the analytics edge

I noticed a recent job posting for a Business Intelligence Analyst, and it struck me how far the BI industry is from where it needs to be.

‘This role will support business analytics for the rapidly growing business by conducting daily, weekly and monthly assessments required to support strategic level decisions by management.’

Sounds like a position that should exist in every company, doesn’t it? The way I read it, though, two things popped out — daily/weekly/monthly, and “strategic level decisions”. Those are two concepts that, when connected, expose the weaknesses of our current business intelligence technologies and have a direct impact on the ability of companies to staff this role.

First, I should explain that I believe strategic decisions - the kind that affect the way a business operates — do, in fact, happen daily, weekly and monthly. Challenges and opportunities do not wait for a quarterly strategic review meeting — they pop out of the woodwork and demand a reaction - instantly. Competitors act, you must react. Acquisitions happen, you must re-evaluate. Markets shift, you must adjust your investment plans. If nothing has become clear in the past few years, it is that your ability to react quickly determines your success in business.

Second, I also believe that information required for strategic decisions actually have a very short useful lifespan. When that unexpected event occurs, you suddenly need to know a few specific things about your own business that you likely weren’t paying much attention to. You need to evaluate scenarios that you did not consider before. You ask for reports you never needed before. But when the decisions are made, that information serves little purpose - maybe a little follow-up to make sure things turned out the way you hoped, but it is not required for operations. The reports languish.

This is real life. This is what Business Intelligence must support. These decisions, these reports, are crucial to company futures. That is why executives invest in BI. They want answers when they need them. IT groups know the scenarios all too well. Some refer to it as the report of the quarter; a high level executive suddenly demands a new report that is based on information no one has ‘cleansed’. A huge effort gets put into cleaning up the data and performing the analysis to deliver accurate information. Then they sit back and watch the report seemingly never get used and wonder why they bothered. A strategic decision was made – the most common one — to do nothing different, but those kind of decisions are invisible. 

So what does a BI solution require of the business analyst position?

‘Create Star Schema, develop cubes, KPIs and dashboards for Products Develop automated reports using SQL Server 2005 Reporting Services Design data mining structure for management to create their own reports Analyse source data and determine fields required Data mapping’

Can you believe that some people actually debate whether IT matters? I thought we were supposed to have left the techie stuff behind years ago and computers were supposed to make it easier to do the things we needed? Sure doesn’t look like anyone told the BI vendors…

It’s time for change. We need a different approach to business analytics. A simpler way.

I came across the first of a two-part article by Ned Madden today that does a great job of exposing the range of opinions about analytics and what it means to different organizations.

I particularly liked the IDC quote defining business analytics software as “solutions used to access, transform, store, analyze, model, deliver and track information to enable fact-based decision-making.” That sums up the general feelings people would associate with the words ‘business analytics’ when not encumbered by dictionary definitions. And it is that emotional attachment to the words that will elicit investments by senior managers. I have to note that the definition could have been written in 1985 and is the fundamental reason behind virtually all business investments in computer technology.

Managers want more information to better run their business, and the role of computers in maintaining and understanding daily operations continues to grow. The speeds and feeds have been producing lots of reports with even more numbers, yet managers are increasingly challenged to find the time to look at them. BI and application vendors know that most of the reports in their systems are never looked at. IT organizations know that a change in management or disappointing quarterly results means a spike in new report requests.

You can see the dynamics in the market — there is a pent-up demand for something new that will make people’s lives better. Mr. Madden has captured some of the excitement building in the industry –terms like custom fit and competitive edge are headlines, and there are lots of niches for vendors developing. I look forward to seeing part II of the Breakthrough in Analytics.

And so it goes…trying to define the word ‘analytics’. Well, maybe we can start off by agreeing that we won’t agree, as James Taylor says. But try we will.

This is predictable, and was evident more than two years ago when the term was just beginning to be picked up by mainstream marketing minds. With the aging of the “BI” term and the changing landscape in the industry, the players need new magic phrases to get a piece of the lucrative IT spending priorities (full disclosure: yes, I work for one of them and I want some of it, too). We have probably all heard that analytics are an enabler in high performing organizations, and books like Competing on Analytics are reinforcing the point in the executive ranks.

Marketing knows what that means — they may not say it outright, but their unspoken desires are building. Every competent executive is uncomfortable with what the future may bring. They need information to make decisions, and they see the word ‘analytics’ popping up in the business press. They don’t try to define it, but they start to feel positive towards it. The more they see it bandied about by others, the more they are attracted to it, and their very own marketing VP is probably talking about things she’s learned from the web analytics.

If someone, like the CIO, picks up on an opportunity to use a buzzword to fix a dwindling budget, then an investment in ‘analytics’ has been born. It could be anything, so long as you can stretch the term to fit the actual spend in some way. Now, I am not suggesting people invest in unnecessary projects just because they used some fancy buzzwords, but anyone that’s been in the IT leader position knows you need to drop those words to get over the resistance hurdle for the projects that the business does need. Once the project starts, vendors pick up on the new ‘analytics’ investment, and they use the same technique: drop the words to get over that initial resistance.

So, should we define ‘analytics’? Absolutely NOT! Leave it as wishy washy as we can get it, and leave the door open to interpretation. Everyone wins! The CEO gets to brag about his new investment in analytics, the CIO gets his project budget approved, the vendors….we get past the gatekeepers to talk to the people that matter.

[oh, and NextAnalytics actually does sell a product that does ‘analytics’ :-)]

As a long-time IT exec, I’ve seen the studies, listened to the sales pitches, questioned the case studies, and lived through the hype cycles. In organizations, there is a perceivable cycle that repeats itself -the current strategy is getting long in the tooth, so management searches for the next ‘thing’.

Someone in marketing recognizes a festering annoyance in a few leading organziations and coins a new phrase that others start to rally around. They know that management spends money to deal with festering annoyances, but they hate pouring money into the same thing, year after year after year. That is why it is so important for marketing to supply their new phrase with some nice sounding “differentiation”.

I remember the need for “reporting systems”, the fancier “decision support systems”, robust “data warehouses”, the more agile “data marts”, prestigious “executive information systems”, the all-encompassing “business intelligence” solutions. All of these were supposed to give management a better understanding of the business while dangling the elusive carrot of competitive advantage.

So that festering annoyance put business intelligence applications at the top of Gartner’s technology priority list for the third year in a row. Management is still not happy with the information they have available to them. They think they can get better reports, and will soon get tired of investing in the same old thing. The BI industry is in a really bad state to respond. The Big BI Consolidation has distracted all the wrong people, and there hasn’t been a lot of innovation to hook a new marketing phrase to.

We all know that the lack of a product would never stop a good marketing group, so is it any surprise that, after Google Analytics helped them understand their web site usage, they are identifying a bigger requirement for ‘business analytics’?

Business Intelligence is dead (as a phrase). Marketing is killing it. Long live business analytics.

Well, three years at least…