Community Action MK: Week 4 — Co-production
Another week in the project and another week filled with team-discussions, explorations, and zooming into how we can structure the data we want to collect, and which tools will be best placed to do it. A lot of internal work in organising the data before we chat to our groups is required.
For Mosaic, it has been another week of Angus talking to our users (the local Voluntary and Community Sector groups in Milton Keynes — see more detailed update from Mosaic below). We wanted to say huge thanks to those of our local groups that have taken time to speak to Angus and give us their insights about their needs and ideas — including, MK CIL, Carers MK, MK Dons, Transition UK, MK Mission Partnership and The Bus Shelter MK (and our CEO, Clare Walton, who has shared some key information with Angus, about the strategic cross-sector work she has been involved with, including the Health & Wellbeing Board, the Women’s Partnership and the Vulnerable Person’s Group — among many others). There will be more opportunities to feed in to the project for more groups in the future, so thanks in advance to the voluntary and community sector in helping us develop this work).
So, about the d a t a…
We spent a considerable amount of time modelling the data we want to collect, drawing on our previously used methodology and tools, including the QuickChat app and the subsequent dialogue collection form (a Google form), which we used to record feedback from individual members of the local community/residents of MK. We listed the users we hope (and are feasible) to engage with as part of the 10-week Catalyst project — mainly the organisations that are part of our Collaborative Action Groups (CAG) network. These groups have already been working closely together, sharing knowledge and grassroots intelligence from their work ‘on the ground’ since the CAG initiative was set-up at the start of the pandemic and have been really engaged in wanting to contribute to the creation of a joint data collecting tool for our local sector that will enhance this work.
We spent time reviewing and updating the way we categorise our groups and the areas of ‘lived experience’ that we want to capture, and we considered the most relevant themes and key-words that can be included in the tool. This is to help us bring together the breadth of the data that is collected by the sector (which is so diverse and multi-layered) — to not only make a sense of it, but also to be able to, in the future, generate robust reports representing a wider picture of the issues and needs being experienced by our local communities in MK. We have to be completely honest in saying that this has been a challenge and, we feel, there is still a long way to go before we figure out the way to bring ‘all the pieces of this data puzzle together’. But that is why having a co-productive approach and talking to so many different types of local VCSE groups has been so helpful — so far each one of those we have spoken to has made us realise about a new aspect/ element of the data that is important/ beneficial to record.
This exercise has also made us realise that it is closely linked to another piece of work we are undertaking at the moment, which is around updating and cleansing our database of local groups (we have around 1330 on our records). And it has also made us think about the links we have recently made with the NHS-led MiDOS tool/database linked to social prescribing. We want to reuse and join-up some of the data methodology that we are working with in both our database and the Catalyst project. More updates bringing these projects together to follow…
…and here are this week’s reflections from Mosaic, the project’s digital partner/ lead….
This week has had two main workstreams:
- Continuing interviews with organisations to try to enrich our understanding of the user-base, expectations and requirements.
- Ramping up the solution design
Taking the interviews first of all, some consistent themes are emerging. People like the idea of the project. The success will depend upon a combination of:
- The tool being stable and working smoothly
- A significant up-take by multiple organisations
- Frequent submissions of data my multiple organisations
- An ability for organisations to see what other organisations are submitting
- The tool to be efficient in its collection of relevant data
On the technology side, we’re trying to design a solution that ticks the following boxes:
- Provides all the functionality that the end-users have identified as ‘must-haves’
- Uses off-the-shelf components where possible
- Has minimal on-going fees (i.e. off-the-shelf components are free or cost very little!)
- Allows the Community Action: MK team to manage as many aspects of the data & process as possible without recourse to technical specialists
- Is able to support future expansion cheaply & easily. This could include the expansion of internal data, the ability to connect to external data sources, the ability to bolt on additional functionality, or to programmatically work alongside additional components.
When you look at this list, we’ve got to build a dream solution — system that does everything required, won’t cost much to maintain, & requires minimal, if any, technical support! Obviously, no such solution exists off-the-shelf, so we’re going to use a number of technologies to achieve this:
Airtable as the core data repository.
This is not as obvious as it might seem. Airtable seems pretty good at solving simple problems, but when we looked at the functionality we needed, there are a number of areas where it falls down:
Pros
- Easy for non-technical teams to get started
- Provides an easy interface for administrators to manage & edit records
- A RESTful API is included
- The Airtable administration screens allows for rudimentary views of the data per table
- There are 3rd party tools that provide form interfaces & reporting functionality
- Hosting & Base support is provided
Cons
- The included front end form functionality is limited & does not support inline javascript & additional validation
- The API is limited to using predefined filters to sort or group single records, i.e. can’t supply SQL queries
- The API is limited to 5 requests per second
- The API has only a limited set of methods
- Using the .js client library means the API key is exposed in-browser, which provides access to the entire BASE.
- Some of the front end plugins might also require a Zapier account to connect to AirTable
The best aspect to Airtable is the ability for non-technical people to maintain data. The interface is pretty good, better than most database interfaces. So this will allow the Community Action:MK team to go in whenever they like and amend the themes & categories as they see fit.
We’re going to implement an AirTable interface layer that will handle all of our interactions with it. There are a couple of open-source implementations of this already, so we’ll take a look at these and see what issues they’ve encountered. Having this server-side means that the API key won’t be exposed to the public.
We’re also testing whether we can build an app within AirTable to allow the team to perform the QA function, i.e. review newly submitted records & amend them if necessary.
The Airtable REST API means we can use our own front-end forms, with decent validation & a helpful interface, for data submission. We have looked at other 3rd party forms, but we only require one, which is quite specific, and having our own means that there’s one less dependency (read: point of failure that’s not within our control).
We also see the API being the way to being able to bolt on 3rd party reporting modules, if charts/graphs/maps are required for e.g. KPI reporting.
The API is quite limited so we’ll need to develop some other scripts within AirTable to help manage the data and the form fields, which will probably include some regular cache-ing as well to overcome the 5-calls-per-second limit.
AirTable is new to us, so we’re spending a few days just checking that we can do all of these things … !
That is all for this week and we’ll be back next week with more updates.
Word of this week — CO-PRODUCTION.
