Let’s fix project overruns

With better internal forecasting

Big projects are unwieldy, hence the adage, “the first half takes 90% of the effort; so does the second half”. 

What is it worth to know how much you actually get done in a week? What if there was 1 question per team per week? What if everyone answered it anonymously? Would you want to try it out?

We (Nathan and Terry) are interested in collaborating with organisations to implement this. We’ll join you in your kickoff meetings (or sprint planning or for part of a 1-1) and ask you and your team to predict predict whether your work will get done by the next time we talk. Then we’ll see whether over time your team can get better at predicting what they’ll get done, and whether that will be useful for you. Then we’ll help you automate the process.

Bad forecast are a big problem. McKinsey say the average cost overrun of projects is 45% and the average time overrun is 7% (see the graph). Better forecasting is therefore worth a huge amount to companies.

I don’t normally trust consultancies, but honestly here the number could be too optimistic

Why is progress so hard to track? Here are some reason:

1. It’s hard to say ‘No’. People want an easy life so say “Sure I’ll get that done'' even if they probably won’t. Likewise, often it's obvious that everyone understands what needs to be done (until they don’t).

2. Situational awareness gets distorted across organisations. Each management layer makes sense of situations slightly differently and will ‘sell’ its take to the next layer up as best it can. But like a game of telephone, the story changes slightly with each telling. At some distance (in time or space) the story doesn’t match reality.

3. The planning fallacy. People are prone to optimism on project timelines. Perhaps it’s because engineers think they will solve problems better this time, or because managers want to create a sense of urgency, but there is a bias towards timelines that are too short.

4. Discovering something has gone wrong often leads to blame.  This often stops people saying “I don’t think that’s going to happen”

How can team forecasting bring clarity?

First, it works with questions every team member can answer. “Will you complete your tasks by the end of the week?” Yes/ Maybe/ No. It doesn’t take anyone out of their portfolio or need sophisticated forecasting skills. “How surprised would you be if this didn’t get done?”. Most people have a gut answer to that, even if they don’t want to say it. 

Second, this type of forecast gives great feedback. If you discover you are too optimistic, then try a little more pessimistic next time. If you aren’t getting the work done, set fewer tasks. Forecasting isn’t giving you an extra thing to think about, it’s clarifying what you already think about - can I trust Ian to get his tasks done - and giving you a way to track that. Hopefully you can think about it less. 

Third, it draws together perspectives  from across the organisation. It’s not just your team who can anonymously forecast, but other teams too. Many others have information on whether your project will succeed. And perhaps they know that their dependency isn’t going to get done and they’ll say anonymously what they wouldn’t put their name to.

Finally, forecasters do not simply sample sentiment; the test is reality. Are the tasks getting done? Jan Filochowski, who turned around a number of failing NHS Trusts notes that in organisations recovering from failure, actual performance can exceed the perceived performance, especially after months of bad news. 

Why do poor progress estimates cost us so much?

Poor estimates create more work everywhere. Not only is there more work the first time, but other teams are delayed and perhaps you have to go back, discover what went wrong and rework from there. Traditional risk management with contingency budgets can help, unless your team starts to take them for granted or the sales team cuts into it to win the client.

A knock-on, however, is that gap between perceived progress and actual progress is often invisible until some significant milestone has been missed, by which time effort not related to the failure itself may also have been wasted. Delay and delusion are expensive partners in failure.

How might forecasting cut the cost of failure and boost the value of success?

Real-time, blame-free, task-oriented predictions of progress might take project management a big step forward. Transparency means that all levels in an organisation share a common situational awareness, too, even if they ascribe slightly different reasoning to how it came about.

As the forecasts are calibrated against real projects over time, the ability to foresee problems and correct them, or find additional gains will likely cut contingency, shorten project cycles and build internal confidence.

Bottom line

We want to test a platform to forecast your progress week by week. If successful, what you do with the extra time and profit will be up to you and your sales team, but you will see it coming. 

Let us know if you want to use this in your organisation. You can reply to this blog, DM wherever you saw it, or e-mail¹ us on nathanpmyoung@substack.com or terry@datchet.consulting.

The authors

Nathan Young builds forecasting platforms, works with teams of forecasters and delivers predictions.

Terry Young ran innovation projects, taught software project management and consults. He is also Nathan’s father.

1. The authors disagree on the spelling of this word