«

»

Building BI into the Budget

Share

Building capability for BI into the corporate budgetThere are a lot of presentations and publications around the glamourous areas of realising the value of BI, and measuring the value that BI delivers to the organisation in terms of outcomes affected by additional, quicker or more focussed insight. However not many focus on the slog work of getting the funding to get BI off the ground or even just paying to keep the lights on. In this post I focus on aspects of BI that should feature on the CAPEX and OPEX budgets.

The organisations that are mature in their adoption and implementation of BI are grappling with issues such as determining and proving the value-add of implementing a BI strategy, or trying to determine BI’s value-add to the business. However, even more organisations are still battling to get BI properly off the ground – and their battle often centers around securing funding. These organisations are typically still trying to estimate the size of the initiative and determine the cost of the investment required.

The reality is that many organisations are still feeling slightly overwhelmed with the amount of data they are producing, and are still trying to figure out what they need to do with that data in order for it to actually be valuable to the organisation. So in order to get any value out of BI, you first need to do the initial work to frame up a BI strategy, and then start implementing according to that strategy. This initial work needs to be budgeted for. Well, how much is that on the budget? At a thumb suck, I would venture two pretty senior consultants for 3 months to get a high level BI strategy in place. Shorter than that, you hardly have enough depth to start running BI seriously, but longer than that, you are dropping into too much detail too soon.

Next comes the step to plan for and budget for the actual BI implementation. As a key part of the BI strategy – before even drawing up a BI budget, the organisation needs to decide on the scope of the intended exercise – whether you want BI to be spread across the entire organization from the start, or whether it will first be implemented within one department. It may be a politically loaded decision, however, if you slowly start implementing BI one department at a time, the new processes won’t seem so frightening and employees and management will start to understand the concept faster – and the initial budget won’t be so daunting either. So you can approach your budget by starting with a small but crucial focus, and then build from there. Of course, you need to keep in mind, and prepare the sponsor and stakeholders, that there will be a bigger budget requirement over a much longer term, but some of those implementation costs can be deferred until some justifying quick win values have already been gained in order to offset those costs. (But naturally, in order to offset value gained against future costs, the value must be justifiably measured.)

So what do you need to budget for? The following is a good starting point:

  • Hardware costs for development, testing and production environments, which includes either the initial costs (CAPEX) plus support and maintenance (OPEX), or an on-going leasing arrangement (OPEX). You will be surprised how many organisations start off BI on a single production environment. Don’t go there. It is hell to reverse the culture and break the bad habits to move to a proper set of managed environments.
  • Software licences, at least for data warehouse database, ETL, reporting, dashboard and visualisation tools (CAPEX) plus ongoing maintenance and support (OPEX), across all three environments and for all expected user populations (as determined in the BI strategy). User numbers may grow over time, and later on you may want to add an advanced analytical modelling tool too. Think about metadata management, MDM possibly (although that should fall outside of BI) and a decent data modelling tool too.
  • Personnel costs for at least a BI solution architect / data modeller, a business analyst / tester / trainer, an ETL analyst and a report / dashboard developer (thus 4 FTEs if the multiple roles are confusing.) That is the smallest team you can run with, unless you want to bargain on a one-man or two-man band to do everything, which is not very practical in terms of delivery capability. Bigger organisations with bigger data warehouses and BI implementations scale this team up – anywhere from 6 people up to about 30, and I’ve heard of organisations with BI teams of over 50. Also think about DBA type skills to manage the environments, databases, tools, and so on. I would not couple the BI team to a specific period estimated for the development of the BI capability – BI is an on-going program, so include the team costs in your on-going OPEX budget.
  • Training costs – it is very important that your BI team and your key users be trained on the technology they are going to use to create and use the BI capability. Skimp on training and you forever have teething and adoption problems. Of course the user training budget can be scheduled across years as the BI strategy is implemented business area by business area.
  • Consulting services, in case you need some assistance to get and keep going. You may want to keep the two consultants that did the BI strategy on for another 6 months to flesh out the full strategy and ensure your team stays on track. It will be good to do an objective BI Health Check at about the one year and two year marks too.)
  • Keep in mind that somewhere down the line (like at about the one year mark, if you have a very efficient team), you will have to start budgeting for on-going maintenance and enhancements costs (in addition to the development team listed above). As your team puts more and more BI capability in place, it all has to be managed and maintained, through application changes, business changes, infrastructure changes, etc. Keeping up with business changes and catering for technology changes come at a cost. How much you may ask? For a small team like the one I listed above, I would add an FTE after a year, and another one at the 3-year mark, depending how fast your BI solution and its adoption grows. Obviously you have to scale this up for bigger teams.

So, once you have decided how you are going to implement BI, the BI manager and CIO needs to sit down with the CFO and the CEO and decide how much budget they should spend on a BI strategy and the subsequent implementation. Be warned though, going cheaper is not always the best approach, as it is critical to have a BI strategy that ensures that BI meets the organisation’s goals.

In a white paper published by Sisense, they state that most BI projects should cost around 1% of your annual operating budget and about a quarter of that in the years to come. Most companies see these figures as a bit discouraging, however, it is important to note that while it may initially seem expensive, remember that there is no room for error with BI – you either often have incorrect and untrusted information, which is basically worthless, or alternatively you have a BI solution will give you the correct and accurate information every time.

Concluding thoughts

So given the above, surely more organisations should be able to budget for a simple BI strategy and then budget for the actual BI implementation based on the BI strategy. Doing it two steps like that can alleviate some of the pressure on the organisation.

Keep in mind, although BI has to meet current business needs, and you draw up the budget accordingly, the BI strategy should be flexible enough to cater for any changing data and informational needs. The budget needs to accommodate that level of flexibility too.

Leave a Reply