Analytics are Big Business: IDC estimates that the worldwide analytics revenue in 2017 exceeded $150B with about half of that amount having been generated in the US. Further, with an expected increase of over 12% annually, the revenue is predicted to reach $100B in the US alone.
Given this enormous amount, the question should be asked: Is Analytics worth it? Of course, Analytics service providers will answer: "Yes, Analytics will solve all your problems (and even give you World Peace for free)". A critical potential client will respond: "I would like to believe you but show me the money!". And yet another (quickly dissipating) faction will object: "Analytics is just a fad, I can do without it".
Hence, it is worth examining the return-on-investment (ROI) for Analytics projects, or to put it differently, is it worth to spend the money? As an example, let's examine a business case.
A Quick Service Restaurants (QSR) chain wants to find out if its advertising campaign is effective. A typical QSR generates an average profit of about $80k annually. The items with the largest profit margins are beverages, which have a 60% profit margin. Let's assume that 25% ($20K) of the total profit is generated by beverages. Case studies show that even simple optimization of an advertising campaign driven by data analytics can increase the sales of beverages by 8% ("In-Store Messages Optimized by Analytics - Case Studies and Results", PRI Journal of Retail Analytics, 2014). Given the figures above, the increase would be about $1600 annually per restaurant. Hence the QSR chain could spend up to $1600/restaurant on Analytics to break even. Assuming that the chain owns 100 restaurants, and has an Analytics budget of $100K, the ROI from analyzing and optimizing the ad campaign would be 60%! The economics would be even better for a larger chain.
Now, let's have a closer look at the overall economics of Analytics. The total cost of running an analytics project is composed of various components: There are operating costs for the computer equipment (on-prem or in the cloud) to consider, and there are costs associated with gathering and storing the required information and data. Last but not least, acquiring and using the necessary expertise either in-house by hiring data engineers, business analysts and other data analytics experts, or engaging outside consultants will account for a significant part of the total expenditures. To limit the financial risk and to keep the complexity low it is always wise to start simple. Hence an Analytics project could follow 4 distinct stages:
Reporting: Often, business analytics projects start with the question: What has happened? These backward-looking projects examine existing "historical" data and visualize them
in ways that are easy to understand for non-technical decision makers. Interactive dashboards and written reports are the tangible outputs of such projects. Often, generating such reports is still a tedious, sometimes manual matter. Investing in automation leads to faster (and often more accurate) reporting which in turn leads to better business outcomes faster. Nucleus Research estimates, that for each Dollar spent on business analytics automation, close to $2 are gained in business value.
Decision Making: Once frequent, accurate reporting is available the next step is to extract information that allows the business to act on. At this stage, business analysts with deep domain knowledge get involved. Usually, they focus purely on driving organizational efficiency in isolated areas by making changes limited in scope and reach. However, when the data and information contained therein is used to improve all decision-making the ROI is close to 400% (Nuclear Research). This is an important lesson: When every business decision is weighed against historical data, and the results are benchmarked against available data, impressive dividends from Analytics can be achieved. But the opportunities don't end here.
Strategically Aligned Analytics: Broadening the approach to use internal data for analytics to align business decisions across the entire organization and strive to forecast the effect of actions via modeling and prediction ("If we do this, what is likely to happen?") increases the ROI by a factor of 10.
Looking over the fence: Finally, the most expensive, but also most rewarding step is to extend the reach of Analytics beyond the limits of the own organization. Incorporating third party data, such as data from suppliers, organizations monitoring the purchase behavior of the customer base, and information gleaned from social media can increase the ROI to a staggering 1200%.
It is essential for each organization to honestly assess where on their analytics journey they are on. Attempting to skip the evolutionary steps will inevitably lead to wasted resources, higher costs, lower returns and a lot of frustration. At each stage of the Analytics journey, the success to achieve business goals (or lack thereof) due to the deployment of Analytics itself should be analyzed. That way, misguided approaches, dead-ends, and other inefficiencies can be eradicated quickly. Tackled in that prudent and targeted way, Analytics projects evolve from a being an expense to an operating cost that is judged against its financial return like all other business costs: Instead of being a drag, with a positive ROI Analytics drives the business forward.