Subscribe to DSC Newsletter

6 Often Forgotten ROI Factors In Data Analytics

Big data. Small data. Doesn’t matter. When it comes to ROI in working with data there is really just one question:  “What value do we glean from analyzing our data compared to our costs in doing so?”

The fact remains that, while we hear of some glamorous wins, and also of some less-than-stellar results, we still have lots of questions outstanding in the market push to data analytics success. There is no one proven and replicable success formula. At least, not yet.

Call me old-fashioned, but I calculate ROI with real, tangible numbers derived from real projects. 

Specifically, what did (or will) my data analysis tool or BI system or database infrastructure solution cost me?  How much cash to buy it?  How much time tying up my internal resources while implementing and learning it?  How much more out-of-pocket for customization and on-going support?

Now, weigh these against the true value received -- in two categories.  Either or both of:  How much directly connected incremental revenue did this initiative achieve?  How much money did I save as a result?

As in many things, whereas the questions may be straightforward, the answers may not be so easy to calculate.  However, I suspect that there is a correlation between the ease of obtaining tangible answers and the achievement of a positive ROI.

Perhaps the most difficult part of ROI is taking into consideration all of the appropriate factors in the calculation. Here are 6 ROI factors that people sometimes forget or under-value:

  1. Establish proper baselines for comparison. How much time does it currently take to do analysis, reporting or other data-related task(s) using your existing data manipulation tools?  How much of that time is backlog or waiting due to resource-related bottlenecks in your area or elsewhere? How much of that is data prep to enable analysis?  How much is rework and error correction?  For sales or marketing initiatives, what are the baselines for revenue or other sales activities today, before the new data analysis-driven campaigns?  Without clear before and after numbers, how will you know if something is successful? 
  2. Your people are your most expensive resource. Don’t forget that. Because their salaries are already budgeted it’s surprising how often this is ignored. If they can be twice as productive, then that’s twice their value. The time you can save them from doing mundane and error-prone manual tasks (like data-prep and reporting) is a huge consideration. For the purposes of ROI, time saved translates into money saved.  
  3. Intangibles are exactly that. Do not rely on hype, assumptions, ‘a hunch’, or use any factor that is not objectively calculable for your ROI. If you can’t measure it and compare the number to your baseline, then it belongs in the comments and not the calculations. 
  4. Plan for the ‘F’ factor. No matter how well we research and plan, things go awry. Simply speaking, we’re talking about scope-creep and other unexpected factors that arise in the time between a decision to move forward, the completion of implementation, and then overall by adoption by users. An ROI proposal that is too tight will often not deliver because of factors unknown at the time of its creation. The ROI plan needs to be rock-solid, yet have a contingency plan as well.
  5. Build a true ROI proof of concept. Your situation is unique, so test your theories about how data analysis can deliver ROI thoroughly before you implement a large or a single solution. Apply your concepts and potential solutions to multiple and different types of small tasks as test initiatives. And be aware that different tasks may require different solutions. Start small and build from success.
  6. Calculate and plan for the full and true costs of your initiative. In addition to the initial cost of the solution, human resource costs (especially time considerations and their rippled-out impacts), specific customizations, and long-term maintenance and enhancements are the factors usually under-calculated.

Of course, this is not the exhaustive ROI factor list, just some of those that I tend to see under-represented. Please add your additions to this list as well – especially the items usually forgotten about. Enhancing the value of data analytics is an important goal for all of us!

David Lefkowich is the VP Sales and Marketing for FreeSight Software, a data integration, cleaning, analysis and reporting tool. (

Views: 3421

Tags: Analytics, Big, Data, Preparation, ROI


You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Thomas Lincoln on March 10, 2016 at 4:20am

LOL!  as a long term finance modeler for m&a, investment projects, etc, and understanding the strengths and weaknesses of the models and the behavioral finance theory, the ROI arguments read well, but aren't applicable in real life.  Its only for people to feel good about making and defending a decision to people who aren't strategically focused, but control focused.  an easy gotcha, Steve Jobs never did an ROI on any product he developed, why should you?  second, if the investment is required to stay competitive, to keep up with competition, doesn't matter what the ROI is, either you invest to stay competitive or you don't, and if you don't you suffer the consequences in the future.  third, the future is always uncertain, and sometimes more uncertain than others, so how to quantify benefits and resulting displacements or substitutions are difficult and are educated guesses at best.  I read one ROI software justification which quantified eliminating steps to the filing cabinet as a time saving.  Were there cost savings? not for more than 5 years.  There was another actual BI implementation start which failed at a cost of $2M and no deliverable resulted, using external professional development resources, and the in house validation server costing $35K and one person's time (mine) ended up being the mission critical application left standing. 

Ultimately, the technology revolution, like the industrial revolution, is creating enormous productivity, but the social consequences of a growing population with less jobs available and more free time, means that there is less money available for workers and want to work unemployed, and has created an excess labor capacity.  Guess what happens with an excess capacity of any commodity?


  • Add Videos
  • View All

© 2020   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service