A distributed approach to co-operative data

[Summary: rough notes from a workshop on cooperative sector data.]

Principle 6 of the International Co-operative Alliance calls for ‘co-operation amongst co-operatives’. Yet, for many co-ops, finding other worker owned businesses to work with can be challenging. Although there are over 7,000 co-operatives in the UK, and many more worldwide, it can be challenging to find out much about them.

This was one of the key drivers behind a convening at the Old Music Hall in Oxford just before Christmas where cooperators from the Institute for Solidarity Economics, Open Data Services Co-operative, Coops UK and Transformap gathered to explore opportunities for ‘Principle 6 Data’: open data to build up a clearer picture of the co-operative economy.

We started out articulating different challenges to be explored through the day, including:

  • Helping researchers better understand the co-operative sector. With co-ops employing thousands of people, and co-operatives adding £37bn to the UK economy last year, having a clearer picture of where they operate, what they do and how they work is vital. Yet information is scarce. For researchers at the Institute for Solidarity Economics, there is a need to dig beyond headline organisation types to understand how the activities of organisations contribute to worker owned, social impact enterprise.

  • Support trade between co-operatives. For example, earlier this year when we were planning a face-to-face gathering of Open Data Services Co-op we tried to find co-operatively run venues to use, and we’ve been trying to understand where else we could support co-ops in our supply chain. Whilst Coops UK provide a directory of co-operatives, it is focussed on business-to-consumer, not business-to-business information.

  • Enabling distributed information management on co-ops. Right now, the best dataset we have for the UK comes from Coops UK, the membership body for the UK sector, who hold information on 7000 or so co-operatives, built up over the years from various sources. They have recently released some of this as open data, and are keen to build on the dataset in future. Yet if it can only be updated via Coops UK this creates a bottleneck to the creation of richer data resources.

My Open Data Services colleague, Edafe Onerhime, did some great work looking at the existing Coops UK dataset, which is written up here, and Dan from ISE explored ways of getting microformat markup into the Institute for Solidarity Economics website to expose more structured data about the organisation, including the gender profile of the workforce. We also took at look at whether data from the .coop domain registry might provide insights into the sector, and set about exploring whether microformats were already in use on any of the websites of UK co-operatives.

Building on these experiments, we came to an exploration of potential social, organisational and technical challenges ahead if we want to see a distributed approach to greater data publication on the co-op sector. Ultimately, this boiled down to a couple of key questions:

  • How can co-operatives be encouraged to share more structured data on their activities?

  • How can the different data needs of different users be met?

  • How can that data be fed into different data-driven projects for research, or cooperative collaboration?

There are various elements to addressing these questions.

There is a standards element: identifying the different kinds of things about co-operatives that different users may want to know about, and looking for standards to model these. For example, alongside the basic details of registered organisations and their turnover collected for the co-operative economic report, business-to-business use cases may be interested in branch locations and product/service offerings from co-ops, and solidarity economics research may be interested in the different value commitments a co-operative has, and details of its democratic governance. We looked at how certifications, from formal Fairtrade certifications for products of a co-op, to social certifications where someone a user trusts vouches for an organisation, might be an important part of the picture also.

For many of the features of a cooperative that are of interest, common data standards already exist, such as those provided by schema.org. Although these need to be approached critically, they provide a pragmatic starting point for standardisation, an example with Coops UK Co-Operative economy dataset can be seen here.

There is a process element of working out how data that co-operatives might publish using common standards will be validated, aggregated and made into shared datasets. Here, we looked at how an annual process of data collection, such as the UK Co-operative Economy report might bootstrap a process of distributed data publishing.

Imagine a platform where co-operatives are offered three options to provide their data into the annual co-operative economy report:

  1. Fill in a form manually;

  2. Publish a spreadsheet of key data to your own website;

  3. Embed JSON-LD / microformat data in your own website;

Although 2 and 3 are more technically complex, they can provide richer and more open and re-useable data, and a platform can explain the advantages of taking the extra steps on these. Moving co-operatives from Step 1 to Step 2 can be bootstrapped by the form driven process generating a spreadsheet for co-ops to publish on their own sites at a set location, and then encouraging them to update those sheets in year 2.

With good interaction design, and a feedback loop that helps validate data quality and show the collective benefits of providing additional data, such a platform could provide the practical component of a campaign for open data publication and use by co-ops.

This points to the crucial **social element: **building community around the open data process, and making sure it is not a technical exercise, but one that meets real needs.

Did we end the day with a clear picture of where next for co-op sector data? Not yet. But it was clear that all the groups participating will continue to explore this space into 2016, and we’ll be looking for more opportunities to collaborate together.