Access to Experts — CNN Product Strategy

Fri, 08 May 2026 00:00:00 +0000

Problem

CNN’s most engaged readers — what we called the Triangulator, a reader who visits four to five times a week but cross-references across multiple outlets — were leaking attention at the moment of peak intent. 49% of readers spend 10+ minutes per major story cross-checking sources outside CNN: Substack writers, niche podcasts, X threads. Meanwhile, 61% named expert analysis the single most valuable content a news outlet can offer, and 52% said it’s what they want more of from CNN specifically.

CNN had the brand equity, global reach, and trust to be the destination for this audience. It didn’t have the platform.

Why this is critical

Revenue. The all-access subscription strategy depends on subscribers deepening, not just returning. Minutes spent at a competitor during a breaking story erode the case for paying CNN.
Positioning. Competitors monetizing CNN’s lost attention — individual experts on Substack, podcast networks — are turning CNN into a starting point, not a destination.
Compounding loss. Each breaking-story session is habit-forming. The longer the leakage continues, the harder the loop is to close.

Why the framing took work

CNN’s brief was deliberately ambiguous: audiences seek diverse perspectives, but lack a single trusted destination. “Diverse perspectives” could mean a dozen different features. Two methodology choices disambiguated the signal:

Mixed-method validation. A survey alone would have misled — the “users want more perspectives” headline sounds like a request for more voices inside the story. 15–30-minute interviews surfaced that readers actually mean credentialed experts who can tell them what the story means.
External cross-check. CNN’s own internal user research independently confirmed the same signal. Their live-blogging product had recently added per-post bylines with profile-page links and saw CTR to those profiles lift measurably — the same underlying need, surfaced on a different product surface.

The strategic question that shaped every downstream decision: what do CNN’s most engaged readers actually mean when they say they want more perspectives — and is that something CNN should build, or keep ceding to Substack, podcasts, and X threads?

Approach

A 12-week product strategy engagement. The output: a single recommendation — a credentialed expert reaction layer inside CNN articles — and a defensible rollout. Five decisions defined the recommendation.

Decision 1: Build for interpretation, not multiplication

A “more voices” feature would have satisfied the survey signal and missed the interview signal. The product is short, attributed reactions from credentialed experts inside articles — experts who explain what a story means, not who repeat what it says.

Decision 2: Use CNN’s existing TV archive, not commissioned content

CNN’s TV bench is already full of credentialed experts. The Triangulator is a text reader — they don’t see those experts because they consume CNN through articles, not television. Porting transcripts and clips from the archive is cheap, fast, and avoids standing up a freelance roster. The lift is in surfacing, not sourcing.

Decision 3: Ship credential cards immediately, independent of the rest

Credential cards — the small author-authority widget under a byline — are trust infrastructure, not a feature test. They cost almost nothing to build, apply to every article on the site, and compound over time. Gating them behind the larger expert-reactions test would be a category error.

Decision 4: Sequence engagement before monetization, with KPI gates between each stage

The bet:

If CNN puts credentialed expert reactions inside the articles its triangulators read, engagement will increase — and a subset of triangulators will be willing to pay to unlock unlimited access.

Engagement is the leading indicator; paywall conversion is downstream. The rollout is gated stage-by-stage so the bet is falsifiable at each step:

Stage	What ships	Gate to advance
0	CTA-only teaser to measure interest before building	≥5% opt-in
1	MVP: one story, one expert reaction (text), free	≥30% read-through
2	3–5 reactions per story across 3–5 weekly themes, free	15% consumption, ~60% read-through, lift vs. control
3	Format expansion — text + video + audio from TV archive	3–5% click-through, hold stage 2 metrics
4	General availability with all-access paywall	2–4% paywall conversion, ~5% basic→all-access, 90% retention

Conservative break-even: ~$1.1M annual cost, ~17,000 incremental subscribers at 2% paywall conversion, ~70,000 monthly exposures.

Decision 5: Tie to the existing all-access subscription, not a standalone SKU

Slotting under all-access uses three existing levers — acquisition (free → subscribed), upsell (basic → all-access), and price — and lets the feature compete for budget against other all-access initiatives on the same terms. A standalone SKU would have fragmented the bundle and forced a separate willingness-to-pay test.

Results

The final pitch landed in front of CNN’s senior product leadership — a cross-functional group spanning growth, retention, engagement, subscriptions, content operations, and editorial × product. Two reactions from the room:

“Your user research is fairly consistent with what we’re seeing.”

“You have anticipated almost every question I would have… you’ve honed in on the obvious question, which is: would you give us money for it? Those are two different questions.”

Specific endorsements:

Credential cards validated as a ship-now item, independent of the rest of the rollout
TV archive first validated as the right starting point
Stage 0 collapsed into Stage 1 — feedback was that a tiny MVP is cheaper to build than a synthetic CTA test
Engagement-first sequencing affirmed as the right way to pitch this kind of feature

Key Insights

The survey lies about what people want; interviews tell you why they actually want it. More perspectives meant credentialed interpretation. Taking the survey at face value would have produced a feature that solved a nominal problem and missed the real one.

Atomize what already exists before building new. The product isn’t new content — it’s existing content surfaced where the user already is.

Trust signals are infrastructure, not features. Features get gated by KPIs. Infrastructure ships.

Engagement comes before monetization in the pitch, even when monetization is the actual goal. Conversion is a downstream consequence of deeper engagement, not a metric to optimize directly.

Go-to-Market | Taewan Yoon