Changing production models: corporate psychology vs. customer satisfaction
Author: Jared Dolich is a Certified Master Anaplanner and Partner and Founder at Retailitix.
Consider this cartoon from Clowncurls.
Maybe a bit exaggerated but sadly not too far from how Anaplan change requests come in. The question is how should you respond?
Long ago I was a service desk manager for a 5000-store retailer, and we received 900,000 contacts per year from store associates, customers, and corporate employees. Fortunately, I was exposed to the Information Technology Infrastructure Library (ITIL) and Lean Six Sigma (LSS), to help ensure quality responses to most of these contacts such as incidents, problems, and service requests.
I am fond of a concept from LSS that, to this day, is predominantly used by most IT organizations which is to create a 2x2 impact effort matrix regarding how to prioritize service requests. The Anaplan Way uses the same approach as part of the planning poker and sprint review. It would look something like the chart below, but instead of effort, we’ll replace that with YOUR tolerance and discipline to best practices.
The key to your response to any production change is understanding customer expectations. So, let’s assume your goal with any Anaplan change is customer satisfaction. In that light, let’s shine intense beams at each quadrant of this 2x2 grid and see what consequences manifest.
But first..
.
I want to mention all the Anaplan best practices related to production changes here to be thorough. I won’t go into detail of each one, rather leave that to you to look them up or refresh your understanding.
- DISCO
- PLANS
- ALM
- The Planual: Misbah and Rob’s newest version
- The Anaplan Way: especially user stories with success criteria, sprint reviews, testing, and planning poker
- Data Hub: especially hub to spoke synchronization.
- Center of Excellence best practices: unique to each organization (e.g., naming conventions)
Did I mention testing?
- Test scripts
- Unit testing (this discussion on Testing in Anaplan may be helpful!)
- User Acceptance Testing
- Dev-QA-Prod migration testing
- System integrated testing
Awesome. Now let’s look at the four quadrants with our best practices in mind.
- Anaplan Champion. Yeah, this is intense and takes a lot longer than usual, but you cannot risk making a mistake, especially one that cannot be recovered. Consider a process that has 100 to 200 actions to synchronize your data hub to your applications and across multiple workspaces. There’s not much chance of recovery.
- Gambler. This is where your professional integrity and self-interests are going to collide. The boss wants the changes now! Will you be a hero, quickly migrate your half-baked changes to production, and leave it to the next modeler to deal with the short cuts you took? It’s possible no one will remember it was you that made this change, but one thing is certain. If Anaplan is causing business delays while the next modeler is figuring out what you did, you’ll tarnish Anaplan’s reputation. Be professional. Use best practices. You’ll win the long-term reputation for you, and Anaplan.
- Sinister. It’s called sinister because the change is trivial and the process to migrate the simplest edit is immense. Sinister to you and frustrating to your customer short-term. Will you cave for short-term heroic status or make the investment to maintain a highly functioning application that will last a lifetime? Look superficially beautiful now or have personal satisfaction of professional work later? Not an easy decision, especially if the ones demanding the change are intimidating.
- Naïve Quick Hit. The easiest of the four in my opinion because you should only encounter this quadrant when you’re training, sandboxing a POC, or doing some research. This is your personal quadrant. Have fun with it. Just make sure you never promote anything from this quadrant into production as-is. Like what you see? Great. Now move this to the next sprint and use first quadrant.
Summary
Only you can decide where you will draw the line between corporate expediency and long-term sustainability. Unfortunately, the two rarely complement each other. What I can tell you is that the best models out there are built using the best practices, PLANS in particular, and tools mentioned above. If you’re fortunate enough to be working with an Anaplan CoE then you should be able to find colleagues that will support your interest in avoiding Anaplan deprecation. At the very least, leverage the ongoing sprints to get your changes prioritized along with everything else that’s happening.
Comments
-
Thanks as always for your insightful (and entertaining) thoughts! The key that cannot be understated is assessing the impact before you decide the approach. I've seen many a good intention be undermined because proper care was not taken to proactively understand the dependencies and production data impacts.
0
