Evolution

If you are going to analyze data successfully, you need to keep in mind that analytics are evolutionary. Analytics need to evolve along with an organization’s applications, business processes and even the organization itself. But this is easier said than done – If you are going to enable an evolutionary analytical process, you need an agile development process, you need iterations.

Analytics change all the time. For example, the day before Lehman Brothers went out of business I’m sure people looked at all kinds of reports. Do you think they looked at the same reports the day after? Typically when organizations first deploy BI they may analyze say 10 things within their organization. After a while, that number grows to be hundreds – and as it gets to know its business even better, that organic growth of metrics and data proliferates.

You may be asking how do organizations get analytics to evolve? Well, fundamentally, organizations need a capability (something you have to develop and build) comprised of people to lead the process and guide it towards business value, and they need a process which needs to be iterative. Lastly, they need a technology that can support the evolution.

End users’ requirements for insight always increase over time because they need access to more data sources to build a more complete view of the business and they need more ways to analyze and interact with that data for them to make effective decisions.  Unfortunately, while the pace of requirements change is high, Legacy BI is very slow to deploy and update or enhance. This means that the first phase deployment is often 6-12 months in duration or even more. Subsequent releases typically take 6 months or so. This pace of enhancement means that the business users’ requirements will never be fully met – forcing the business users to “fly blind” in key areas and this ultimately lowers the effectiveness of the overall decision-making process.

Agile BI offers a much better way for companies to meet their end users’ demands for insight over time. By enabling rapid initial deployments and more rapid iterations over time, Agile BI enables the end users to explore a more comprehensive view of their business in more flexible ways more quickly and affordably than Legacy BI – enabling the BI system to keep pace with the end user requirements over time. By better keeping pace with end users, Agile BI enables better and more effective decision making in more areas of the business more quickly and affordably than traditional BI solutions. This difference enables end users to drive more business value for their organizations over time.

Fundamentally, I believe there is a right way to do analytics. Whenever I turn something on and a user says, “I want this and this,” inevitably, when I deliver it –  the user says, “This is exactly what I asked for… but now that I see it, it’s not what I want!”

If a user wants changes, like simple things such as a different chart type, or a new KPI, or something more complex such as adding a new data source, etc., with the traditional approach to BI, each of these changes takes months to execute. Typically an organization may spend 9-18 months on a project delivering everything a user wants, but what are the odds that the requirements are the same from the outset of an 18-month project? Can you afford to take a year to start over? No. Business doesn’t move that way and neither can BI. We need to be agile and iterative.

When we say iterative, we know that reports and dashboards are relatively easy to change, but we also know that the underlying data is not. Organizations need to evolve reports and dashboards over time and need to continuously ask, is this the right data? If you added Netsuite to Salesforce data, for instance, what new things could you measure? Having the mind set that we need to iterate is key, because you can’t be successful with your analytics strategy if you don’t take this approach.

What about value? How do you drive value? Some organizations are better at this than others. Some are still learning.

When you start an analytics program, sometimes you see Mr. CFO just wants certain reports and inevitably these will need to be delivered. In an ideal world, the CFO identifies the key metrics he or she is accountable for. The same holds for other roles where key metrics will vary from organization to organization, and from department to department within the same organization. They could be related to revenue or lead generation; maybe they are related to inventory turns and the supply chain; or to new customer acquisition. Whatever the metric is, the key is finding the drivers for it and in turn, finding the actions that move these in the direction you want them to move. It’s these actions that will drive your report and dashboard development. In turn you will find that key metrics in one department are drivers for another and so you will be able to connect the organization together through analytics.

To be successful with analytics, you need a strategy that enables consistency and freedom of analysis for end users, but that ALSO enables rapid development and provides richness of features, and you need to drive value for your organization.

What does this mean for you and your organization? If I was going to sum it up, here’s the nitty-gritty of what you need:

  • An analytical platform that enables organizations to deploy rich, capable analytical applications quickly.
  • The ability to combine ETL and Logical modeling into a single, automated step.
  • To be able to build a robust data model / data warehouse infrastructure quickly – and then deliver end users several ways to receive and interact with data by using a single BI platform.
  • To be live fast – let’s say 2-4 months (or faster if you can) and you need to be able to deliver iterations at least monthly thereafter.

I know one company that provides all this and more and offers it in their Cloud and as an Appliance that you can run in your own Cloud.

Here’s a hint: It rhymes with “First”.

Think fast.