At last week’s OKFestival in Berlin, Kat Borlongan and Chloé Bonnet from Parisian open data startup, Five By Five, moderated an interactive speed-geek session to examine how startups are building viability using open data and open data APIs. The picture that emerged revealed a variety of composite approaches being used, with all those presenting having just one thing in common: a commitment to fostering ecosystems that will allow other startups to build alongside them.
Above: Kat Borlongan hosts a packed room of developers interested in open data business models.
The OKFestival—hosted by the Open Knowledge Foundation—brought together more than 1,000 participants from around the globe working on various aspects of the open data agenda: the use of corporate data, open science research, government open data and crowdsourced data projects.
In a session held on the first day of the event, Borlongan facilitated an interactive workshop to help would-be entrepreneurs understand how startups are building business models that take advantage of open data opportunities to create sustainable, employment-generating businesses.
Citing research from the McKinsey Institute that calculates the value of open data to be worth $3 trillion globally, Borlongan said: “So the understanding of the open data process is usually: We throw open data over the wall, then we hold a hackathon, and then people will start making products off it, and then we make the $3 trillion.”
Borlongan — whose Five by Five business also operates as the agency representing the Open Data Institute in Paris — argued that it is actually a “blurry identity to be an open data startup” and encouraged participants to unpack, with each of the startups presenting exactly how income can be generated and a viable business built in this space.
Jeni Tennison, from the U.K.’s Open Data Institute (which supports 15 businesses in its Startup Programme) categorizes two types of business models:
- Businesses that publish (but do not sell) open data.
- Businesses built on top of using open data.
Businesses That Publish but Do Not Sell Open Data
At the Open Data Institute, Tennison is investigating the possibility of an open address database that would provide street address data for every property in the U.K. She describes three types of business models that could be created by projects that generated and published such data:
Freemium: In this model, the bulk data of open addresses could be made available freely, “but if you want an API service, then you would pay for it.” Tennison pointed to lots of opportunities also to degrade the freemium-level data—for example, having it available in bulk but not at a particularly granular level (unless you pay for it), or by provisioning reuse on a share-only basis, but you would pay if you wanted the data for corporate use cases (similar to how OpenCorporates sells access to its data).
Cross-subsidy: In this approach, the data would be available, and the opportunities to generate income would come from providing extra services, like consultancy or white labeling data services alongside publishing the open data.
Network: In this business model, value is created by generating a network effect around the core business interest, which may not be the open data itself. As an example, Tennison suggested that if a post office or delivery company were to create the open address database, it might be interested in encouraging private citizens to collaboratively maintain or crowdsource the quality of the data. The revenue generated by this open data would then come from reductions in the cost of delivery services as the data improved accuracy.
Businesses Built on Top of Open Data
Six startups working in unique ways to make use of available open data also presented their business models to OKFestival attendees: Development Seed, Mapbox, OpenDataSoft, Enigma.io, Open Bank API, and Snips.
Startup: Development Seed
What it does: Builds solutions for development, public health and citizen democracy challenges by creating open source tools and utilizing open data.
Type of business model: Consultancy, but it has also created new businesses out of the products developed as part of its work, most notably Mapbox (see below).
What it does: Open data platform with advanced discovery and search functions.
Open data API focus: Provides the Enigma API to allow programmatic access to all data sets and some analytics from the Enigma platform.
Type of business model: SaaS including a freemium plan with no degradation of data and with access to API calls; some venture funding; some contracting services to particular enterprises; creating new products in Enigma Labs for potential later sale.
What it does: Enables users to design and publish maps based on crowdsourced OpenStreetMap data.
Type of business model: SaaS including freemium plan; some tailored contracts for big map users such as Foursquare and Evernote.
Startup: Open Bank Project
What it does: Creates an open source API for use by banks.
Open data API focus: Its core product is to build an API so that banks can use a standard, open source API tool when creating applications and web services for their clients.
Type of business model: Contract license with tiered SLAs depending on the number of applications built using the API; IT consultancy projects.
What it does: Provides an open data publishing platform so that cities, governments, utilities and companies can publish their own data portal for internal and public use.
Open data API focus: It's able to route data sources into the portal from a publisher’s APIs; provides automatic API-creation tools so that any data set uploaded to the portal is then available as an API.
Type of business model: SaaS model with freemium plan, pricing by number of data sets published and number of API calls made against the data, with free access for academic and civic initiatives.
What it does: Predictive modeling for smart cities.
Open data API focus: Channels some open and client proprietary data into its modeling algorithm calculations via API; provides a predictive modeling API for clients’ use to programmatically generate solutions based on their data.
Type of business model: Creating one B2C app product for sale as a revenue-generation product; individual contracts with cities and companies to solve particular pain points, such as using predictive modeling to help a post office company better manage staff rosters (matched to sales needs) and a consultancy project to create a visualization mapping tool that can predict the risk of car accidents for a city.
Composite Models Rule
The majority of those presenting relied on multiple revenue streams in order to build their businesses. For example, for the Open Bank Project, it can take up to 18 months to build a relationship with a new bank customer. While this is viable once the customer is on board, the long time frame for customer acquisition means the startup also provides IT consultancy projects in related security and financial fields while building out its customer base. Snips has an interesting model where it plans to create a “Google Now that works”-type location-recommendation application that can be sold via app stores to generate a base income, while also continuing to pursue discrete, consultancy-based projects with city and company clients.
Of all those presenting, only OpenDataSoft had a pure-play SaaS model where all income to manage the business is coming solely from subscription payments for the service. In order to extend its business further, however, it is looking for venture funding to use to expand its base into new markets.
A Commitment to Equity and Ecosystems
An interesting and heartening discovery among these startups is that each one has built a social good component into its business model. Enigma.io, for example, is able to offer its freemium plan to those wanting to experiment with open data, which has been made possible because it has already built out the infrastructure by working with private clients. Snips, Open Bank Project and Development Seed open source the tools they're building.
In addition, all are thinking proactively about how to support other developers. For example, OpenDataSoft is solely focused on providing its portal product to customers. It encourages customers to link with third-party developers when additional data management services are needed. Much of Development Seed’s work moves beyond consulting and into capacity building to help widen the skill sets of those it comes in contact with, in order for the type of work it does to be replicable and scalable around the globe.
A Sustainable Business Ecology
Open data leaders like those involved with the Open Knowledge Foundation and experts like Joel Gurin from Open Data 500 are working to document use cases that demonstrate the viability of businesses that use open data as a raw material in their products.
Along the way, as these businesses grow using primarily composite business models with multiple revenue streams, they are also finding ways to foster a self-sustaining ecosystem.
All images of the startups were sourced from each business’ website.