Purchasing BI-related software reminds me of purchasing a new car. Do I buy or lease? Zero percent financing or $2,500 cash back? What about an extended warranty? There are direct parallels to these questions within most BI software purchasing decisions. Do I purchase an enterprise license? Per server or CPU? What about named, concurrent or device-based user licenses? Once you've settled on the licenses you need; you have to wade through the support agreement alternatives. The myriad of feature and financing options are just too great. I don't always buy that this diversity of choice is giving the customer the flexibility that they demand, but rather it seems a clandestine attempt to confuse preventing prospective customers from calculating the software's true cost or performing a direct apples to apples comparison with their competition.

Many companies make a concerted effort in selecting their appropriate BI toolset for their organization. However, once they have selected a specific vendor tool, they spend comparatively little time determining their appropriate licensing approach. Hidden just beneath the surface of the seemingly obvious licensing alternative are some key factors that have the potential to significantly reduce the overall software cost and increase your flexibility and bargaining potential over the lifespan of the software. Having been through these negotiations more times than I care to remember, I've developed a few lessons learned that I believe are worth sharing.

Remember, Most Companies Overbuy

The first issue that needs to be addressed is just how many user licenses you need. Correctly calculating how many licenses you need and, just as importantly, when you will need them is the key requirement that will drive whether an enterprise, per server/CPU or user license scheme is appropriate for your unique situation. Unfortunately, overestimating the rate of user adoption and the ultimate penetration of the BI tool is the classic mistake that a large majority of companies make.

One client I worked with a few years ago originally purchased a server-based license for a reporting tool and associated BI portal for its planned 200-user deployment. A year after the solution was initially deployed; the software remained in the hands of only about 20 power users. Six months later it had been deployed to all regional sales offices, and almost two and half years after the initial implementation it maxed out its penetration at approximately 140 end users with the other 60 or so planned users actively resisting the adoption of technology and relying instead on their assistants to produce the paper reports they need. While the project itself is considered a rousing success, in hindsight, the original user adoption plan was overly optimistic. Had this company accurately forecasted their adoption rate accurately, it would have reduced the overall licenses needed and perhaps prompted an incremental purchasing strategy versus a large bulk purchase before the licenses were actually needed.

An Incremental Purchase Strategy

As this company found out, it is actually much harder to scale down that to scale up. Scaling up, on the other hand, can be an effective way of managing the required software investment. While challenging from an annual budgeting perspective, the future incremental costs of additional user licenses are quite often more cost-effective over time than the present value of a larger, enterprise or per server, bulk purchase. An enterprise or per server license may initially seem cheaper but may often prove to the contrary if the end user adoption rate doesn't play out as anticipated.

The first step in comparing the two is to find the maximum user count at which a per user license is no longer the most cost-effective, create a realistic estimate for your ultimate user base and finally assess how long it will take for you to get there. Remember, if your rollout is longer than a year not only are you paying for software you aren't using but more than likely also paying for a support and maintenance plan you don't need. With the average annual support plan costing 20-30 percent of the additional purchase price, you're also paying a hefty interest rate on an unused asset to boot.

Your vendor may try to convince you that an enterprise or per server license is the most cost-effective and try and steer you away from any incremental purchasing strategy. They are doing this for two reasons. One, like any sale, more is better and two, the bigger the purchase, the larger and longer the commitment. Contrary to what the vendor may say, those incredible savings deals that are available only now due to an end of the month goal that the sales rep must hit or the rebate that expires in two weeks will magically reappear in one form or another the next time you call looking to purchase more.

Question the Need for the Support and Maintenance Agreement

Most companies consider purchasing a support plan an absolute requirement for all their software purchases. While I would recommend purchasing a support plan for the first year when implementing a new piece of software, I would caution against merely rubber stamping that same agreement year after year. With a support and maintenance plan costing between 20-30 percent of the initial purchase price, it is only a few years before these costs outweigh the initial investment. The key benefits of any support and maintenance agreement are access to technical support and software upgrades. Is your organization gaining the value that you thought you would out of this? Is the premium technical support offered through your agreement of significantly higher value than that you can find through many of the user communities on the Web?

Assuming your current software version is relatively stable, and therein lies that risk that you need to manage, not having access to software upgrades can also prevent you from falling into the repeating upgrade trap. When calculating the total cost of a software investment, what is rarely taken into account is the cost of implementing future software upgrades. Major releases of the software may add additional features and functionality, but the cost of upgrading from one major version to another can be 50-100 percent of the original implementation effort. BI tools are like many other pieces of software; the average user takes advantage of only 20 percent of its functionality. Do you really need that functionality that's being promoted in the latest version?

One other aspect of cost that is somewhat unique in the BI software space is that software purchase prices are not guaranteed to rise with inflation. The BI software market is continuing to expand with new products constantly being developed. You may find that your vendor of choice is now promoting its new analytic application while the reporting tool you bought two or three years ago has actually dropped in price since then. A client I've worked for recently ceased their annual support and maintenance with one of the top two BI vendors three years ago. Their support and maintenance agreement was at the time costing them $50,000 per year. They are now looking to move to the latest version of the software for an initial purchase price of $70,000. While they will need to "invest" in the software again, their $70,000 cost is outweighed by the $150,000 savings they've gained over the past three years.

ASPs and Upcoming Alternative

One of the interesting and upcoming alternatives that bears monitoring is the rise of application service providers (ASPs) in the BI space. ASPs have been slow to penetrate the BI world due primarily to the custom nature of data warehouse development. However, the introduction of analytic applications is slowly changing that. Even more interesting within the general population of ASPs is a segment that I'll term analytic service providers. These ASPs are not only offering to host and manage your BI application, but they will add their domain expertise in the analysis of your data. This domain expertise can take many forms from data cleansing, data augmentation with third-party demographic information, leveraging their industry analysis expertise, all the way to pooling your data with their other customers' data for data mining purposes. Analytic service providers have started to appear in the healthcare and insurance industries in clinical analysis, fraud detection and healthcare usage analysis and will undoubtedly spread to other verticals. For more on analytic service providers please check out next month's column.

Caveat Emptor

A sales rep, who has worked for the two leading BI tool vendors, once told me that, "Software is not meant to be used; it's meant to be sold." I'm sorry but I just don't buy that attitude. Unfortunately, many of his customers do and now have some very nice shelfware to prove it. I don't necessarily blame the salesperson though. As the old adage goes, caveat emptor or buyer beware. As a potential customer, it is up to you to understand what is being sold to you and how it is being packaged. With a little additional diligence you can potentially save your company a significant amount and increase the overall value of the solution you're implementing.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access