Business Intelligence

TCO: Total Cost of Ownership

There is a business equation that states: Value = benefit / cost.

As at the time of this post creation, the world is reeling from a financial crisis brought about mostly by mismanagement or not considering that at some stage, the piper needs to be paid. For history, here is the full story. Basically, the Pied Piper was hired by the mayor of Hamlin in Germany in 1284, to rid the town of a rat infestation. He did this by very unconventional means, playing a pipe that lured the rats to follow him, where he led them to drown in the sea. When returning for payment, the town refused. So he came back a few days later and while the townsfolk were in church, he played his pipe again, to a different tune, this time luring all the children of the town / village (some 130 of them) who were never seen again.

What does this story have to do with BI? Great question, I am glad you asked.

Beware Free BI software.
There is a certain global massive organization who needs no introduction, that consistently kills every market it moves into and it does this with the lure of free software and alleged standards. If you purchase a database software from this organization, a reporting and business intelligence framework comes free. What this organization is not telling you is that in order to get the software to do what you want, it will take 4 times longer to get things implemented. Also beware the hidden costs of licensing (where there is just a little something in your environment that violates the terms of the free license, which is when you are locked in and have spent significantly and the cost of undoing and starting from scratch is unthinkable). Let's keep in mind that on average, the cost of implementation is about 3 - 5 times the cost of software, at least this is the ratio for ERP / CRM software in general. for BI, there are other considerations, like the data integration (ETL / ELT) will consume at least 50% of the project time, especially for the pilot BI implementation. It is very difficult to provide value based on license fee pricing alone, so when the audience has their eye only on the cost of licensing, the value proposition becomes all the more important. The value has to take the benefit and the relevance of the benefit provided, into account.

Warning, shameless plug about to follow:
One of our customers have a BI systems manager who came from a local government where this previously mentioned global organization's solution was implemented, with 4 people over the best part of 2 years. So to implement a free software solution, it cost 8 man (person) years, at an average of $120K / person-year, just shy of $1 million. Not bad for a free software. There was nothing 'micro' about the cost and most certainly nothing 'soft' about the impact to the bottom line. Also, the sponsors of the project had to live with many compromises.
Disclaimer: any inferences you may have deduced to any named organization is purely by your own association

When our customer tried our solution (we implemented what they needed which did a whole lot more (and faster), and was implemented over a period of 6 weeks with a single person), the reality of the free software's economy hit home. Over a 3 year period, assuming nothing else was done with the project, our solution came out at less than 25% of the cost they would have paid to implement the free software. Also, the rental / subscription nature of our solution gave them freedom to quit at any time with flexible options. The latest comments we have had back from the business users is "don't anyone DARE try to take this information delivery capability away from us!", which is always reassuring to us not so much in the secure revenue, but rather by our key success metric, which is to deliver value to our customers and their users.

Another thing that still baffles me is the tendency to focus on hourly rates instead of value. I once dealt with an organization in Melbourne that employed consultants from a particular organization, because they got them for a certain daily rate (let's assume $800 per day). This team of 6 had been busy trying to solve a performance problem on a large CRM implementation. A performance expert was eventually called in and had the problem solved in under an hour, but the billing rate was A$300 / hour. Management was furious at the ridiculously high hourly rate and immediately demanded a lower billing rate. The response was: "So it is okay to pay $800 X 6 people X 5 days (A$24,000) and not get a result, but it objectionable to pay A$300 to see a problem solved?"

My question here is: Which approach delivers value?

Let's put it another way: assuming you needed heart surgery, would you opt to have a surgeon operating on you that will charge $10,000 for his expertise and experience or an almost-qualified surgeon who charges $50 (because he does 20 surgeries a day and can thus cover costs more efficiently)? The analogy might seem extreme, but you get the principle.

As we move into BI projects, we need to be aware that we are NOT dealing with another ERP / CRM type application. BI needs specialists (or of course you could hire the specialists when you have to redo the project) and getting experienced high-calibre people will yield a better overall value when all is said and done.

Remember: Value = benefit / cost

Cost should never be measured only in terms of license fees. The License fee is usually a small fraction of the total implementation and ongoing cost anyway, so focusing on license cost alone is like focusing on getting cheap running shoes to the detriment of your feet, spine, knees and then embarking on running a marathon.