Learning to scale up Local Welcome
Local Welcome started in 2015. We were trying to bring refugees and local people together to get to know each other. After several false starts we hit on the idea of meals where everyone cooks and eats together.
The initial work was small-scale. One-off meals in random locations. Mad dashes to get people together. Learning to run great meals.
In 2018 we got funding from the The National Lottery Community Fund to expand our meals. This money was to experiment with using human-centred design, a multi-disciplinary team, and lean and agile methods to scale up across the UK.
In February 2019 we opened our first Local Welcome group in Cardiff. We got some things right, but we got lots of things wrong. Through 2019 we opened groups in Birmingham, Thornton Heath, Derby, Liverpool, Wakefield, Glasgow, and Belfast. Each time we learned a little more.
Now, 12 months on, we have learned three important things about scaling up our approach to Local Welcome. They are about:
We have more to learn. We’ll make more mistakes. But the Lottery money gave us time to learn how to systematise and scale our approach. We hope that sharing what we’ve learned might be useful.
1. Launching new groups in new locations
We are now confident that we know how to launch a Local Welcome group in a new location. We did this 8 times in 2019. None of the locations were in places we knew well. This was deliberate!
The launch process for a Local Welcome group is:
Identify a suitable location
Find an appropriate venue
Run Facebook ads to generate interest
Set up partnerships to find refugee guests
Run meals to recruit diverse and capable leaders
Turn out paying members to the meals
Hand over on-the-ground work to leaders
Grow the number of meals until we’re running weekly.
These steps look obvious with hindsight! Trust me they were hard-won. Each involved a huge amount of experimentation, failure, learning, and - eventually - finding something that worked well enough to use.
The most important things we learned was how much it costs, how much effort it requires, and how long it takes. This means we now have more realistic plans for how we’re going to expand to 80 groups.
The hardest thing we learned was how to stop sending one of our team to each meal. It took months to get all the tacit knowledge that was in our heads and communicate this properly to local leaders.
Now we’re ready to launch lots more groups. We’re also ready for continuous improvement because we know which parts of the launch process aren’t working perfectly and are itching to fix them.
2. Planning our work using lean and agile methods
The Lottery money, and the flexible conditions of how it was released to us, helped us learn to plan our work using lean and agile methods.
‘Lean’ and ‘agile’ are horrible jargon. I thought about not using these words. But part of the Lottery funding was about experimenting with lean and agile methods in the charity sector. So here is what they mean to us:
‘Lean’ is a way to organise work to minimise waste and maximise productivity. We do this by giving every member of the team the power to make improvements to the way things operate. For these improvements to be helpful, everyone on the team needs to understand the way the whole organisation operates.
‘Agile’ is a way to organise work to do a tiny bit, see what happens, learn from it, and then go back and change it for the better. We focus on making the most important things good enough (but not perfect). For this to work, everyone needs to know what’s most important, what’s not working, and what’s ‘good enough’.
One part of working like this is that our whole team makes plans together. We all need to know where Local Welcome is going and how we’re trying to get there. Otherwise we can’t make good decisions on our own. At the end of every four month phase we do these things together:
Phase retrospective (a ‘retro’) - reflect on what we’ve done, celebrate the successes, work out what’s been going wrong
Roadmapping - re-plan the next two years from our current position based on what we know (based on a Jamie Arnold post)
Objectives and key results (‘OKRs’) - set clear priorities and measurable indicators for the first 4 months of the roadmap
Phase kickoff - discuss and prioritise the work we’ll do to meet these objectives (based on Pete Herlihy’s inception workshop)
Another part of working like this is making sure we encourage, receive and act on feedback from our leaders, members, guests, and partners. Otherwise we won’t know what’s not working. We do:
User interviews with leaders before they sign up
Debriefs with our leaders after every meal
Automated feedback surveys for members and guests
Regular phone check-ins with partners
In-person reports from our team members
Periodic depth research and analysis into our impacts.
The most important thing we’ve learned to do is a weekly debrief on a Monday morning for the whole team. We go through the feedback from the weekend’s meals to analyse feedback and agree changes together.
The hardest thing we’ve learned about lean and agile methods is how to measure impact. Lots of Local Welcome outcomes are slippery to measure. We didn’t start measuring until we had 6 groups because we wanted enough valid data. This left little time to measure our impact before we needed to write new funding applications. Stressful.
We are a small team. Using lean and agile methods helps us plan work together and respond to feedback as a team. We think it pays off when our team performs as more than the sum of its parts.
3. Building internal systems to support our work
Finally, we learned how to create, document and operate complex internal systems. This will allow us to scale up the number of groups we run without scaling up the team at the same rate.
One of my favourite quotes is from a systems thinker called John Gall:
“A complex system that works is invariably found to have evolved from a simple system that worked. A complex system designed from scratch never works and cannot be patched up to make it work. You have to start over with a working simple system.”
John Gall, “Gall’s Law”
At Local Welcome, we started 2019 by trying to design complex systems from scratch. Gall’s Law was right. This did not work:
We made an elegant automated workflow to recruit leaders. But it recruited no leaders because it didn’t speak to their needs at all.
We started by asking leaders to recruit 7 friends as members. But it turned out that almost no-one had 7 friends they could co-opt.
We launched refugee tickets using email addresses for booking. But refugees rarely have the internet, let alone email addresses.
We learned from these failures.
Now we start with the simplest system that can be operated by one of the team. This forces that person to understand the people that use it by coming into direct contact with other humans and the context of use.
When the simple system is working well - in the sense of achieving its goals, not in the sense of being highly efficient - we document it in Notion. (Notion is incredible for a team who are learning as they go).
When it’s documented in Notion it means that any of us can do it. This is great for when we go on holiday or are off sick. It’s also great for thinking about which things we can automate to become more efficient. Eventually we automate most things. Otherwise we can’t scale up.
After doing this over and over again, Local Welcome has now evolved a set of simple systems into a complex system that works. We operate:
Weekly tasks centred on running meals, monthly processes tied to arranging meals, and ad-hoc checklists for launching new groups
User journeys that take place over several months involving contact with multiple members of our team
Automated tasks in Hubspot and Zapier that now do in seconds what used to take us hours as humans
Manual processes documented in Notion
Dashboards and alerts to keep us informed about what matters.
The Lottery money enabled this because it understood that we couldn’t design this complex system upfront. Instead, we had to start from simple systems and then evolve them into a complex system.
We are now ready to scale up to 80 groups
Our funding from the National Lottery Community Fund helped us systematise our approach to Local Welcome in three big ways:
We know how to reliably launch groups anywhere in the UK
We’re running a team that is more than the sum of its parts
We’re operating a complex system that changes as we learn.
It took us all of 2019 (and a little bit of 2020) to learn these three things. It’s only the start of our journey if we’re going to grow to 80 groups and get to self-funding by the end of 2021. There are huge challenges ahead.
We wouldn’t be here without the way that the Lottery funded us. Working with agile and lean methods doesn’t lend itself to traditional funding models based on known outcomes that are fixed to points in time. It took a lot of bravery for the Lottery to fund us without those things.
We are thankful. And so are our leaders, members, and refugee guests.
Say hello on @myddelton. Posted originally on the Local Welcome blog. This post started life as an answer to one of five questions posed to us by Cassie Robinson. We’ve been meaning to share more about what we’ve learned over the last 18 months. These questions helped with that.