Estimation — Effort vs. Complexity

Marcel Stieber
Nov 3 · 4 min read
Complicated or complex?

In my first article about estimation I explained that it should not serve a self purpose but is a tool to start conversation in the team. If you haven’t read it yet, I suggest to do this first to be on the same page here. I experienced it several times that there are discussions about what to estimate. Effort or complexity. Having in mind what the purpose of estimation is this question is actually not fundamentally important. It’s a question of purpose and the root. If you only estimate to get everything clarified and everyone on the same page it might not be needed at all. But if you want to use the number for your planing, for reports and for improvement you have to ask yourself what you want to achieve here.

Using the estimated numbers for you sprint planing means you want to use numbers fitting to a time scope. Easiest way to go here is to estimate for time effort as this fits easily right away. Estimating complexity will not help in a short term, as you can’t say that you are able to handle 13 complexity in two weeks. Complexity itself is simply not time related. Also, having one story with 13 points is not equal to having 13 stories with 1 point. And even not equal to another story with 13 points, if you refer to the time needed. This will only be helpful if you have enough sprints as reference to derive an average to use.

Going with effort is a good fit if you do not have a lot of sprints with a stable team. It’s easy to establish and to understand. If you go with something like “3” means one day of work, everyone will immediately understand this. On the other hand you could stymie yourself, because you committed yourself to a time. Depending on stakeholder communication this could be an issue. But frankly, if this will cause any issues, the root cause is not the estimation. Comparing this approach to relate numbers to time reveals that complexity is also a time based estimation in the sprint context. You will still plan story points for a specific time scope.

Coming back to the duration of establishing complexity estimation. The need of several sprints to create a foundation is not only valid for the average sprint limit but also for the set of reference stories that will help you to classify new stories and put a number to it. This again needs a stable team setup. Otherwise your reference stories will be outdated pretty fast as new team members will not know them. It’s the circumstances that are crucial to decide if you go for complexity or effort.

Putting it together you can decide between a tool that needs a stable team setup over time to be established properly, a tool that gives you an emergency exit as the number is only correlating with time or an easy and immediately established process which could commit you to your time prediction. But no matter what you choose: Estimation will not prevent from surprises. There will always be things you have not seen coming and thing that are not working as expected. So no silver bullet here. Same for the guarantee that the effort will not blow up. It’s not the guarantee to deliver on time or to achieve a better predictability. And it could definitely not be used to communicate numbers to stakeholders. The only thing we are doing here is labeling the output of our conversation. This could be a number or even a color. And at the end it still is a best guess which will get fuzzier the further you look ahead.


Conclusion

It is a matter of preferences and circumstances what should be estimated. More interesting is WHY you are estimating. It could be stakeholders like marketing or dependencies to other teams. So there is most likely a proper request and estimation the weapon of choice. But it means following this request blindly and trying to give a prediction how long it will take. It’s better to turn this around and ask why stakeholders need to know a specific date? Maybe marketing needs to prepare a campaign? So, how long do they need to prepare? Three weeks? Four weeks? So wouldn’t it be working to give them a ping if we are sure we only need four or five more weeks to finish? Or it is the team itself which needs to plan resources more proper. The wording of the question is already pointing to planning and not estimation. It’s a question of preparation and not labeling. Only with a stable foundation estimation could help you. And: do not over-complicate. Too many rules and references are an overhead that distracts from the base task. Deliver a good product to your customer.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade