Over and over again I find teams and organisations that have a chronic problem with ‘estimation’. It is one of the biggest causes of disharmony and mistrust in organisations I have worked with. Many tout ‘agile estimation’ as a process to help address this and in some aspects it is helpful, yet many teams still struggle.
But it doesn’t have to be so, here are 3 handy tips that can help you. But before you read any further, I’m going to need you to do something – you and all the people seeking estimates and the people doing the estimating.
You have to get together and agree : For what reason(s) are you estimating?
If it is simply to have a number to go into a spreadsheet that you can then use in the defense or prosecution case then stop and go here, you will get far more satisfaction.
If it is to identify risks and inform a collaboration into whether what you are trying to build is worth taking those risks for then read on.
1. Start Estimating Early
Figure what the minimum you need to start the conversation and then start it. I have found overwhelmingly in over 100 teams I have coached, teams asking to be involved much earlier on and many times the conversation starts with a one-liner and a sense of value. Then teams and the business can explore the risks together and collaborate on ways to minimise them.
The trouble I see most often is that information is not brought to the people who can best use it until the item becomes very important and increasingly urgent. This is the zone no one wants to be in. It adds pressure to the system that hinders collaboration and strains relationships.
2. Elaborate and Estimate Often
Big bang estimation sessions are motivation killers and reach a point of diminishing value around the 90 minute mark – much sooner if there isn’t the information available for a meaningful conversation. Meaningful information is the key.
If we have nothing new to talk about then let’s not waste everyone’s time talking about it.
The focus is on relentlessly identifying what we do not know about risks that we can reasonably discover without actually building the thing. Then we go find more information about those risks – and in turn discover more about what we should be building, even how we should be building it.
Then we get back together and based on this new information we either know enough to quantify the risks and a sense of time or we have more stuff to go find out.
Most teams that I know getting great results from regular estimation – although they call it elaboration or grooming – are doing it 2-3 times week for no more than 1 hour. For some, they do it right after their daily stand up for 15- 30 minutes.
3. Estimate Quickly
Time and again I find groups sitting mindlessly together, everyone individually frustrated at being in an estimation ‘meeting’ that is wasting everyone’s time. This is often a signal to me that the conversation has gone on far longer than it was useful. Of course some things need long engaging conversations, intertwined with times of contemplation where people can think about what has been said – but if you’ve been in those conversations you realise that everyone knows it is valuable. I’m not talking about those.
The trick – if I can call it that – is to be quick. For each item you are estimating, start by asking “what don’t we know about this that we need to know”. There are 3 possible responses to this:
We know all we need to know
Then go ahead and estimate using whatever technique you choose to derive your number (planning poker, affinity grouping, bucketing etc). Once estimated, move on.
We don’t know what we don’t know
Consider building a prototype to kickstart the questions or visit a customer or do a gemba walk. Do something to get to the first or last response. If you’ve tried everything and you cannot proceed, don’t build whatever you were asked to estimate. If you cannot find the answers to reasonable questions now, how might you fare when you have unreasonable questions and the meter is running? Decide what kickstart activity you will do to answer the question “What don’t we know that we didn’t know”, plan how to start (remember you have nothing else to talk about!) and then move on.
We need to know…
Write down as many things as people say, then quickly frame them as questions – using ‘What/How/Can’ – then prioritise them in some logical order.Then take between 3-5 from the top, figure out who will go hunt for the answers and then move on.
Whatever the case try not to spend more than 10 minutes per item. If you find yourselves approaching this limit, stop and ask yourselves is there any more value in carrying on.
I believe all estimation is waste – some might even be necessary, by preventing even greater future waste.
Did you find these tips useful? I know estimates and estimating are a controversial and potentially scary subject – how do you estimate, what are your stories. I’d love to learn and share.