Winning at Dependency Twister

I’d like to dig into that question more.

Simple, simple example…

GrowthCorp has three high priority new feature initiatives (F1,F2, and F3). A single shared Operations team (Ops) is slated to support these initiatives, along with other mission critical work. As usual, Ops is under the gun… under-resourced, under-trusted, and overworked.

After the quarterly planning dust has settled, Ops creates cards to support the feature work (Support 1, Support 2, and Support 3). There are two boards: Feature Work, and a private Operations board (because when Ops makes their board public, it gets micromanaged.)

At the start of the quarter, it looks like this:

It’s getting near the end of the quarter, and things look like this:

Bar Team hasn’t even started F3, and Ops has finished Support 3. What if Bar’s F3 needs change? Acme Team is waiting around for Ops to finish Support 1. F2/Support 2 is way behind schedule for both teams. Meanwhile, Ops is having to do production support for shitty feature implementations from last quarter (“unplanned”). And…of course…Acme, Foo, and Bar are complaining about Ops. WHEN IN DOUBT BLAME OPS.

Operations leadership has been advocating for headcount for ages, but someone always plays the process and efficiency card.

First, let’s put it all on one board and add Ops as a collaborating team for F1, F2, and F3. We can indicate their tasks as sub-tasks or related stories (not shown).

Next, let’s indicate that at any given time we can only really have Ops working on two items (Doing), and be committed to two items in To Do. Note here that Ops has a spare magnet for a To Do item. We move two of our feature items into Doing (F1 and F2), and the teams work on these together.

Low and behold, the feature issues creep up (“unplanned”). Ops has a working agreement that requires them to shift to production issues. Given that they cannot collaborate on F1 and F2, the feature teams swarm on the unplanned work to help knock that out.

And things keep rolling…

Teams subject themselves to all brands of masochism in an effort to “thread the needle” and eek out every last grain of efficiency. This happens ALL THE TIME with shared teams like Ops, UX, Data Science, Research, etc.

But in the process of trying to do this optimization (often as part of quarterly planning), they reduce transparency, build up planning inventory, increase rework and canned efforts, and institutionalize big-batch planning cycles.

Without a unified view, the shared teams will always struggle to make the case for headcount, or to embed themselves with the feature teams (or tool up, automate up, etc.). Somehow we persuade ourselves that we can get more out of them.

So …. the big point:

Related posts:

No business can work without an effective stock of money. It is the backbone of all associations and the shared factor by which most business execution is estimated both inside and remotely. The…

No es un secreto que uno de los sectores más apetecidos por estudiantes y jóvenes profesionales para vivir son los barrios aledaños al Centro Comercial Alamedas. La cercanía con rutas de transporte…

I got the opportunity to be a part of Google Techmakers Scholarship Program 2017. I am writing this post to share the details and my experience so that we can together come up with a diverse tech…