And when you’re negotiating pricing with BI software vendors, are there any things to know as your negotiating pricing or looking at licensing structures? Are there any changes or trends in the way that BI software is sold?
There are definitely a lot of changes in the pricing structure. All of the users of business intelligence and actually all enterprise applications and enterprise software are demanding simplification. Traditional pricing structures and options have been extremely complex, starting from multiple SKUs that one had to choose from and options that one had to go through.
So there is definitely a trend for enterprise software vendors to consolidate multiple SKUs into at best one or two product offerings. That’s one trend that users should look for and ask for from the vendors. There’s also a trend of moving away from inflexible pricing structures such as named users or server based pricing structures. Enterprises are very agile, and they’re always on the move so a person within an enterprise constantly moves from one role to another.
Enterprises constantly add hardware, reposition, and redeploy hardware. So the best pricing options are now coming from the side of software of a service subscription models or concurrent user licenses that really do not look at named users or fix the price to specific server usage.
These are some of the best practices that most of the leading vendors are moving towards and buyers of technology should ask for these as part of the negotiation. Cite competitors and the competitive advantage of other vendors when negotiating with other vendors who are less flexible.
Let’s assume that somebody has evaluated their software. They’ve successfully negotiated with a vendor. They’ve tied it to their business initiative. What are some of the challenges when it comes to actually implementing BI software?
Well, even before we implement the BI solution, and let’s say if it’s a good example, we’ve created the data governance organization. We took ourselves through all of the right software evaluation steps. One of the intermediary steps that I always recommend that very few enterprises follow, and I think it’s a very indicative step towards what vendor to choose, is to actually ask the vendor whether they use their own software internally.
So for example, if they’re trying to sell you a dashboard that will show you how your enterprise is performing, ask them if their CEO is using the same dashboard, the same technology to run his or her business. Sometime I would even recommend making an unannounced visit to that CEO’s office. Walk into the office. Take a look at their PC and see if they actually do have that dashboard running and if they don’t, there may be something wrong with what that vendor is telling you their product can do.
So once that step has passed, I think some of the best practices in terms of implementing business intelligence and any enterprise software applications is to create small, manageable steps. I am a big opponent of a big bang implementation approach as we talked about a vision and the road map so the strategic road map needs to be divided into tactical steps, and each tactical step needs to be a small department, 10 to 20 users and no more than that.
Keep it very specific. Very focused business requirements may be one particular aspect of that department. Let’s just say sales metrics and sales business intelligence. Don’t try to cover the whole CRM gambit which would include service and marketing metrics, as well. So pick a subject area. Pick a small department. Roll out a proof of concept. Turn that proof of concept into a prototype, and once the prototype is successful, move onto the next steps.
This works well for many reasons, number one manageability, but number two is that in most of the cases when one implements a new business intelligence environment, it can be implemented two ways. One is top down. An end user says what technologies they need, and technologists go out. Find the sources of the data, and make it available. But unfortunately very often the end users don’t know every single question that they will ask in the future from the business intelligence environment, so typically they say to the technologists, well, go get everything that is out there, model it, I’ll start using it, and then I’ll tell you what I want.
That’s a much more typical approach that’s obviously riddled with problems because one cannot just take every single piece of data and information, and model it without some particular requirements. So I say start with small steps. Let the end users play with that environment. It’s almost like a sandbox environment because only when they have the first data mart, the first set of reports they can touch and feel, only then will they start providing the technologist with the real requirements.
Now let’s talk about final advice for those with the tasks of evaluating purchasing or implementing BI software this year. We covered many of the areas for best practices in terms of starting with business first, data governance and data organization. Data stewardship is a very important piece of the puzzle. Read up on research pieces on data warehousing and business intelligence best practices. Definitely take advantage of them. Take a look at those research papers.
You can also search out Webinars like these where many of the best practices for implementing enterprise software including business intelligence solutions are discussed. We talked about data governance. We talked about a road map, and then taking small steps. Probably the last advice I will give our listeners is that unfortunately business intelligence still cannot be categorized as a science.
I cannot predict and categorize every single step, every single requirement, or every single best practice. It is still much more of an art. So it still takes a skilled experienced professional who has done it before, who has lessons learned, documented them and who has seen successful projects and unsuccessful projects.
So I would say if experience like that does not exist within the organization, definitely seek advice from a professional system integrations firm. There are many great system integrators out there who have done this for 20 or 30 years, who started back in the main frame business intelligence days and traced that whole process, who do know the lessons learned and do know the best practices. I would definitely recommend starting any significant BI initiative by partnering with an experienced business integrator.