Suitability of impact map for non-feature work (a migration project)

[Rob Harrison]

Hi,

Can/Should i try to use impact maps for a migration project? I'm a scrum master and have been very interested in impact maps of late, with not much opportunity to use one until now (perhaps).

We are undertaking to migrate some 90 micro services for our loan platform to the Amazon cloud. The 'why' is clearly "Save money on infrastructure (by moving to Amazon)", the actors are the various teams involved, plus third parties.

My problem is when we get to the 'what' of each actor that will perform migration to the cloud because the 'what' is basically a list of "migrate service x" for each team performing migrations.

Adding the 'how' to each 'what' will generate a list of tasks which should map to user stories.

Does this sound like a good use of an impact map? 

I'm using mindmup to experiment with this concept because i love the idea of a large map showing what needs to be done however the inter-dependencies involved are going to make this quite messy.

I appreciate this may be hard to visualise, i can link to my current map if that would help, however i'd need to obfuscate some info.

Regards

Rob 

[Gojko Adzic]

Hi Rob,

The short answer is yes, but most likely not the way you started :) 

I'm currently interviewing organisations on how they used impact maps for a new book, and there is a pattern emerging around restructuring the scope of a project -- where the scope is known (or they think they know it) upfront, but there is a lot of work, potentially dependencies, so alignment on the sequence of work is important, and ensuring that people keep the eyes on the prize.

in this case, the actors aren't the individual teams... they might fit around deliverables, but it's not their behaviour you are likely trying to impact. if the goal in the middle is to reduce operational costs, then by migrating some service and associated set of features, you may help someone (eg an inventory manager) perform their tasks (eg generate inventory reports) cheaper. breaking things up like that might help you figure out which parts of the infrastructure aren't needed any more, which are just not that important, and which would be low risk to start first (= those impacting low risk activities) or which would be high reward to do first (= those with a high operational cost). in addition, this may help you build up vertical slices of your infrastructure that's moving, and reduce business disruption. for example, you might discover that for a particular activity, you don't need to migrate the whole database but only a few tables, or certain columns in certain tables. then for another activity you may need to migrate additional tables etc.

not sure if this maps to your case or not, but if it does, that's the approach I'd try. 

--
You received this message because you are subscribed to the Google Groups "Impact Mapping" group.
To unsubscribe from this group and stop receiving emails from it, send an email to impact-mappin...@googlegroups.com.
To post to this group, send email to impact-...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Rob Harrison]

Thanks for the reply Gojko, I did start to experiment with the map and in fact landed on a version very similar to how you describe, so that's massively reassuring, thank you.
Already the map is starting to show a lot of good info that the excel wasnt(at least visually), and coupled with the use of mindmup is shaping up nicely so far.