Photo Credit: Ernesto Andrade Licensed under CC BY 2.0

Imagine that you’re in the middle of an analytics migration project (as Dell was).

Hundreds of users and projects all over the world are in transition from your legacy analytics product to the new one (as ours were).

Everyone is heads-down-focused on following vast, detailed project plans with a jillion moving parts (as we were.)

Suddenly, out of nowhere, an opportunity to fix Something Else swoops onto the scene (as it did onto ours).

That Something Else is kind of a mess, and this would be the ideal time to fix it, but everybody around you is urging you to focus, focus, focus on migrating projects and users. The Something Else has to do with the tools and processes on either side of the analytics function. It’s not exactly the same issue as replacing your company’s advanced analytics product, but it’s closely related.

What do you do: stay focused on your original project or devote some cycles to dealing with the Something Else?

ETL, data extraction and reporting. And the timing belt.

At Dell, we were waist-deep in migration from a well-known analytics product we had used for decades (you can probably guess which one) to Statistica, a product we had recently acquired. As I posted a few weeks ago, our migration project team discovered that a lot of people were using a Ferrari to haul dirt – that is, using a powerful analytics tool just for data manipulation – so we made some organization and tool changes as part of the migration.

But the Something Else we discovered was that people were using dozens of tools for some of the main functions around analytics:

ETL (Extract Transform Load) Process Automation – Microsoft SQL Server Integration Services, Microsoft Visual Studio

Data Extraction – Microsoft SQL Server, D3’s JavaScript library, Adobe Site Catalyst

Reporting – Microsoft SQL Server Reporting Services, Microsoft Access

We had the opportunity to consolidate or replace these and stop the tool-creep, and it seemed as though we’d never have a better time to do it. Once everyone saw the inefficiency, the whole migration team wanted to deal with it, but it wasn’t part of the original plan.

It was like the timing belt story:

“Well, ma’am, your car has 90,000 miles, so we should replace the timing belt. And while we’re in there, we’ll have everything apart, so if there’s a problem with your water pump or the tensioner or the front cover gasket or the seal, that’s the best time to take care of it.”

It’s tough to bite that bullet and deal with the Something Else. But you know if you don’t deal with it and you have to go back in again later to fix it, you’ll kick yourself.

Actually, you won’t have to, because your boss will do the kicking for you.

The Great Analytics Migration – new e-book

So what did we do at Dell?

We went the extra mile and did the consolidation. It’s the kind of company we are: we can’t look at an inefficiency and not do something about it. Our companywide Business Intelligence Council ran a survey that found dozens of tools at work. The council identified seven of the top ten of those additional tools for migration to appropriate Dell technologies. We’ll get to the rest of them eventually.

Should you migrate users from other tools in the same project? We can’t tell you how to make that decision for your company, but we’ve put together an e-book, “Statistica: The Great Analytics Migration , Part 3: Technology, that tells you how we made it at Dell. Read the e-book for unique insights into how we managed our migration. We know quite a bit about it.

Just don’t ask us about your timing belt.

 

David Sweenor

About David Sweenor

From requirements to coding, reporting to analytics, I enjoy leading change and challenging the status quo. Over 15 years of experience spanning the analytics spectrum including semiconductor yield characterization, enterprise data warehousing, reporting/analytics, IT program management, as well as product marketing and competitive intelligence. Currently leading Analytics Product Marketing for the Dell Software Group.

View all posts by David Sweenor | Twitter