Business information service Owler has launched an API suite that opens up the startup’s dataset of over 10 million companies so that developers can access company information programmatically via API.
Owler’s New APIs
The initial API available is the free CompanyAPI that allows basic searches by company name. This returns the company’s full trading name, their address, phone number, and website.
Additional (free-for the-time-being) APIs are also available. The first extends the CompanyAPI to enable parameter searches so that users can also retrieve some financial data about a company (estimated revenue, stock exchange ticker code, employee counts, etc.) and the name of the CEO.
Two other APIs are available: one to surface competitor information (CompetitorAPI), and the FeedAPI that enables a feed of all blog posts and articles about a company to be displayed (including news on companies sourced from ProgrammableWeb’s news content). Eventually, Owler hopes to release the full range of data available via the web application as programmatically-accessible data endpoints.
Managing Company Information by ID
All requests for an API key require a personal registration process. Tim Harsch, Co-founder and Director of Product at Owler, says that eventually, the basic free API will be available by self-serve signup, but the strong interest by competitors in the closed beta API stage has led to a cautious release of the data via API, in order to best track usage levels.
Jim Fowler, Co-founder and CEO of Owler, believes part of Owler’s success will hinge on the eradication of proprietary company identifiers such as the DUNS number used by Dun & Bradstreet. Last week, at API Strategy and Practice, Founder of open data business information service OpenCorporates, Chris Taggart spoke of the dangers of locking company information behind proprietary identifiers. While Taggart says that this can contribute to poor data quality and vendor-locked data, Fowler is more concerned with usability for those searching for company information.
Fowler explains:
One of the things we are really pushing is the concept of the URL as the primary identifier of a company. Any database that sells company information has a corporation identifier. The DUNS number is a closed and proprietary number. We think the whole world will adopt our idea of the URL as company identifier.
Fowler hopes that Owler’s top-down data collation model (where a company is shown as one entity based off its main site URL) will complement OpenCorporates’ bottom-up approach (where a company’s subsidiaries, sub-branches, and additional trading names are being documented to give a more comprehensive picture of a company’s open and hidden behaviors). “At some point in time we will meet in the middle and tie it together,” says Fowler.
While the URL-as-company identifier is at the core of the data model, to surface company records using the APIs from Owler, developers will still need to get the Owler database's unique record ID.
So first, a developer would make a CompanyAPI call using the name, URL, or stock exchange ticker code as the query. This returns the basic information, including Owler’s database record ID number for that company. It is this record ID that is then used to perform additional API queries for financial details, competitor information, or to access the feeds service.
The API Business Model
“We are figuring out the pricing right now. There are a number of other APIs out there that are priced by an annual subscription or by usage,” says Fowler. For now, during the onboarding process, developers can explain which APIs they need access to and will be able to come onboard while pricing strategies are formalized.
The lack of pricing for now also speaks to a key goal of the API strategy: to raise the profile of Owler’s company information database. “We want people to start to realize the high quality of the data that we have, and making people aware of where this great data is coming from,” says Fowler.
As a result, the initial API releases really focus on the FeedAPI, which developers can use in their public integrations, as long as they show a “Powered by Owler” logo with backlinks to Owler.com.
At present, there are around 35 companies integrating the APIs from the beta release.
API User Friendliness
Limitations to the APIs include the fact that users can only return 30 responses for the basic company search. This means that if developers search for all companies with the word “green” in the title, Owler only returns the first 30 items, even if the database has many more.
“In early usage we rarely saw users go beyond 10, so 30 was chosen to give us a wider buffer,” explains Harsch. “We will be adding pagination in the future however, to allow more generic queries. One other note is that our search is made extra powerful by the community, in that companies with more followers will appear at the top so if you're searching for “green” you’ll first get the most popular results which hopefully means less pagination required (as we’ve seen to date).”
Harsch confirms that advanced search (such as by company name AND industry OR organizational type, etc.) is on the roadmap to allow much more complex queries and for generating lists of companies that meet certain query criteria.
For now, this is a bare bones API launch. The developer portal allows users to track their API call usage, and has simple documentation that explains the model schema and search parameters. A testing environment is provided within the documentation to allow users to immediately check how to query an API. Response headers and bodies to the query are shown, as well as the cURL syntax and the request URL and response codes:
The documentation is straightforward, and the testing environment easy to use. Given the current level of data being exposed, this is a clear way to show what the API is capable of for early adopters, without needing a full bells-and-whistles developer portal in place.
The greatest challenge for adopters may be from the data itself. In several cases, competitor data was limited for some companies searched. Owler’s data-as-a-service model relies on crowdsourcing by users to flesh out company information such as who exactly is a company’s main competition.
Fowler described one use case for the API where businesses could identify their customers’ competitors and use that as a list of potential new target companies for their own products. But to do this effectively via API will require more of Owler’s users to enter competition information into Owler in the first place. However, with Owler's website visits growing just over 25% in the last month (from 860,000 visitors to over 1 million, according to SimilarWeb statistics), this may not be a limitation for long.
Developers can register for an API key at Owler’s developer portal.