For my end-year review one of our user researchers asked me to talk more openly about the mistakes I’d made throughout my career. She said it was important for senior people to show that they didn’t have all the answers.

I loved this feedback. It struck a chord with me for another reason too.

I’m about to leave GDS to go and work for the Home Office. I’ve been wondering how to mark this change. So over the last week I’ve been giving this talk to say goodbye.

My GDS best bits

Even at my most open and vulnerable I can’t just post up my mistakes! I’m human after all. So here are the five things I’m most proud of from my time at GDS.

Working as a user researcher on GOV.UK Notify
When I joined GDS I spent four months working on GOV.UK Notify as a proper user researcher. It was the first time I’d ever worked on an agile product team. I loved it. It taught me a huge amount in a short time.
Helping GDS think about service teams as users
I stood up in front of hundreds at an all-staff and talked about the idea that civil servants were our users. It was terrifying. I said difficult things. But people listened. That’s when I fell in love with GDS.
Understanding the user needs of service teams
Given civil servants were our users we needed to understand their needs better. We interviewed 150 teams from huge online services to tiny offline transactions. This work changed how we think about our users.
Exploring the biggest unmet need of service teams
Three things are common to all services - publishing to all (GOV.UK), talking to individuals (GOV.UK Notify) and collecting information (unmet). We ran the Submit discovery to look at collecting information. It’s the best work I’ve ever done.
Building a team of user researchers
The reason it’s OK for me to leave GDS is the team of researchers we built. We took risks and hired non-seniors. We trusted and supported them. Now they’re flying on their own. Could. Not. Be. More. Proud.

My biggest mistakes at GDS

Now that’s all out of the way let’s get down to these mistakes.

#1. Ignoring hypothesis-driven design for too long

I’ve always thought in terms of research questions. I was wary of hypothesis-driven design because I worried that hypotheses are ‘validated’ or ‘falsified’ with too little attention paid to the details. But I’ve learned that hypotheses are a great way to link problem-oriented researchers with solution-oriented designers and developers.

#2. Failing to write up my most important research

After interviewing 150 service teams we presented our results. We got a lot of people excited. We took that momentum and carried it into a discovery. But what I failed to do was stop and take the time to write up those results for others to read later. I get requests pretty much weekly for the write-up. It’s still on my to do list. Huge mistake.

#3. Holding alpha workshops when I wasn’t ready

At the end of our Submit discovery we held design workshops with people from across GDS. We wanted to use their creative thinking to widen our alpha pool. But I wasn’t ready to hold these workshops. Instead of cancelling, I went ahead and ended up talking at these talented people for three hours. Wasting their time. Horrible.

#4. Leaving the Submit team after discovery ended

This one is hard to admit. On the Submit discovery I was the product manager on top of my normal role as lead user researcher. We finished with huge team momentum. But I decided to walk away and go back to my normal job - for good reasons - and I became part of the reason that momentum was lost. Momentum is a precious thing.

#5. Forcing researchers to recruit participants

Civil servants are our users. We can’t recruit using agencies. We can’t pay incentives. The best answer two years ago was to recruit people ourselves. But I failed to tackle this at a structural level. Now our 10 researchers spend too much time recruiting participants. I was scared of the size of the problem to solve. That’s a poor excuse.

#6. Doubling up our researchers without thinking

I introduced a senior and non-senior researcher to some teams. A move that let us recruit juniors, open career progression, respond to new projects, and be resilient when people left. But I didn’t think about how our researchers would get on with each other or split work between them. We’re humans. Humans are complicated.

#7. Failing to record sick leave and annual leave

This one sounds like a little mistake but was a huge one. I pride myself on being a good manager. But I didn’t always log sick leave and annual leave properly for the people I managed. Knowing what I now know, this is unfair to the people I managed and undermines much of my other good work. I’m never making this mistake again.

#8. Talking down the work of our teams in public

My worst mistake to own up to. Part of my job is to critique the product strategy of our teams. This is a natural - if difficult - conversation to have with the team leaders. But there were times where I let my frustrations show to other members of the teams. I upset people I cared about. I burned bridges I needed and they needed. I’m sorry.

#9. Failing to sell ideas to our senior management

The Submit discovery felt unstoppable to me. I thought it responded to a clear user need, had significant cost benefits and opened up strategic opportunities. But I got carried away by being too close to it and failed to communicate these things to our senior leadership. I’m still kicking myself for this. Because I know how to do this.

Other mistakes in my career

In writing up my biggest GDS mistakes I realised something. Many things that I did well at GDS were linked to mistakes I'd made beforehand. These are some of them.

#10. Writing discussion guides at 2am

I did a lot of this at cxpartners. I used to say this was because I wanted to spend time working on the prototypes. But these days it feels more like a control-freak mechanism to ensure I didn’t have to take feedback into account. At its worst, my intern turned up to moderate his first research session to find me still writing a guide.

#11. Researching my designs in a leading way

I used to be both researcher and designer. I was an excellent researcher when I was testing designs made my other people. I found all sorts of problems. But when I tested my own designs they magically performed much better. Yes, I was a good designer, but no, this wasn’t the whole truth. Don’t mark your own homework.

#12. Presenting a huge list of problems to people

When I started out as a user researcher I wanted to hold on to every observation. Partly because I felt this was the way to prove myself. Mostly because I didn’t have any idea how to prioritise. But if you present a huge list of problems to a group they’ll smile, nod, and ignore the whole list. Much better to show them four or five things.

#13. Expecting people to work the same way I work

When I started managing people I was out of my depth. I tried to turn people into clones of me. Leaning over to edit documents. Holding them to timelines I’d set myself. But people don’t respond well to being managed like this. State the goal, let them go, trust them, and give useful feedback. Let them make their own mistakes.

#14. Sharing the wrong things with people

Our society and culture encourages us to be transparent and open. To avoid being ‘two-faced’. I tried this with the people I managed. But it unsettled them when I told them things that were unclear or too far outside their normal world. Now I take this advice from a trusted friend - be judicious with information, emotion, and opinion.

#15. Being angry at others when out of my depth

At cxpartners I was handed an impossible project. I tried to deliver it and spent months being furious with the people who set it up. But what I didn’t realise was that I had the power to say no. That though anger is a useful sign that something is wrong, it’s rarely a useful thing to hold onto for any length of time. I’m still learning this.

#16. Binding my identity to my work too closely

Lots of these mistakes come from linking my self too closely to my work. I worry that saying no will stop people loving me. Once upon a time this was useful because I had loads of capacity. But these days there are far more requests than I have time to spare. Learning that the world doesn’t fall apart when I say no has been huge for me.

#17. Burning out in 2014

I've written before about burning out. I used to think that I could cope with anything work threw at me. But during the impossible project, trying to make a webfont work at 3am, I broke down in tears. I went to New Zealand for four weeks and spent half of it recovering. I did counselling. I had coaching. I learned to deliver the impossible project. It took me six months to regain my confidence. Then I came to GDS.

So, yes, I’ve made mistakes

We all make mistakes. All of the time. Every one of us. It's unavoidable.

It doesn’t matter that we make mistakes. What matters is whether we are able to recognise the mistakes we made. This is hard work for lots of reasons:

It’s not always possible in the moment.
It requires us to take a long, hard look at ourselves.
It means putting aside our anger with others.
It can involve someone being brave enough to bring something to our attention.
It means we need to find space to reflect on things away from everyday work.

After we recognise our mistakes we can change things. Of course we will always make more mistakes. But at least we won’t be repeating the ones we made before.

I hope that sharing this prompts you to think about the mistakes you've made. And then, having done this, to try talking them through with someone you trust. I think you’ll be surprised at how helpful people can be when you do...

Let me know what you think on @myddelton. Thanks to Amber Westerholm-Smyth for prompting me to do this and Mia Kos for once saying we should talk more about our failures. Finally, I want to give a huge thank you to the wonderful people at GDS for giving me the space to make these mistakes and learn from them. I will miss you all.