Industry Dive’s Foray Into Generative AI Content Licensing
By: Haniya Rae
For many publishers, finding a way to generate revenue from major generative AI platforms is a challenge. But even if they haven’t yet gotten a deal done with one of the platforms, some publishers are exploring how large language models (LLMs) open up opportunities to sell vast reservoirs of content, just slightly repackaged.
Launched in September, Industry Dive’s DiveAccess API allows developers to access more than 4 million articles from the publisher’s 37 industry-specific publications, for a fee. By using the API and building on top of an LLM (DiveAccess lists OpenAI and Claude as two examples, but any LLM would work), developers can retrieve a handful of relevant articles from the database and provide an answer to a specific question using only Industry Dive’s content library.
“Content licensing has been an important part of our business for several years, Some of its traditional. Some of it manifests itself as a content marketplace,” Industry Dive’s VP of Product Megan McCoskey told AMO. “Part of the thinking with DiveAccess is, what does content licensing look like in the future?”
How the DiveAccess API Works
Instead of combing through thousands of articles spanning the past decade, this approach allows any company to gain real-time access to reported and factual information. For example, a company looking to support a sales team that sells across a variety of industries may traditionally spend months researching and writing a white paper on a specific topic. With access to the API, the team can pull a number of stats and relevant information quickly with less time spent on research. All summaries provided by the generated responses are required to include citations to the specific Industry Dive publication.
Because the access to Industry Dive’s content is only based on the query presented — this is a retrieval augmented generation (RAG) pipeline built on top of an LLM — it doesn’t allow an LLM to fully train off of the company’s vast library of content. The narrow scope keeps the generated information accurate, preventing the hallucination effect some LLMs face when trained on wider swaths of content.
It also doesn’t give the developer up-front access to all of Industry Dive’s content. Instead, a company pays more as they scale, giving Industry Dive more control of how much they charge for retrieval. DiveAccess has tiers based on the number of API calls. A free version of the plan gives 500 API calls, allowing developers to build a prototype and test it out but not much else, according to McCoskey. A company needing to support a full sales team may pay between $20 for 5,000 API calls and $100 a month for 30,000 calls; eventually, they may need to opt for the custom pricing depending on the scope and size of the team.
If and when this type of API becomes more popular among publishers, the Industry Dive team may reassess what the content is worth. But for now it’s a test to see how interested companies use the product. “There aren’t a lot of great examples in the market yet for pricing,” said McCoskey. “So, we decided to throw some stuff out there.”
What This Means for Industry Dive
If it takes off, DiveAccess could be a revenue driver, a traffic driver, an audience development tool, and a play at brand awareness.
Because the content comes only from Industry Dive’s reported articles, and because summaries require some type of credit pointing back to the Industry Dive publication as the source of the information, the publisher gains brand credibility. If the summary is used for a site or application with links, they all point back to the respective Industry Dive publication.
DiveAccess is also useful for Industry Dive to know more about their customers. Each query from a developer is logged, so theoretically, Industry Dive can see who’s requesting what and how the information is being used. This could in turn help Industry Dive publications make content decisions, akin to looking at Google Analytics and deciding to write more about specific topics that are driving a lot of traffic. Specific queries may also inform larger trends within certain industries.
Interestingly, other than the initial press release, McCoskey doesn’t think Industry Dive will promote the tool or do much to push it out there. So, it’s still a big ‘if’as to whether DiveAccess takes off more generally. But McCoskey said the tool was built based on interest from current sponsors who were already wanting to pay.
“Businesses who need trustworthy and domain-specific information as part of how they are leveraging LLMs are the right fit,” said McCoskey. In other words, businesses already building chatbots and wanting accurate information quickly already see value in Industry Dive.
“We haven’t seen this in the wild yet,” said McCoskey. “Prior to it launching, we had side conversations. This is much for people who specifically already know the content.”