Tuesday, October 4, 2011

"We haven't really thought about that"

If you have been dealing with BI related sales activities or are searching for that "right" BI tool, you will find that most organizations:

* Use manual tasks, including desktop query and reporting tools, to answer their business questions.

* Have "something" in place that they are not happy with or is costing them too much money.

* Have data in multiple silos that they need to access, consolidate and optimize.

Hence, they are usually looking for a low cost BI alternative that can provide them with the answers to their business questions, as well as ease of use and functionality within their budget.

Don't believe me? Join the many BI groups available in LinkedIn and other social networking type sites and you will see the barrage of questions from those looking for recommendations on BI tools.

So...I was on a call yesterday with a well known organization where the "prospect" stated: "We need basic reporting with the ability to access all of our data without moving it or massaging it."

"Okay? That is absolutely possible, however do you understand the pros and cons that are associated with that?" I replied.

[...] Silence...not only can you hear crickets on their phone, but you could hear them in the next conference room over. I took the proverbial saying "Silence is Golden" to another level. It became so uncomfortable that the Account Rep felt he should interject. I interrupted promptly to allow them to answer the question. After about a minute of what appeared to be hours of silence, they responded, "We haven't really thought about that." - BINGO! Case closed! Next!

Hmm... "We haven't really thought about that."

That's the problem, no one is taking the time to be proactive and think about what it is they need and are rather just reacting. "Let's see a demo." "I just need Reporting". "We need Dashboards." If that is the case, I would recommend you watch a video demonstration which may intrigue you to start thinking about what you really need. Then come talk to me when you have more criteria that will support your BI initiative. :-)

So I digress...in turn, I took this as an opportunity to educate by asking pointed questions that would help them see what it is they actually need vs. what they thought they needed.
  • Is the data you need to access all in one location? - No

  • Does the data you have support a majority of questions that will be asked of it? - Don't know

  • Would you like answers to questions that occur on a regular basis? - Yes

  • Would you like your users to answer their own questions on a random basis? - Yes

  • Would you like your users to explore and discover answers to questions they did not think to ask? - Yes

  • Do you have a predefined set of KPIs to manage and track business performance? - Yes

  • Would you like your executives to see an at a glance view of those KPIs? - Yes

  • Would you like to be aware of "something" when a defined threshold is met? - Yes
Alright, now we are getting somewhere. Each of those questions and responses clearly identifies that their needs are more than just simple reporting as originally desired. They require a solution that encompasses both Data Integration and Content Delivery. (ETL, Reporting, Analysis and Dashboards)

I further probed as to why they wanted to access all of the data "without moving it or massaging it". They replied: "Because building a Data Warehouse takes too much time and costs too much money."

Wow! Clearly a response most likely seeded by a competitor whom believes they can access all of the data where it sits, without building a DW. Which may be true for some of the competition out there. However they usually leave out the fact that they are still "moving and massaging" the data - they just don't call it ETL or refer to their process as Data Integration or even use the words "Data Warehousing".

I further explained that Data Integration (ETL) does not have to be about building an EDW, Enterprise Data Warehouse. It can be about building operational data stores that are refreshed periodically to support questions that the business users want to ask. It can involve federated queries where the data is accessed from the source without having to stage the data. It can also be about normalizing data in to a small data mart that supports speed of thought analytics for the power users.

Upon those points I provided a demonstration of Pentaho's "Agile BI" capabilities which involves a rapid, collaborative and iterative approach to building BI applications. At completion of the presentation, the prospect was amazed and pleased. They stated: "This is exactly what we need." Music to my ears.

People, you cannot throw a BI tool in your organization and expect it to stick without asking some important questions. It is those answers that will help guide you to the right solution. And most importantly, you cannot put a BI tool on top of all as-is data without knowing what questions are going to be asked of it. It is impossible to know every question that may be asked, but at least have those that are important to tracking your business performance.

On the majority of calls that I participate in, it seems that organizations don't have the time to properly plan and discuss the criteria needed to implement a decision support system. Why? Because everyone is doing more with less these days and researching a BI tool is usually an ancillary responsibility for them. If that is the case, allow us to help you with your research and we will ask those question you haven't really thought about.


Michael Tarallo
Director of Enterprise Solutions