Product and business context
OpenWeb already powered article conversations across publisher sites. Publishers wanted a persistent destination for participation, first-party audience relationships, and monetization beyond an individual article.
Forums extended the existing identity, moderation, Conversations, and advertising platform into a publisher-owned community product rather than a separate social destination.
My role and decision scope
Product direction was mine; engineering leadership remained with the engineering lead.
I held decision rights for product vision, definition, roadmap, and priority. I led discovery, partner alignment, requirements, launch readiness, and the operating cadence across Engineering, Data, Analytics, Support, Customer Success, Marketing, and other product teams.
- Product vision
- Roadmap and priority
- Partner discovery
- Beta and launch readiness
The problem
The product had to create useful community activity without becoming an empty, unsafe, or cannibalizing destination.
Publishers needed a credible community experience that readers could understand immediately, staff could operate safely, and implementation teams could adopt without a costly rebuild.
The first beta also exposed a cold-start problem, weak homepage engagement, taxonomy mismatch, and conflicting requirements around moderation and SEO.
Three important decisions
The choices that most clearly show the product judgment behind the work.
- 01
Protect the beta date instead of rebuilding for speculative SEO value.
Supporting one prospective partner’s SEO requirement credibly would have required server-side rendering and a major architectural change. I escalated the tradeoff and kept the beta on the existing platform so the team could learn on schedule.
Preserve learning velocity over speculative scope. - 02
Replace the convenient taxonomy when it blocked adoption.
The MVP reused standardized IAB categories, but publishers organized audiences around brands, teams, products, and local language. I moved custom taxonomies and category management into the roadmap.
Let publisher reality override technical convenience. - 03
Layer moderation instead of rebuilding a legacy workflow.
Full pre-publication review would have delayed every thread and consumed the beta. We combined registered authorship, automated screening, and publisher-side removal after publication.
Balance brand safety, user speed, and team capacity.
What shipped
- A feed-first, mobile-web-first Forums homepage and persistent discussion threads
- Publisher-defined taxonomies, followed categories, profiles, and Top Voices
- Registered thread creation with layered automated and publisher moderation
- Caia Autorepost, replacing synthetic AI seeding with authentic article activity
- Recirculator and Publisher Ecosystem Search across community and editorial content
- Localization, analytics, monetization, onboarding documentation, and launch controls
Results and evidence
- 3×
- up to higher time on site
- 2×
- up to growth in registered users
- 21%
- up to higher overall engagement
- Global
- public launch as part of Community Core
Behavioral results were reported by OpenWeb across selected beta partners; they do not isolate Forums as the sole cause.
What I learned
Shipping a customer request is not the same as proving a lasting reader habit.
I am proud that I maintained a coherent direction while stopping work that evidence did not support: synthetic discussion starters, standardized-only taxonomies, the navigation-heavy homepage, and further SEO investment without demonstrated value.
I would now validate the repeat reader proposition earlier—especially what a publisher-owned community can offer that a global community cannot.
Want the decisions, flows, and evidence in detail?
The full case contains the complete product story, supporting visuals, and the reasoning behind the work.