I believe to make the (often significant) investment in a technology and then not to invest in having it configured properly for your business or having your people skilled up on it properly. I’ve seen this a number of times over the years and there’s no hope of realising a decent ROI from the technology if companies under-invest at this critical early stage. The same principle also applies to the next category of cost: people.

In my opinion, people are the most important investment that an organization can make in developing its analytics capability. There can be trade-offs between developing an in-house capability versus out-sourcing to an alternative provider, but at the end of the day it will be the quality of the people that will count. There is also a balance to be made between acquiring technologies and acquiring skills. Smart people can do a lot with relatively simple technologies or open-source systems but you need the smart people if you are buying in smart tools. Otherwise it’s like buying a high-performance car and giving it to someone who’s just passed their driving test. It’s only going to be a matter of time before there’s an accident.

So the investments in analytics can be assessed, but what about the returns? And importantly for any specific analytics investment, what is the marginal return for the marginal cost? I think assessing the returns can be quite a tough job and broadly they break down into two types: the direct benefits or returns and the indirect ones. I’ll be looking into these in more detail next time.

Image via Shutterstock.

Related reading

Vector illustration with a magnifying glass focusing on a pie chart, a graph line trending upwards, and other metrics symbols.
Checkboxes on smartphone screen. Hand hold smartphone, finger touch screen. Checkboxes and checkmark. Modern concept for web banners, web sites, infographics. Creative flat design vector illustration
Screenshot shows a Google search for outdoor grills, the shopping ads shows images with “in store” showing the product is available nearby.
How numbers affect conversions