Is Open Data Product Specification the Swagger of the Data Economy?

Jarkko Moilanen (PhD)
Open Data Product Specification
8 min readOct 23, 2022

--

According to Nordic APIs, here are some numbers from the API Economy 2022: ​​

  • 83% of All Internet Traffic Belongs to API-Based Services,
  • 93.4% of API Developers Are Still Using REST.
  • More Than 93% Of Communications Service Providers Use OpenAPI

Roy Fielding published his dissertation 2000 and described what we now know as REST APIs. In brief it took around two decades for REST to hit over 90% in adoption rate. Of course the adoption of APIs in general has affected the adoption rate. The amount of APIs started to explode around 2010–2011.

What about Open APIs then? Do not confuse open content with the term Open APIs. The openness in the name refers to the standard licensing rather than content of APIs. We who had the privilege to live through the beginning of the API economy remember Swagger.

When the REST APIs started to boom there were multiple competing standards on how to describe REST API in machine-readable format. The three most common formats were RAML, Swagger and Blueprint. The Swagger API project was made open source in September 2011. While these competing products had stronger financial backing than Swagger, they initially focused on different use cases from Swagger, and as of mid-2014, Swagger interest was growing more quickly than the combination of the two others.

As we know now, the descendant of Swagger known as Open API was victorious. In November 2015, SmartBear Software, the company that maintained Swagger, announced that it was helping create a new organization, under the sponsorship of the Linux Foundation, called the OpenAPI Initiative. On 1 January 2016, the Swagger specification was renamed to OpenAPI Specification. Now 2022 more than 93% of communications service providers use OpenAPI.

Data Economy

Data products are now considered to be significant part of the data economy regardless of taking the monetization approach or not. Even in data exchange, which is more about sharing the data across various borders, data products are now the thing. Data product is one of the cornerstones of much hypeted Data Mesh.

If data products is the “REST API” of the data economy which is going to boom the data usage in architectures and business solutions, then what is the “Swagger” of the data economy? Where is the metadata model which is generic enough to cover most common use cases? What are the uses cases it should cover?

One of the emerging candidates for the “Swagger of the Data Economy” is the Open Data Product Specification (ODPS).

The Open Data Product Specification is a vendor-neutral, open-source machine-readable data product metadata model. It defines the objects and attributes as well as the structure of digital data products. The work is based on existing standards (schema.org), best practices and emerging concepts like Data Mesh. The reasoning is that we reuse and proudly copy instead of reinventing the wheel.

The model in ODPS is quite generic and contains elements and attributes to describe data product from business, legal, ethical, and technical perspectives. ODPS is not the only solution in the markets. There are probably multiple rivaling solutions. The situation reminds me of the beginning of the REST APIs driven growth of the API Economy.

Data Economy is about moving the data across various borders. To be efficient in that, it needs a shared understanding of the metadate model used. This is the core reason we need ODPS or alike standard at the moment. Later on, when the data monetization kicks in, we need the standard to enable efficient monetization of data. At this point of the story thoughts were turned into the history of ODPS which has not been written anywhere. It does have some similarities compared to the story of Swagger.

Short history of ODPS

History written in 3rd person so to make it more usable in multiple contexts. Wikipedia article would be nice.

Business tools ignited the development process

Jarkko Moilanen started developing Data Product Toolkit in March 2019 on his own time during evenings and weekends. The toolkit at that point was a set of canvases for business people. The target group of the canvases is the data product managers which are leading the development of data products from business-first perspective. After some rounds with customers Jarkko discovered the need for a standard which defines the product and options for example with pricing plans.

Slow confusing early days 2019

At that time Jarkko Moilanen was working as Chief Development Officer at Platform of Trust and tossed the idea of standard data product model to the data platform development team. The initial commit to the emerging standard took place in 22 Nov 2019 by Jarkko Moilanen.

The team probably had other more acute issues to focus on and standard development was often put aside. Because of this the definition of the standard was hard to understand, lacked a lot of the basic elements, and solid base. It was going nowhere.

This is also when Jarkko started to consult companies on data economy under “Data Product Business.” The monetization angle was not valued in the Platform of Trust context. The business aspect was valued by Jussi Niilahti and Toni Luhti with whom data monetization business has been taken further.

Reborn as Open Data Product Specification 2020–2021

After slow progress with the standard Jarkko decided to “fork” the process in March 2020 and rewrite the whole thing again from scratch on free time. Nothing was taken from the old spec. The spec was taken out from Platform of Trust development.

It’s worth mentioning that the ODPS development started from the business element standardization and then moved on to legal, ethical and technical parts.

The development was moved forward hidden under the radar. Around July 2021 the need to define “final” version 1.0 was becoming evident. The rise of Data Mesh concept pushed Jarkko to kickstart finalization process of ODPS 1.0. Jussi Niilahti was invited to the process to become one of the editors. In Feb 2022 the version 1.0 was published. It’s also important to notice that Toni Luhti had joined the development as data economy professional and also provided some funding for the development. Also Fuug association provided a small grant to finalize the spec version 1.0.

Write a book and take the next step

While finalizing the version 1.0 Jarkko also wrote a short academic article about the specification to the 17th International Conference on Design Science Research in Information Systems and Technology. It was accepted and that article caught the attention of a few players from Fortune 500 company Grainger and Bayer the medical giant among other smaller players.

At this point Jarkko wrote a book called “Deliver Value in the Data Economy” together with Toni Luhti and Jussi Niilahti. The ODPS standard is discussed in the book and tied to larger Chief Data Officer as a Service model.

At this point it became obvious that there is a clear need for the metadata model and this seems to be the right moment. People talking about the specification referred to it as the “Jarkko’s model” and that was the next thing to change. Around here discussions with Agile Lab from Italy started and we discovered that we both have same goals in general and we both have worked on a standard.

Road to Open Data Product Initiative

There was a clear need to institutionalize the standard and development of it. Jarkko contacted old friend Antti “Jogi” Poikola, with whom he had worked in the past. Poikola was oen key player when MyData movement went global. He knew how a good idea is grown into global joint movement.

In June 2022 Jarkko and Antti sat down on Friday afternoon in a terrace in Helsinki for a few beers to discuss the ODPS. Antti shared the story of MyData movement and Jarkko told the history of ODPS and future needs. The need for more piloting around the spec was seen as a must have to show the value and see what needs to be focused next.

Piloting cases under work or in progress

As a result, Jarkko established Open Data Product Initiative in July 2022 to make it possible for the specification to grow and become institutionlized. The ODPI was taken under the wings of open source chapter of Open Collective. Then the first groups were established and first meetings held. The initiative was born and ignited!

Growing interest towards the open data product specification

After the summer ODPS gained more and more interest. For example MaaS Alliance which operates in the Mobility as a Service field wanted to know more and asked Jarkko in their meeting. In August 2022 Jarkko started to collect piloting ideas and have meetings with various organisations to ignite several pilots. These could be seen as use cases which were in crucial role in the success of Swagger. But have we selected the “right” use cases for ODPS? At this point ODPS was also acknowledged as one of the data economy ecosystems in the OPEN DEI report listing the EU data economy building blocks for data sharing.

Becomes part of the European data economy

It has taken now three years for ODPS to see the daylight and become part of the data utilization discussions in Europe and in US. It is now listed as part of the European data economy building blocks. It has it’s own neutral organization. More and more evaluation of the model is ongoing and pilots are emerging. ODPS has come far from an idea in one person’s head 2019. We are on the verge of something significant.

By 2026 Open Data Product Specification dominates the data economy

We need a longer term aim. For the specification penetration against any competition we should aim for same numbers that has happened with OpenAPI spec in API Economy.

--

--

Jarkko Moilanen (PhD)
Open Data Product Specification

API, Data and Platform Economy professional. Author of "Deliver Value in the Data Economy" and "API Economy 101" books.