On Business Intelligence, Customer Relationship Management and Analytics ROI

Here, on the threshold of 18,000 visits to my blog, I'm writing a different kind of article. You may have guessed that already, with your first question being why I'm linking Business Intelligence, Customer Relationship Management and Analytics in one article. Well, the reason is really quite simple. It's because the common thread is data, and a primary objective of all three applications is to leverage the data asset to the benefit of the organisation, or alternatively, in an increasingly customer-centered paradigm, to the benefit of your customers.

In the bigger scheme of things, this article could also link to data warehousing, because without a quality data warehouse to begin with, it's really pretty difficult to achieve success in any of Business Intelligence, Customer Relationship Management or Analytics. However, for the purposes of simplicity, let's make the rather big assumption that a quality data warehouse exists.

Talking about poor ROI, note that according to Gartner, only between a third and a half of all Business Intelligence, Customer Relationship Management and Analytics projects are successful.

---

For anything to have ROI (return on investment), it suggests it needs to have a return which should generally be able to be measurable in financial terms. This of course does not mean all benefits are financial, but if we're talking ROI as accountants, then yes, they are. The nature of this return could include:

• an increase in customers - which means an increase sales, or in customer equity if you're assessing ROI from a discounted cash flow perspective
• an increase in brand awareness or reputation - which means an increase in brand equity and potentially conversion into future sales
• a decrease in costs
• a decrease in operating risk, which could be measured depending on the risk being mitigated, such as shrinkage. 

Now if the benefits of your Business Intelligence system are based on the premise of improved decision-making, then it is quite difficult to assess ROI, even though it may have really improved decision-making. That's because ROI assessments need specifics. What financial impact did that decision making result in? Increased sales? Reduced costs? Optimised Risk? By how much? Without concrete answers to the latter questions, you would be hard pressed to assess the ROI of improved access to information, either historically or into the future. Note that according to Gartner, less than 30% of Business Intelligence projects meet their objectives.

On the other hand, your Customer Relationship Management system may have a more direct relationship to returns, and these may thus be much easier to measure, especially if there is direct sales person or account executive interaction with your customers. However, not all sales may be the result of your CRM system, and it can become tricky to decide when a sale is a result of the system, and when it was the result of a different activity. Given these and other challenges, note again that according to Gartner, less than half of Customer Relationship Management projects meet their objectives.

Analytics is a little more tricky to determine ROI on, because unless the outcomes of your analytics programme, such as that fancy new predictive model of yours, find themselves integrated into your CRM system, the outcome of your analytics programme will be little more than a theoretical exercise. Once again, according to Gartner, 70%-80% of Analytics projects fail.

Why all the failures? 

My assessment (as a management consultant, and having served as CEO, EVP, board director, chairman, and even as a business school lecturer on Corporate Data Management) is firstly that there is no big picture strategy explicitly aligned with the corporate strategy that integrates all the different data initiatives, including the data warehouse, and secondly, that many of these projects are (often inadvertently) hijacked as being technology projects rather than being business projects that should be delivering a certain suite of results. Indeed, in all the write-ups of the various issues surrounding these disciplines, I'm sure the amount of text and effort devoted to the technical aspects of these applications way exceeds the text and effort devoted to the business aspects by a considerable margin. Note that the first reason falls in the domain of a failed Enterprise Architecture, specifically the non-technical component of it. As an aside, the ideal Enterprise Architect is not a technical specialist (leave that to the Technical Architect), but is rather someone as well versed in business as he/she is versed in technology. 

Returning to ROI, besides quantification, the stumbling block to truly professional ROI assessments is that there is often no direct relationship between the outcome of the Information System and the business beneficial outcome. In other words, as much as you may want to, it's not really feasible to claim that all current sales are a result of your new CRM system, since many other factors could be at play that resulted in that sale, such as brand equity or a previous encounter with your sales executive. Unless there's a direct trackable relationship between the cause of the sale and the sale, you may need to err on the side of caution more often than not, at least in the beginning.

In conclusion, leading on from the above paragraph on failure, a practical way to ensure that all of your future Information System projects yield an ROI is to begin with the end in mind, which is business/customer benefit. While it's easy to think that you need a Business Intelligence system to improve decision making, or that you need a Customer Relationship Management and Analytics environment to improve leads, what you're really saying is that you need these systems to improve the financial performance of your business, and that performance is a result of improved sales, lowered costs, and optimised risk, which are the actual outcomes of improved decision making and lead generation.