It seems that after 20 years we still have articles being written on “Why do Business Intelligence Projects Fail?”
Many of these articles summarize the reasons typically within 3 general areas:
– were too expensive, took too long to implement
– were not easy to use, complex to adjust/modify
– they had great visualization but poor data management
– great datawarehouse but visualization was poor
– Users/company resists change, user adoption is to blame
– More comfortable with status quo/spreadsheet
– Always needs IT to make changes
– Management got sold on a premise that BI will yield specific financial results
– There was no process and BI needs specific process definitions to work within
– BI is a “culture” or a philosophy
– Data was poor/bad quality, garbage in-garbage out
This is by no means an exhaustive list but in essence it summarizes most of the reasons why, according to Gartner, over 70% of these types of projects fail.
Then come the specialists, consultants and vendors with the solutions to the above issues, myself included in all of these groups, having been a business and IT client charged with implementing BI in many companies, a BI consultant charged with implementing several BI solutions, and a vendor of a BI solution. This serves first as a disclaimer and also as a way to declare that I do have some experience in the subject matter.
What do the specialists, consultants and vendors typically offer do border on a “magic” bullet with promises of increased revenue or financial results?
If you think about it BI is no different than CRM and ERP systems. These are not systems that will yield better financial results, these are in essence, management systems. You need these to manage a group of people, customers and suppliers, their activities, efforts, costs and payments. You need them because your brain and memory is not enough to track all the details and volume of data, spreadsheets and word documents becomes unmanageable, and hiring more people to handle the paper work, processes, and tracking it all is costly.
So we look at software to make our company run smoother. If done correctly, with the right software and right people and clear processes, it will provide operational savings. It will not help you sell more. It will not magically give you an increase in profits unless your sole reason to do so was to lay off staff which is already a sign that you are heading down the wrong path.
You implement CRM, ERP, BI and other software to give you clarity on how you are running your business, both from a transactional and analytical point of view.
– There is no point in having invoicing processes fully automated if you can’t summarize and aggregate how many invoices are due and what is their amount, how quickly do clients pay you, etc. etc.
– There is no point in having a sales force contact management if you can’t summarize the activities of that sales force by day, week, month, or to track their progress against objectives.
– There is no point in having an expense tracking system if you can’t monitor at a grouped level who is spending how much on what and when.
All these aggregations, summaries, totals, are needed from a management perspective. Furthermore, it is not just about monitoring each individual process but the entire company across the different divisions, departments, teams, client groups, products, and so forth.
BI software is not an optional tool and should not be sold as a way to increase your profitability or guarantee sales results. BI software is as critical as a dashboard in your car, the control panel of an aircraft, the GPS and radar in a ship. You need to know where you are and where you are headed when managing a team, a product or an entire company.
Without it, you are might as well take your invoicing back to paper or manual word documents, have your field force schedule appointments with clients in their personal agendas, track expenses in an Excel spreadsheet. After all, if you don’t use this data to build dashboards and reports, in conjunction with other relevant data, to make decisions why track it in the first place?
Here is advice #1 – if you are looking at implementing ANY software. Start with the metrics – the dashboards and reports that you expect to have once the tool is implemented – and work backwards. If at the end that tool is unable to provide the data that can be transformed into those dashboards and reports then that is not the right tool.
Selecting a BI tool is not easy, the difference between data visualization and data management, plus all the hype about mobile BI, social BI, big data, in memory, NOSQL, OLAP, OLTP, and all the other acronyms, fads and marketing words is enough to confuse experts and beginners alike. But there is a new class of software applications that are different than most. Many people call them Cloud applications or SaaS (Software as a Service).
That in itself is not enough since many applications these days are cloud/SaaS in one way or another and they have the same pitfalls as any other tool. Being Cloud/SaaS does help many of the implementation costs and time associated with putting in place such a system but that is not enough.
Cloud/SaaS “Agile” applications is what you should be looking for. These are applications that provide faster results (or not) than their non Agile equivalents. The key here is FAST. If you can’t do it, the data does not fit, the visualization sucks, you can’t publish, it is too expensive, find out why FAST, don’t wait until you sign the deal, buy the servers, get the consultants on board to determine at the end of a certain period that “something” does not work.
Cloud/SaaS Agile BI tools allow you to:
1. Create an account in a few minutes
2. Load some sample data in a few more minutes
3. Build a few dashboards to see what you can do with your data
4. Looking good? Continue.
5. Go back to 2. with some more data, or more real data or more variety of data and repeat process.
Iterate through the development of your own dashboards and build it quickly or fail quickly, but either way you will know immediately the weaknesses and strengths of the tool. And you can do this before you even spend money and large amounts of your entire company’s time on the topic.
So, while many say that the issues in implementing BI successfully in an organization is only partially due to the tool, I only agree if the tool that was selected is the wrong one. The right tool should help you structure your process and make it easy on your people to use the tool with minimal effort.
So here come my tips for your 2017 BI success:
Work backwards from the end result, the reports and dashboards that you need to manage your business.
Select an agile/cloud tool that is easy to implement, available at an accessible cost and that can deliver all the core functionality you need in one tool – data connectivity, management and visualization.
Built it gradually and enhance it as you go, do not make mega projects out of them because they will fail or at best, cost a lot of money.
This is the best low risk approach to avoid a failed BI implementation.
Good luck and all the best for 2017!