Greg Bloom, Open Referral, on co-creating data standards.

FabRiders’ Network-Centric Resources Project supports the development of people-powered and participatory resources that establish assets for networks and communities. Our past online discussions have focused on developing content with and for communities and network’s, such as Soraya Okuda and EFF’s Security Education Companion and Heather Leson with IFRC’s Data Playbook, or building an event in collaboration with 2000 individuals with Sarah Allen on MozFest.

On February 20th, we had an online discussion with Greg Bloom from Open Referral on how you motivate health and human services providers to share information. The main takeaways from the conversation:

  • To solve problems around sharing information, it’s less helpful to focus on building a great tool and more helpful to focus on cultivating a healthy information ecosystem.
  • To build a healthy information ecosystem, you need:
    • Data standards
    • Open (interoperable) platforms
    • A community of practice
    • Sustainable business models
  • Standards can help foster cooperation by helping stakeholders achieve their own objectives while aligning around common goals. But they take work!
  • Open Referral was able to establish itself as a ‘trustworthy’ community in the face of skepticism because it was clear about its values and principles – in language that would resonate with all stakeholders.
    • The values (i.e. what Open Referral works to bring into the world): accessibility, reliability, interoperability, & sustainability.
    • The principles (i.e. how Open Referral does its work): open by default, users prerogative, learning, and kindness.

View the online discussion

You can also

The Raw Notes

When Greg worked for ‘Bread for the City’ in DC, he came across the problem of how do you see the social safety net? While Bread for the City has their own services, they also kept a database of other service providers in the area in Access. They would maintain the database every Friday afternoon. People were coming to Bread for the City, simply because they had the best information on where they could go to get help in DC. This ended up adding to their workload. Other organisations would come to Bread for the City looking to get access to the data, and while they were very open to giving other organisations access, this seemed crazy to Greg. They were making copies of their database and then these copies would need to be redundantly managed.

Bread for the City was not alone, many service providers were developing and maintaining their own databases of local services. It was a very fragmented field. There were also emerging web-based startups that were scraping information off the internet and then reselling that information at a premium.

All these organisations have different tools to collect all the same information in redundant fragmented silos. Everyone is tackling the problem on their own and it becomes a competitive market failure. It’s hard to understand what exists, where are the gaps and how to make improvement. It’s hard to innovate without the data.

Greg had been given advice over and over again to just scrape everyone’s data and build a ‘killer app’ that everyone would use – solve the problem by winning the market. But that’s the same logic that got us into this mess; it won’t get us out.

Instead of building the best system – Open Referral is working to develop healthy information ecosystems. This entails the development of infrastructure, which is hard to get resourced because it’s hard to explain and ‘see’ – if infrastructure is working right, almost nobody realizes it’s even there.

The components of Open Referral’s vision:

  • Data Standards
  • Open Platforms
  • Community of Practice
  • Sustainable Model
The goal is to prove that this is not inevitable.
We do that through cooperation in the face of competition.

You solve collective action problems by re-aligning competitive incentives into cooperative ones. It’s not about you creating the ‘best’ standard that ‘beats’ all of the others, but rather developing tools that enable interoperability across this fragmentation.

In 2013 proposed a schema for civic services services that was accepted by the W3C as a standard for publishing data on the web. It was good timing: Greg had just published an essay about this problem in Code for America’s Beyond Transparency book. Greg managed to bring the authors of’s civic services schema, along with Code for America and other startups, into the same conversation as A.I.R.S and others in the conventional call-center field. Through this conversation, he got them to agree that in principle that all these different technologies be able to talk to each other. Once they agreed, Open Referral set out to build the bridges.

They developed a first draft of the ‘Human Services Data Specification,’ then built prototypes that would use this format to organize service directory information. They got technology providers like Purple Binder to engage with and use this format for their projects. Since then it’s been used by many different service providers and iterated and improved.

In 2018 the Board of the Alliance of Information and Referral Systems “move to promote the adoption of Open Referral’s Human Service Data and API protocols”. This means Open Referral is now an industry standard.

There is still the big question: ‘If it takes resources to maintain resource data, but everyone should have open access to it, then how can it be sustainably produced?’ They are exploring how to answer the question, through specific pilot projects.

To get people together in a room – you talk to people, figure out their interests, then what the commonalities are – and communicate that as the reason for bringing them together. Greg would go to meetings and talk to people – he had to prove to people that he understood their interests and their concerns, so just showing up and listening was a major factor in success.

Funders don’t necessarily get this work – but they do want to fund convenings.

Greg learned early on not to talk about “Data Standards” – it carries to much baggage or it doesn’t resonate at all with people. He’s starting to talk about standards again now, because Open Referral is now a standard! But still, the goal is to enable people to find this information wherever they might need it, and the objective toward that goal is to establish interoperability (and accountable systems) and standards are only one tool used in the service of that objective.

For data standards, it only works if everyone is willing to co-operate, but that still has a cost to adopt a data standard – so nobody wants to be the first. Indicators that it was being successful – iCarol – the market leaders for call centre software adopted Open Referral pretty early on. After iCarol adopted it, governments started adopting it, NYC was a big one.

And now some funders have begun to require service providers’ information to be published in this format.

We also talked about Shine Theory and Lizzo. Open Referral succeeds if others succeed. “Find the people that want to cooperate and help them win”

If the provider data gets out there – there is concern among information intermediaries that untrained people might do the wrong thing with the data, and this is one of the justifications for not publishing data to be more accessible. Greg’s answer to this is that the better path is to improve the quality of the information and promote better design with it, rather than to try to gatekeep it from the public.

This is public information about services.

Centre for Humanitarian Data had created a standard called HXL

Open Referral is trying to build guides and share learnings on how to get this started in communities and with governments outside the US.

That shared value proposition – have a sense of where you want to go

How can we demonstrate the value of this work, with the least amount of work, to secure resources to get the work done.

One of the first things Greg did when he was starting Open Referral was to write down their values and principles.

The values are: accessibility, reliability, interoperability, sustainability.

Principles: Open by default. Users perogative should guide our work, our work should be focused on learning, and treat everyone with kindness and respect.

These serve to draw a border around what they are trying to do. It proved tactically important as it helped people understand what they were trying to do.

IF you have a network-centric resource that people want to share, you have to put boundaries around it. Here’s what’s inside the resource, here’s what falls outside.

He feels the era of open-data as a strategy in and of itself, as something that we should assume is inherently good – that era is over. You can’t assume that just because you are working on open-data, something good will happen. That’s now insufficient as a working model.

We get better results when people cooperate with each other. Also deliberation and discovery – we are creative agents, we can find solutions.

Everyone actually has these capabilities but our culture papers over them and says cooperation and creativity isn’t possible. That you have to be an entrepreneur to have value in creativity.

Open Referral’s work tries to show that cooperation is not only possible but generally the best strategy.

We still need to speak to organization’s bottom line. We just need to show that the best way to address the bottom line is also often by doing the right thing.