Weeks 2 and 3: depth and scale
The last two weeks have been full on, hence a bumper two-week edition of weeknotes. Lots of stuff happening:
- user research
- a digital survey
- flyers
- driving traffic to alpha.gov.bb
- polling
- a roadmap
- new heads of product and delivery
- some snow and skiing
All of that is against the backdrop of Estimates season.
User research
The team spoke to 68 people about their experiences using government services. That’s the first time the team have done research on this scale, and with mostly people who haven’t done user research before.
This research will feed into the roadmap for the next year.
Digital survey
We asked similar questions to our user research. I’m not normally a fan of surveys, but we got some good, helpful data and volume, which confirmed what we were learning in research.
Good learning for me to be open to trying new things.
Flyers and driving traffic to alpha.gov.bb
The survey is on alpha.gov.bb, so we needed to drive traffic to the site, obviously. Visits had been low until now, but we ended up getting 700+ visits a few days in a row, and 490+ survey submissions.
We’ve been using AI to analyse the data. The tool we’re using, Tally, gives us a quick view of the submissions, so by using AI to do the analysis we can sense-check the results.
We’ve been happy with the results, and if we’re 80% going in the right direction, I think that’s good enough for now.
Polling
Last weekend we had a polling company out in Barbados going door to door asking very similar questions again. It’s helpful to get data at scale: they’re going to be speaking to 1,500 people.
That’ll take us to over 2,000 people. Not bad for our first round of research, and it’ll help direct further research in the future.
With the user research and the survey, this now gives us both depth and scale, which we’ve not had before.
Roadmap and 7.5 hours on Zoom
Wednesday was a mammoth day. We needed to magic up a roadmap for Estimates. We’ve been over-indexing on delivery lately and haven’t had the data to feed into a roadmap process.
But now, with the data from the user research and survey, and a list of 2,220 service candidates of varying quality, we used AI (Claude), working in co-working mode, to build a roadmap across 15 themes for us to work on.
Two important caveats here.
One: we went into this with our eyes open, knowing that it could, or would, be wrong. Much of the 7.5 hours on Zoom was spent checking and re-checking the results.
Two: we know we’ll need to validate all of this, but going into Estimates without a plan for how the money would be spent just wasn’t going to fly.
This wasn’t about being right as much as it was about being credible.
And we know we’ll have to change this as we find new information, especially as we don’t have any better data. I think we got to a point where we’re confident in the results. All of this will change anyway. That’s just how the game works.
New heads of product and delivery
At the same time, we’ve welcomed two new key hires to GovTech: a head of product and a head of delivery. Both have got up to speed quickly.
Outside of work
Finally, and outside of work, it snowed in the northern Alps last week, the first time since the beginning of December.
Last Saturday was epic. I skied until I physically couldn’t walk any more.
Unfortunately, there isn’t any snow in the forecast for the foreseeable future. Frustrating.
I hope you all have a great week.