OptimalBI AgileBI Man

Agile is trendy!

It’s all the rage these days to hold a few standups, start referring to weeks as “sprints”, change your mind frequently, and claim to be Agile. As we like to say around here “that’s ad hoc, not Agile!”
And if you search “Agile BI” you’ll find many attempts to apply Agile principles to the world of business intelligence. Apart from some excellent books by people like Lawrence Corr and Ken Collier, plenty is hot air: vendors trying to sell you the “Agile” version of their same old tools.

Agile is tough, especially in the world of data

The real difficulty is tying it all together: Agile is a solid set of ideas, but implementing it in data projects against moving targets (changing data, requirements and people) is tougher than designing screens for a mobile app.
For the past few years OptimalBI has been on a journey towards AgileBI, and boy is it tough. Like everyone else, early on we fell in to the “rename things but don’t change process” camp, with predictable results: no change in behaviour. In fact we quickly found that just pretending to be Agile is actually slower and riskier than Waterfall, because all process goes out the window, producing chaos.

How we became experts at AgileBI

We turned a corner about a year ago upon discovering the BEAM✲ methodology developed by Lawrence Corr. It filled the first of our “Agile BI circles” (requirements) with a robust and easy methodology, tailored to data projects. Like many things Agile it starts with post-it notes and whiteboards, but also produces usable documentation.

Since then we’ve slowly built out the rest of the circles and hardened the way they work together. Some of the challenges are technical, but most are human: how can we make business intelligence a truly team sport and deliver value for money faster and safer? No toolset on the market makes that automatic.
Some of our customers have shared this journey with us, and together we’ve “walked the talk” on AgileBI. But we firmly believe that true knowledge comes from teaching something, which is why we (eventually) turn everything we learn in to a workshop. That’s how we hardened our use of BEAM✲.

What we offer you

Last week we reached a critical point: we proved our knowledge to ourselves and a customer by taking them through a one-day overview workshop on AgileBI. That’s “talking the walk” on AgileBI; sharing our journey, our recommended workflows, and the collection of cheap and free resources that help glue it all together.
Here’s an overview of the day, module by module:
0: Defining AgileBI, the common problems it solves, and how it changes our work
1: Implementing AgileBI with dual workflows, Scrum-based “Team Flow” and Kanban-based “Fast Flow”
2: How AgileBI changes scoping and stakeholder management (and what to do about it!)
3: Gathering requirements with BEAM✲, rule patterns, and wireframes
4: Using “data layers” to increase resilience to breakages and make changes to data visible
5: How test-driven development, continuous integration and peer programming make all this possible
6: Developing reusable content “early” and using AgileBI to iterate improvements
It’s safe to say I wouldn’t be writing this post if I didn’t think the day was a roaring success, and I’m sure the team of 25 that joined us for the day think the same.
We’re ready for the next customer, so if you’re interested in starting your AgileBI journey with us, get in touch!
Until next time, keep asking better questions,
Shaun – @shaunmcgirr
Shaun blogs about analytics, machine learning and how data solves problems in the real world.