Navigating the Uncertainty of Agile Estimations

Do We Re-Estimate Split Stories? - Mike Cohn

Someone asked this question at a recent webinar I hosted on user stories: Do we re-estimate the user stories after we split them? The short answer? Absolutely.
When you break stories down during product backlog refinement (or during sprint planning if you’re trying to make a story smaller to fit inside a sprint), the team will want to re-estimate them. And there is no guarantee the sum of the new estimates will equal the old estimate.
That’s OK and expected. Ideally, the estimate will even out over a bunch of stories and splits but it won’t be perfect. Just adjust the burndown to account for the rise or fall in estimated points, and keep moving.
The Product Burndown Is Like an Airplane Touchdown
For a longer explanation, I’ll use an analogy. I flew into Heathrow a few months back on one of my periodic trips to London. I was looking out the window, idly watching the mechanisms on the wing as we made our final descent. It was a windy day, so I was curious about how the pilot would handle the landing.
I know that the pilot follows a glide path towards the runway (shaped much like the ideal burndown line). But from my point of view, we didn’t hurtle down an imaginary diagonal line and magically land in the perfect spot. We descended a bit, then flew what seemed like level for a bit, then got a bit lower.
At one point, when we were close enough where I could see the runway, we were buffeted up by a strong gust. I didn’t panic, and thankfully neither did the pilots. They just made some sort of an adjustment (it felt like they added a bit of power) and resumed their steady progress towards our destination.
Every commercial runway in the world is painted to help give pilots visual cues about their safe landing zone. In vastly simplified terms, there’s a threshold (after that point it’s safe to land) and a touchdown zone (a range of safe touchdown points in 500 ft / 120 m increments). The safe range varies depending on the size and weight of the airplane.
I don’t know precisely where we landed in the touchdown zone, but the pilots greased the landing and got us to our gate 10 minutes early. So despite the literal ups and downs, and the visual estimates they had to make on their way down, they succeeded.
A product’s burndown chart is like that airplane’s approach into Heathrow. It rarely follows a straight line. Some sprints we approach the goal a bit faster than we expect and sometimes we reach it a bit slower than we expect. Occasionally, we might bump slightly upwards and have to descend again.
Predicting exactly when we will touchdown at our product goal is impossible. It’s always a bit of a guessing game. But predicting a date range where we will finish is both desirable and entirely doable. And as with landing a plane, it’s a skill that improves with practice.
The more you estimate and the better you track your progress, the more consistent and predictable your team’s estimates and plans will become. And creating accurate agile plans that others can rely on will definitely help you succeed with agile,

How to connect with AgileDad:

- [website] https://www.agiledad.com/

- [instagram] https://www.instagram.com/agile_coach/

- [facebook] https://www.facebook.com/RealAgileDad/

- [Linkedin] https://www.linkedin.com/in/leehenson/