My first agile CRM experience started out as a traditional, sequential, waterfall project for Premier Medical Group that quickly went sideways, almost went backwards, but turned around to become a pivotal moment in my career. In 2008 I was running Increase CRM, a cloud CRM service provider, and I got a call from Debbie Cragg, operations director at Premier Medical Group (PMG). Debbie was looking for a consultancy to configure and host a new CRM system to replace their legacy insurance workflow system.
PMG prepares medical reports for solicitors as part of accident-related medical claims funded by insurers and employers. Their goals were to speed up the process, make it as efficient as possible, and help the insurance companies standardise payouts to reduce insurance premiums. They had 300 agents based in Ludlow (over 3 hours drive from our office in west London) preparing nearly 200,000 reports each year.
The perfect requirements specification
Dan Barber and I spent several weeks driving up and down to Ludlow to meet users, model their business processes, capture their requirements and analyse their data needs. We held lots of workshops and pulled together the best possible requirements specification we could. Hundreds of pages in glorious technicolor. Sure Step's inventors would have been proud, but maybe not surprised by what happened next.
PMG's goals required a complex solution that managed patients' medical assessments in relation to their insurance claims. The medical examinations had to be scheduled in clinics arranged by PMG in multiple temporary facilities across the country every day and the doctors had to be provided with all the patients' files for the day of the clinic in a custom offline synchronizing application that couldn't use the standard client application. CRM had to manage a long-running case and all the related business processes with interactions via telephone, email and postal mail.
After spending several days reading the specifications and collecting stakeholder feedback, Debbie didn't think we had captured all of the requirements or their complexity. She was expecting a larger implementation estimate.
"Could you double your quote and add a contingency?" she asked. It's not often a client asks me to double my estimate (it's never happened before or since!). Then she added, "But I really don't want to wait nine months before we receive any software for testing. Can you deliver working prototypes before that?"
Let's go agile instead!
I can't even remember where I'd first heard about agile frameworks. (Possibly at a BBQ at Dirk Elmendorff's house one Sunday in 2008?). After a weekend crash course reading blog articles and a couple of days making calls around the Microsoft partner network I got connected to Paul Fox and a team at CIBER UK who offered to partner with Increase CRM and help us deliver PMG's complex CRM system using Scrum. I pitched the concept to Debbie and she was immediately on board. Now all we had to do was convince her board of directors.
Debbie and I confidently presented our new approach to PMG's executive team. But I think it was Shay Ramalingam from Nomura (an investor in PMG) who asked me: "How will we know when you're done? How will PMG know that the requirements have been satisfied if the requirements specification isn't part of the contractual agreement?"
It was an insightful and fair question, which makes it a shame that I flubbed the answer.
Don't use this line in a pitch to executives
"When you've run out of money".
That was my off-the-cuff reply. I tried to back-peddle fast, "In an agile approach the client sets the priorities according to business value, so PMG will control the scope and costs and can choose to keep the development going as long as our team is delivering features with more value than we're charging you to develop them".
And that is how my first agile CRM project got approved.