Agile Estimating has an EVIL side effect

If you’re in an Agile team, and especially if you’re doing Scrum

Chances are that you’re in the habit of estimating work.

And (the process of) estimating has much to commend it.

But, it has a nasty side-effect - borderline EVIL.

And it’s hiding in plain sight.

These - as I’m sure you know - are planning poker cards.

I first came across them in 2012 - when I joined my first Agile team.

And because it was my first Agile experience, there were all sorts of “goings on” that were new to me.

But these estimating sessions stood out for two reasons:

They were long - as a developer, I wasn’t used to spending 4 hours plus in a meeting Our estimates were all over the place.

We had 1 pointers taking an entire Sprint...
20 pointers that were done with a 5 minute config change...

It seemed like accurate estimates were the exception rather than the rule.

It didn’t take me long - maybe two or three Sprints - to draw my first conclusion:

“Estimating is a waste of time”

As time went on, our estimates improved… not at all.

But my appreciation for the process did. You might say that estimating went up in my estimation!

The estimates may have been way off, But the discussion we had in those estimating sessions was good.

Excellent, actually.

I was forced to draw a new conclusion:

“Estimates are rubbish… … BUT the process of estimating is valuable”

And it seems that you agree… (with the second part anyway) Here’s what some of you said in a quick survey a few weeks ago: In retrospect, estimation has almost always been a waste of time. Talking about the work while estimating was the real value, but we could have done that in any context, not just time estimation.

Estimating is often a waste of time, but on the other hand, it has the potential to create a good discussion about what a ticket is really about (scope) and if it can be divided into separate steps (split) ← MOVE

I see it as a tool to discuss, ensure that everyone is aligned on the complexity / risks on a story. The act of estimation often opens the floor to a discussion that might otherwise not have happened in a group setting or at all So far, so… not evil.

For the longest time, I was at peace with estimating sessions. I’d enjoy the discussion… and I’d smile to myself as an (essentially random) Story Point value was typed into Jira.

But I’m not smiling any more.

Much like Pandora's box, the trouble starts when those estimates leak out into the world.

That’s when things start to turn rotten.

I got my first glimpse of it later in 2012.

My Scrum Master asked me about a particular ticket. I said it was just about ready for testing, Then she said, “Oh, I thought that would be finished by now. Didn’t you say it was a 2 pointer?”

That was irritating, I won’t deny it.

And when … when someone from “the management” asked a similar question

Well, that went WAY beyond irritating. I was furious!

What strange alchemy is this, where an estimate is transmuted into a commitment?

Seems I’m not the only one to have experienced it. Again, from the survey:

The estimate is sometimes treated as a promise. Then disappointment follows when it is not met.

Seen as promise, so frustrating that nothing is started anymore

if the estimating does not match the real invested time, some people get mad ;)

It's never accurate and its used as a noose by business folk to try and hang teams

It's great to have an idea of the size of the work, but the estimate is held as "the word of God". Then problems arise when reality is different than the estimate.

Mostly not positive. It highly depends on the how the estimation information is used. Worst case when estimation are used to set hard timelines disregarding the assumptions and knowledge consider which getting to them [did I have another?]

It is abused by management to punish teams when they 'guess' wrong. It is misunderstood by teams who seek to deliver story points not finished stories or value to the customer. It wastes time as teams try to re-estimate half-done stories.

I’m forced once again to revise my position. Estimates are not rubbish. They’re toxic - evil, even.

Which has led me to look at whether we can do without estimates. But what has held me back is the process, the conversation. I think there’s a baby and bathwater danger here/ Throw out the evil, and we also throw out the value.

So about 3 years ago I started looking for an alternative to the “estimating conversation”

And I think I’ve found it. We’ll be talking about that… in the very next episode.