People mapped three food media acquisitions last week: Wonder buying Tastemade for content-as-funnel, People Inc buying Feedfeed for creator-as-distribution, ATK buying Food52 out of bankruptcy. Three deals, three theories of what food content is worth.
The business analysis is clean. The engineering question is messier: how do you integrate creator content into a publishing platform built for editorial?
Nobody has solved this publicly. No reference architecture exists. Here's why it's hard.
These aren't different formats of the same thing. They're different content types with different data models, different lifecycles, and different consumption patterns. Putting both in the same content graph requires one of three moves:
- Normalize creator content to fit editorial schemas. Lossy — you strip what makes it creator content.
- Extend the graph with new content types. Expensive — and you're now maintaining two data models in one system.
- Keep them separate. Connect at the distribution layer. What everyone actually does.
Two Approaches in the Wild
Future plc chose door #1. Their Collab program gives vetted creators a pared-down version of Vanilla CMS. Creators publish directly into Future's platform as structured editorial content. The editorial system absorbs the creator. Result: 4x more social traffic than traditional editorial content. The creator's content becomes Future's content — structured, SEO-indexed, and targetable by Future Context, their contextual ad tech. SheerLuxe, acquired January 2026, is already being brought into Collab.
People Inc chose door #3. They bought Feedfeed — a creator network with 7M followers, 1,000 influencer partners, 600M social views per year — and placed it under Eric Handelsman's food and drink vertical alongside Allrecipes and Food & Wine. Seven months later, no technical integration has been disclosed. The creator network operates alongside the editorial platform, not inside it.
The D/Cipher Blind Spot
This is where it gets consequential.
D/Cipher reads editorial content on owned pages — 12,000 taxonomy terms, 100M semantic connections, intent signals from every page view. D/Cipher orders are now 50%+ larger than non-D/Cipher orders. It's in over half of all direct digital revenue.
But D/Cipher has no signal from Feedfeed's 600M social views. Those views happen on Instagram and TikTok, where People Inc doesn't control the page, the data model, or the targeting. Creator content on social platforms is invisible to D/Cipher.
Future's approach solves this by absorption — creator content enters the CMS, gets indexed, becomes targetable. People Inc's approach preserves creator authenticity but creates a targeting blind spot in its most valuable ad tech asset.
The Graphene-Selene Question
Graphene-Selene was built for editorial articles. It serves two consumers today (website rendering and D/Cipher) with three more arriving (product apps, AI agents, internal AI tools). Adding creator content as a first-class content type would mean extending the graph's schema to accommodate unstructured, ephemeral, multi-format content from external platforms.
The IAB published a Creator Economy Taxonomy in April 2025, providing formal definitions for creator types, content categories, and platform relationships. That's a starting point for any schema extension. Netflix's knowledge graph normalizes talent entities alongside catalog content — but Netflix controls both its talent data and its content. People Inc controls its editorial content. It does not control its creators' platforms.
What I'd Recommend
Don't put creator content in Graphene-Selene. Put creators in Graphene-Selene.
The hardest integration isn't the acquisition. It's the data model.