United StatesMedia and entertainment2024 – ongoingPeople Inc. (formerly Dotdash Meredith)
OpenAI language model added to the D/Cipher ad-targeting tool
Dotdash Meredith said an OpenAI model improved content connections in its D/Cipher ad tool by 30%
Problem
D/Cipher, the publisher's own contextual ad product, infers reader interests and shopping intent from the pages people read across about 40 properties. Matching content by keywords missed articles that described the same topic in different words.
Approach
The company added OpenAI's large language model to D/Cipher. The model finds links between pieces of content by sentiment and subject, so different wording on the same topic is recognized as related. The D/Cipher general manager described the work at a Digiday event.
Results
Takeaway
A language model can sit inside an existing in-house ad product and improve how it matches content. The only figure is the executive's own, and the source does not say how the 30% was measured.
Categories
- Industry
- Media and entertainment
- Business function
- Marketing and sales
- Innovation type
- Product and service innovation
- AI technology
- Generative AI
- Adoption
- Partner co-development
- Evidence
- Company's own claim
Sources
Source summaries
Digiday reported a live podcast recording in which Lindsay Van Kirk, group GM of D/Cipher, explained how an OpenAI language model was added to the ad product. She said it produced a 30% improvement in the connections made across pieces of content. The article also gives the scale of the company's audience and data signals.
Full details are in the original sources.
