LLenny's Podcast
← All episodes
Alex Hardiman (CPO at The New York Times)13 November 2022

An inside look at how the New York Times builds product

6Frameworks
15Insights

Frameworks in this episode

Leadership4 steps

Functions and Missions: The Two-Axis Org

Craft and careers live in functions; delivery lives in cross-functional missions with three flavours

Read →
Entrepreneurship6 steps

Protect the Core Magic: Acquisition Integration

Freeze the loop users love, rebuild everything behind it, and narrate the seams honestly

Read →
Innovation5 steps

Scaling Expert Judgment Through Algorithms

Structure your experts' judgment into a signal, train on that — not only on engagement

Read →
Strategy5 steps

The Solar System Bundle

Anchor a subscription on one sun product, then orbit satellites that share its DNA

Read →
Innovation5 steps

Two-Mode Product Teams

One team, two clocks: ship at the speed of the moment, and build the system between moments

Read →
Leadership6 steps

Wartime Product Management

In a crisis: blow up the roadmap, share context, restate the mandate from the mission

Read →

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Hot Take· 1

Hot Take54:30

At the NYT, PMs work the full stack: content, distribution, and product

Alex argues the key difference from big-tech product work is that NYT product managers work across the full stack: the company owns its journalism and content, its distribution, and its products. At Facebook they built software and distribution but didn't control the content, which created many of the quality problems discussed earlier. The best NYT products come from combining journalism and product levers.

  • NYT owns content, distribution, and product; big-tech platforms own only two of three
  • At Facebook, lack of content control drove many quality challenges
  • Best products emerge when journalism and product levers are combined
  • PMs must blend art and science, valuing editorial judgment alongside KPIs

“you work across the full stack of the product, meaning we own our journalism and our content, we own our distribution”

Alex Hardiman · 54:30
#product-management#full-stack#media-vs-tech

Explainer· 5

Explainer09:30

Why Facebook's binary view of content made misinformation hard to fight

Alex explains that Facebook historically treated content as binary: friends-and-family versus public content, where public content could come from a reputable newsroom or anyone declaring something true. Her team tried to unpack public content to elevate factually accurate, trusted journalism while reducing dubious or misleading propaganda, a hard task because the platform was never built to classify coverage that way.

  • Facebook's content model was binary: friends/family vs. undifferentiated public content
  • Public content lumped reputable newsrooms together with anyone's unverified claims
  • The fix meant surfacing trusted journalism while reducing dubious or misleading propaganda
  • The platform wasn't built for classification, and lacked the right goals and incentives

“You basically had content from friends and family and then you had public content and public content could come from anywhere.”

Alex Hardiman · 09:30
#facebook#content-moderation#journalism#product
Explainer16:30

What 'the product' actually is at The New York Times

Alex reframes the NYT product as its journalism, married to a compelling and useful user experience that helps people act on that journalism. For 150 years the product was a bounded printed newspaper; about 25 years ago the shift to digital opened it up and forced the Times to build a wide array of products to extend its reach.

  • The core product is journalism plus a UX that helps people act on it
  • For ~150 years the product was a finite, time-bound printed newspaper
  • Digital forced a shift from a single bounded product to a wide array of products
  • Early digital products included the website, apps, newsletters, and VR/AR experiments

“our product is our journalism, which we then marry with a really compelling and useful user experience in a way that helps people really act…”

Alex Hardiman · 16:30
#new-york-times#product-strategy#media
Explainer20:30

'News is the sun': the solar-system model of NYT products

Alex describes a solar-system metaphor where news is the sun, the reason the company exists and the source of its brand, trust, largest audience and highest-quality coverage. News gives birth to satellite products that share the same DNA of trusted journalism and expertise, and the third strategic piece is connecting them into one family so entering via news or Wordle exposes everything else.

  • News is 'the sun': brand heritage, trust, largest audience, top-of-funnel
  • Satellite products inherit the same DNA of trusted journalism and expertise
  • Three needs: best news destination, strong beyond-news products, a connected bundle
  • Entering via any product should let readers discover everything else

“we actually have the solar system metaphor where for us news is the sun in the sense that it's why we exist.”

Alex Hardiman · 20:30
#strategy#new-york-times#product-portfolio
Explainer22:30

How NYT structures product teams: functions, missions, and embedded editors

The org runs on two axes: functions (product and design, handling craft and career growth) and missions (cross-functional teams pursuing shared goals). Missions come in three types, consumer, monetization, and platform, and the distinguishing feature versus a tech company is that editors are embedded in consumer missions that directly shape journalism.

  • Two axes: functions (product, design) and cross-functional missions
  • Three mission types: consumer, monetization, platform
  • Editors are embedded in consumer missions that shape journalism
  • Structure otherwise mirrors a tech company, minus the editors

“the way to set that up so that our people can really do their best work is that we have two axes. We have functions…”

Alex Hardiman · 22:30

“the big difference is we also have editors if it's a product space that directly shapes our journalism.”

Alex Hardiman · 23:30
#org-structure#product-teams#new-york-times
Explainer56:00

Training algorithms on editorial signal, not just engagement

Alex explains that the NYT home screen starts with expert editorial judgment, then trains algorithms on data sets like editorial importance scores drawn from journalists, letting editorial judgment scale to many readers while still driving reach, engagement and conversion. She contrasts this with Facebook, where ranking could only be trained on engagement outcomes, never on the quality of the information itself.

  • The home screen begins with expert editorial judgment on the top stories
  • Algorithms are trained on editorial importance scores from journalists
  • This scales editorial judgment while still optimizing reach and conversion
  • At Facebook, ranking could only train on engagement, not on content quality
  • Structuring 2,000+ journalists' expertise into algorithmic decisioning is unique

“we're training algorithms on specific data sets like editorial important scores that actually come from our journalists.”

Alex Hardiman · 56:00

“when I was at Facebook and we were focused on news ranking and feed, all we could do was train pieces of in- information based…”

Alex Hardiman · 56:30
#algorithms#editorial#ranking#personalization

Story· 5

Story09:00

Leading Facebook's news product as 'wartime product management'

Alex joined Facebook months before the 2016 election and moved to lead its news product and engineering teams right after, amid intense reckoning over misinformation and platform responsibility. She frames the experience as 'wartime product management' and credits the team's humility for enabling an honest diagnosis of what was actually happening on the platform.

  • She joined Facebook in late 2016 and shifted to the news effort in early 2017, right after the election
  • The work centered on misinformation, disinformation, election integrity and platform responsibility
  • She says humility was needed to first diagnose what was really happening before fixing it

“I think about wartime product management,”

Alex Hardiman · 09:00

“there was just this incredible humility, right? That was needed to really understand and first first diagnose like what was actually happening on the platform.”

Alex Hardiman · 09:30
#facebook#misinformation#leadership#crisis
Story28:30

The article that showed you your own 'productivity score' as you read it

Alex points to reporter Jodi Kantor's investigation into employers tracking remote workers via productivity scores. The piece was designed to display a reader's own productivity score in the moment as they read it, an effect she calls visceral and creepy, and an example of magic that only happens when dedicated designers and engineers sit with a reporter on special one-off features.

  • Jodi Kantor's story covered employer surveillance of remote workers
  • The article surfaced the reader's own productivity score live as they read
  • Autonomous newsroom teams (graphics, visual/interactive news) build these one-offs
  • A separate storytelling product team scales winning formats across the report

“She recently did a piece on how employers are tracking and monitoring remote workers with tools like productivity scores.”

Alex Hardiman · 28:30

“And it was super visceral, really creepy in the most effective way.”

Alex Hardiman · 29:00
#storytelling#journalism#interactive#design
Story34:30

Why The New York Times acquired Wordle in a matter of weeks

Alex recounts learning about Wordle from a NYT column about creator Josh Wardle, and how games GM Jonathan Knight had already reached out before publication. Wordle shared the DNA of NYT word games like Spelling Bee and the Mini crossword, and games fit the subscription strategy as a counterpoint to news; the acquisition closed in weeks, faster than any deal she'd been part of.

  • A NYT reporter's column about Josh Wardle sparked internal interest
  • Games GM Jonathan Knight had already contacted Josh before the piece ran
  • Wordle shares DNA with Spelling Bee and the Crossword Mini
  • Games are framed as a 'counterpoint to the news', a break, not empty calories
  • The deal closed in weeks, faster than any acquisition she'd done

“we really see games and demand for games as this basically like it's like a counterpoint to the news. It gives people a chance to…”

Alex Hardiman · 34:30

“the acquisition talks were incredibly fast. I mean, the whole thing took place in a matter of weeks, which is way faster than any other…”

Alex Hardiman · 35:00
#wordle#acquisition#games#strategy
Story37:00

Migrating Wordle's backend, and the day the answer was 'fetus'

Wordle arrived as a simple web game with no backend, so stats and streaks lived in local browsers and vanished when users switched devices. The Times connected it to a free NYT account to protect streaks and expand surfaces, but mid-migration the pre-loaded solution one day was 'fetus', right as the Roe v. Wade draft leaked, and because users were split across old and new versions they couldn't change the word for everyone. Alex says transparency about the migration defused the rumors.

  • Wordle had no backend; stats/streaks were stored in local browsers and easily lost
  • The Times linked it to a free NYT account to protect streaks and add surfaces
  • Mid-migration, the day's pre-loaded word was 'fetus' as the Roe draft leaked
  • Split old/new versions meant the word couldn't be changed for everyone
  • Public transparency about the integration process defused the coincidence

“when we acquired Wordle, it was a simple web game with no back end.”

Alex Hardiman · 37:00

“the Wordle solution the next day was fetus, which is just an extraordinarily bad coincidence because the word had been loaded into the game by…”

Alex Hardiman · 38:30
#wordle#integration#engineering#transparency
Story42:00

Building COVID products from a laundry closet in March 2020

Alex returned to the Times just before the pandemic and describes the COVID onset as a wartime product moment: she packed her kids off to Vermont and disappeared into a laundry closet for two weeks as Slack blew up. With institutions failing to explain the basics, the team blew up roadmaps to build a public COVID case dataset from a spreadsheet, launched zip-code-level infection and vaccination tools, and made critical public-safety coverage free.

  • She rejoined the Times in late 2019, just before the pandemic
  • Team built a comprehensive public COVID case dataset, starting from a spreadsheet
  • New tools let people check infection and vaccination rates at the zip-code level
  • Most important COVID coverage was made free to everyone
  • In March 2020, roughly half the country came to the New York Times

“I literally just went into a laundry closet, and I didn't emerge for 2 weeks, because my Slack was blowing up about all of the…”

Alex Hardiman · 43:00

“In March 2020, we saw that half of the country came to the New York Times.”

Alex Hardiman · 46:30
#covid#data-journalism#crisis#product

Tool· 1

Tool1:00:00

Alex Hardiman's book picks: Stripe Press plus a novel

In the lightning round, Alex recommends Stripe Press books as evergreen references, naming High Growth Handbook and Will Larson's An Elegant Puzzle, plus the more topical Revolt of the Public. She always pairs a pragmatic work book with fiction, currently rereading James Baldwin's Giovanni's Room, arguing that stepping away from practice-related books sparks some of the best product ideas.

  • Stripe Press titles as evergreen references: High Growth Handbook, An Elegant Puzzle, Revolt of the Public
  • She deliberately balances pragmatic work books with fiction
  • Currently rereading Giovanni's Room by James Baldwin
  • Believes fiction sparks ideas that feed back into building news products

“I love uh Stripe Press. And so I think a lot of the books that they have are just such good references.”

Alex Hardiman · 1:00:00

“I also always try to balance books and my own reading time and recommendations”

Alex Hardiman · 1:00:30
#books#recommendations#reading

Takeaway· 3

Takeaway19:30

The 'essential subscription' bet: 135M addressable, 15M subscribers by 2027

A year-long strategy exercise sized the Times' addressable market at ~135 million people worldwide willing to pay for high-quality journalism-based products across news, games, cooking, sports, shopping and audio. The goal is to become the essential daily subscription by bundling these categories, targeting 15 million subscribers by 2027, up from just over 9 million at the time.

  • Addressable market estimated at ~135 million people globally
  • Categories: news, games, cooking/recipes, sports, shopping recommendations, audio
  • The Athletic was acquired to serve the sports category
  • Target: 15 million subscribers by 2027, from just over 9 million

“there are about 135 million people around the world who we believe are willing to pay for the type of high-quality journalism-based products that The…”

Alex Hardiman · 19:30

“We're just over 9 million today and I really think we can do it.”

Alex Hardiman · 22:00
#subscription#strategy#bundle#growth
Takeaway47:30

Fighting burnout by being ruthless about what to stop doing

Alex calls burnout one of the hardest ongoing challenges. Beyond time off and financial and daycare support, she argues the biggest lever inside the job is reducing context switching by being disciplined about the small number of important things that truly must happen at any moment, and being honest about what the team is happy to stop doing.

  • Context switching, within a job and across job and life, is a core burnout driver
  • Baseline support: more time off, financial help, daycare and health benefits
  • The real lever is focusing on a small number of truly important things
  • Being explicit about what to stop doing is as important as what to do

“context switching is one of the things that is really, really difficult.”

Alex Hardiman · 47:30

“the places where we can be so much more focused and thoughtful about the small number of important things that we must do at a…”

Alex Hardiman · 48:00
#burnout#leadership#focus#management
Takeaway58:00

A broader definition of impact than 'move this metric'

Alex contrasts Facebook's focus on scale, engagement and revenue with the Times, where business goals serve the mission of seeking truth and helping people understand the world, not the other way around. Impact means growing subscribers, but also when a deeply reported story triggers a policy change or a new law, giving product managers a broader aperture on the relationship between business and mission.

  • Facebook centered on scale, engagement and revenue
  • At the Times, business goals are in service of the mission, not vice versa
  • Impact counts subscriber growth and real-world outcomes like policy change
  • PMs get a broader aperture on business goals vs. mission-impact goals

“our impact and like our business goals are in service of our mission, which is to seek the truth and kind of help people understand…”

Alex Hardiman · 58:00
#impact#mission#product-management#purpose