LLenny's Podcast
← All episodes
Yuriy Timen (Grammarly, Canva, Airtable)01 September 2022

How to grow a subscription business

7Frameworks
15Insights

Episode overview

Lenny Rachitsky interviews former Grammarly growth leader Yuriy Timen about building subscription growth through paid acquisition, SEO, virality, and product-led loops. Timen explains how product economics, network effects, programmatic content, attribution, and company stage determine which channel deserves investment. He also emphasizes focused experimentation, customer research, onboarding, and delaying diversification until a primary engine has reached meaningful scale.

Key ideas

  • Subscription businesses should select growth engines according to their economics and product structure: high LTVs can support paid acquisition, inherent network effects can support virality, and unique content, data, or programmatic surfaces can support SEO.
  • Network effects are difficult to manufacture; referral and viral loops work best when each user's utility rises as relevant users join, though beloved brands can also sustain incentivized referrals.
  • SEO generally has a longer return horizon, so teams should seek a differentiated editorial, programmatic, or proprietary-data angle and design a time-boxed experiment that lowers the cost of learning.
  • Paid acquisition can either produce efficient growth or rapidly test messaging and positioning; teams should be explicit about which purpose they are funding.
  • Use focused experimentation with predefined guardrails: give a channel a credible test, evaluate its incremental promise and opportunity cost, then pause rather than permanently discard it.
  • Most admired companies still depend heavily on one dominant growth engine; early-stage teams should exploit a working engine before diversifying, while scaled companies must manage concentration risk.
  • Onboarding is especially valuable for complex prosumer products because tailored guidance can make professional-grade capabilities accessible without overwhelming new users.
  • Customer interviews, surveys, and other research consistently create clarity, while attribution investment must be proportional to company stage, channel complexity, and potential payoff.

Transcript available · source text is retained privately and is not published

Frameworks in this episode

Strategy3 steps

Focus With Guardrails

Concentrate on one channel but set abandonment criteria up front so you know exactly when to move on

Read →
Marketing4 steps

Onboarding Leverage For Robust Products

Use customized onboarding to reconcile simple-enough with pro-grade, and unlock 2-4x activation

Read →
Marketing3 steps

Paid Readiness Diagnostic

Read a company's funnel to judge whether paid can be a profitable channel or only a learning tool

Read →
Finance3 steps

Subscription Funnel Benchmarks

Category-dependent conversion thresholds for judging whether a subscription funnel is healthy

Read →
Marketing4 steps

The 2-of-3 SEO Opportunity Audit

Decide if SEO is worth pursuing by checking three angles and time-boxing the test to three months

Read →
Strategy3 steps

The Diversification Timing Test

Know whether it's too early to diversify growth channels or dangerously overdue

Read →
Marketing3 steps

The Three Growth Engines Selection

Diagnose which of virality, SEO, or paid your product can actually grow on before spending on any

Read →

People & resources mentioned

Attributed to the moment in the episode. Timestamps are approximate.

People · 13

  • Lenny RachitskyMentionstoday we've got another very special compilation episode something I've been pulling on more and more with the podcast and the newsletter

    “yuri welcome to the podcast”

    ▶ Watch at ~3:30approx.
  • Yuriy TimenMentionsyou were head of growth at grammarly you spent nine years there

    “you were head of growth at grammarly you spent nine years there”

    ▶ Watch at ~4:00approx.
  • Elena VernaMentionsElena Vernon who's been on the podcast is just like a huge proponent of everyone will be product LED self-serve eventually

    “i've had casey winners on this podcast and elena verna”

    ▶ Watch at ~6:00approx.
  • Casey WintersMentionswho's had the most influence on you in your career well we mentioned him before so I gotta bring him back up Casey Winters

    “i've had casey winners on this podcast”

    ▶ Watch at ~6:00approx.
  • Casey WintersMentionswho's had the most influence on you in your career well we mentioned him before so I gotta bring him back up Casey Winters

    “casey especially he's such a mensch when it comes to just being generous with his time”

    ▶ Watch at ~6:30approx.
  • Ethan SmithRecommendsone of the most innovative disciplined first principles seo thinkers and i've met is ethan smith from graphite

    “one of the most innovative disciplined first principles seo thinkers and i've met is ethan smith from graphite”

    ▶ Watch at ~22:30approx.
  • Viktor FranklCoinedif the freckles a man search for meaning it's just a remarkable memoir

    “if the freckles a man search for meaning it's just a remarkable memoir”

    ▶ Watch at ~1:01:00approx.
  • Adolf HitlerMentionswinston churchill's response in 1941 when you know hitler started marching through europe

    “winston churchill's response in 1941 when you know hitler started marching through europe”

    ▶ Watch at ~1:01:30approx.
  • Winston ChurchillMentionsChurchill's a really good writer. And he has a whole history of World War II that he wrote

    “winston churchill's response in 1941 when you know hitler started marching through europe”

    ▶ Watch at ~1:01:30approx.
  • Volodymyr ZelenskyyMentionsthere have been a lot of references drawn between like the president zelinski and his response in this war

    “there have been a lot of references drawn between like the president zelinski and his response in this war”

    ▶ Watch at ~1:01:30approx.
  • Erik LarsonCoinedi read a book called i think the splendid and the vile by eric larson

    “i read a book called i think the splendid and the vile by eric larson”

    ▶ Watch at ~1:01:30approx.
  • Adam SandlerMentionsso i watched hustle with adam sandler i loved that

    “so i watched hustle with adam sandler i loved that”

    ▶ Watch at ~1:04:00approx.
  • Mark FiskeMentionshe was leading growth and marketing at credit karma for a while

    “he was leading growth and marketing at credit karma for a while”

    ▶ Watch at ~1:05:30approx.

Resources · 26

  • GrammarlyMentionscompany · Grammarly

    “you were head of growth at grammarly you spent nine years there”

    ▶ Watch at ~4:00approx.
  • Otter.aiMentionssoftware

    “there is otter dot ai who else flow health”

    ▶ Watch at ~5:30approx.
  • LinkedInUseswebsite · LinkedIn

    “honestly linkedin is really the only place i i'm pretty low-key otherwise”

    ▶ Watch at ~9:30approx.
  • AirbnbMentionscompany · Brian Chesky, Joe Gebbia, and Nathan Blecharczyk

    “like airbnb from your days obviously you know marketplace very strong product network effect dynamics”

    ▶ Watch at ~14:00approx.
  • MintMentionssoftware

    “think of it as like you know a new and improved version of mint”

    ▶ Watch at ~19:00approx.
  • SimilarwebUsessoftware · Similarweb

    “just like plug things into like similar web and and trying to figure out the opportunities there”

    ▶ Watch at ~21:30approx.
  • GraphiteMentionscompany

    “one of the most innovative disciplined first principles seo thinkers and i've met is ethan smith from graphite”

    ▶ Watch at ~22:30approx.
  • RecastMentionscompany

    “the company that's leading the charge of bringing the media mix modeling methodology of like the traditional advertising era”

    ▶ Watch at ~29:30approx.
  • RecastRecommendscompany

    “so i've heard really good things i haven't tried them with any of my companies yet”

    ▶ Watch at ~29:30approx.
  • MeasuredRecommendssoftware

    “one is measured can be found at measure.call amazing domain name”

    ▶ Watch at ~31:30approx.
  • INCRMNTALRecommendssoftware

    “then the other one is incremental but in incremental outcome no no vowels except except the last a between the t and the l”

    ▶ Watch at ~31:30approx.
  • YouTubeMentionswebsite · Google

    “oh youtube we tried it it doesn't work”

    ▶ Watch at ~34:00approx.
  • EppoMentionssoftware · Eppo

    “this episode is brought to you by epo epo is a next generation a b testing platform built by airbnb alums”

    ▶ Watch at ~34:30approx.
  • EppoMentionssoftware · Eppo

    “check out epo at getepo.com”

    ▶ Watch at ~35:30approx.
  • TikTokRecommendssoftware · ByteDance

    “tick tock has so many users and it's still so relatively unsaturated with advertisers that like your audience is on that”

    ▶ Watch at ~57:30approx.
  • EssentialismMentionsbook · Greg McKeown

    “it's a book called essentialism i forget the author's name i think his last name is mccown or something”

    ▶ Watch at ~1:00:00approx.
  • EssentialismRecommendsbook · Greg McKeown

    “i would say there are a couple of books that stuck with me over the years”

    ▶ Watch at ~1:00:00approx.
  • Man's Search for MeaningRecommendsbook · Viktor Frankl

    “if the freckles a man search for meaning it's just a remarkable memoir on you know perseverance”

    ▶ Watch at ~1:01:00approx.
  • UkraineMentionsplace

    “i'm originally from ukraine i believe you are as well”

    ▶ Watch at ~1:01:30approx.
  • The Splendid and the VileRecommendsbook · Erik Larson

    “i read a book called i think the splendid and the vile by eric larson”

    ▶ Watch at ~1:01:30approx.
  • All-In PodcastRecommendspodcast · Chamath Palihapitiya, Jason Calacanis, David Sacks, and David Friedberg

    “i would say the all-in-one i feel like it's like a cool it's a cool way for me to just catch up”

    ▶ Watch at ~1:03:30approx.
  • HustleRecommendsfilm

    “so i watched hustle with adam sandler i loved that so good”

    ▶ Watch at ~1:04:00approx.
  • Credit KarmaMentionscompany

    “he was leading growth and marketing at credit karma for a while”

    ▶ Watch at ~1:05:30approx.
  • ReforgeMentionscompany · Brian Balfour

    “he shows up in the reforged chats a lot”

    ▶ Watch at ~1:05:30approx.
  • Lenny's NewsletterMentionsnewsletter · Lenny Rachitsky

    “they can also find me inside of lenny's newsletter”

    ▶ Watch at ~1:06:30approx.
  • Lenny's PodcastMentionspodcast · Lenny Rachitsky

    “you can find all past episodes or learn more about the show at lennyspodcast.com”

    ▶ Watch at ~1:08:00approx.

Spot an error or want something removed? Request a correction or removal.

Insights & moments

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

Myth Buster· 2

Myth Buster38:00

Admired Companies Aren't the Diversified Growth Engines You Assume

Yuriy busts the idea that famous companies run highly diversified growth engines — 80/20 usually applies, with one strategy working overwhelmingly well (Grammarly's performance marketing, Canva's programmatic SEO). The nuance on timing: he talks early-stage teams out of diversifying too soon when their winning channel is still small, and pushes late-stage teams that are 90%+ reliant on one channel to start carving out resources to de-risk.

  • 80/20 applies: one strategy typically drives the majority of growth
  • Grammarly leaned on performance marketing; Canva on long-tail programmatic SEO
  • Even some later-stage companies are 'one trick ponies' 90%+ reliant on one channel
  • Diversifying too early distracts from a winner that's still small — lean in first
  • Late-stage single-channel reliance is a blind-spot risk that needs deliberate diversification

“definitely the 820 applies there's usually some kind of strat this is a kind of strategy that's working overwhelmingly well and there is a scramble…”

Yuriy Timen · 38:30

“don't don't get distracted with diversification we'll get there lean more into this hit these growth rates”

Yuriy Timen · 41:00
#growth-strategy#diversification#80-20#risk
Myth Buster56:30

TikTok Isn't Just for 15-to-22-Year-Olds

Yuriy pushes back on brands writing off TikTok as a Gen Z channel. He's seen it emerge as one of the most efficient digital channels for some brands, and because it's still relatively unsaturated with advertisers, unexpected demos are reachable — including a 40-plus, married, $200k+ household-income audience.

  • TikTok is increasingly one of the most efficient digital channels for some brands
  • It's still relatively unsaturated with advertisers
  • The 'it's only for teens' assumption is wrong — your audience is likely there
  • Yuriy has reached 40+, married, $200k+ household-income demos on TikTok

“i've worked with brands that you know their core demo is like 40 plus married you know making 200k plus in in household annual income…”

Yuriy Timen · 57:30
#tiktok#paid-acquisition#channels#audience

Hot Take· 2

Hot Take24:30

The Paid-Media Contraction Is Creating Winners

Yuriy argues the market downturn is squeezing paid budgets — acceptable payback windows collapsed from ~12 months to 6 months or less, and unattributable spend is getting cut. But that contraction is an opportunity: disciplined companies with strong cash, unit economics, and payback (he names Grammarly and Canva) benefit from decreased competition and lower CPMs as looser-spending rivals pull back.

  • Payback tolerance shifted from ~12 months to 6 months or less
  • Spend that can't be attributed to sales is being cut first
  • Well-instrumented, disciplined companies win as competitors retreat
  • Fewer competing bidders means lower CPMs on ad platforms
  • A short-term contraction opens room for smarter attribution investments

“so if before for a consumer subscription company you know 12 month payback was decent now it's like you better pay back your paid media…”

Yuriy Timen · 25:00

“the companies that have strong cash positions had strong unit economics strong pay back periods already right”

Yuriy Timen · 27:00
#paid-acquisition#attribution#downturn#payback-period
Hot Take46:30

Paid Has Two Jobs — And Confusing Them Wastes Money

Yuriy reframes paid acquisition as two distinct plays: driving profitable returns at efficient unit economics, and buying fast, at-scale learnings on messaging, positioning, and design. His frustration is companies that can't say which one they're running — burning $100k/month 'inefficiently' without even structuring experiments to extract the learnings.

  • Paid does two things: profitable acquisition, or fast learning on messaging/positioning/design
  • It's hard to get faster at-scale learnings than A/B testing headlines and search copy
  • Trouble starts when a company can't name which of the two jobs it's paying for
  • Yuriy can judge paid readiness from funnel conversion, retention curves, LTVs, and churn
  • Weak funnel/low LTV = not time for paid; healthy funnel + less-saturated channels = worth a go

“the other way to do it is because it's a very quick way to get learnings on messaging and positioning on designs”

Yuriy Timen · 46:30
#paid-acquisition#experimentation#funnel#unit-economics

Explainer· 5

Explainer10:30

The Archetypes of Subscription Growth (and What Consumer LTVs Really Look Like)

Yuriy sorts subscription companies into a few growth buckets: those with strong consumer LTVs that can afford paid, those with network effects that can lean on referral/viral loops, and those with a long-tail programmatic SEO angle like Canva. He gives concrete LTV context: most $5-7/month consumer subs cap out at ~$50-60 LTV, while prosumer buyers can be worth $120-130 a year.

  • Growth archetype depends on whether you have strong unit economics, network effects, or an SEO/programmatic angle
  • Cheap ($5-7/mo) consumer subscriptions typically cap LTV at $50-60
  • Prosumer buyers (using a personal tool for work) push LTV into the hundreds of dollars
  • If you have healthy LTVs and convert 5%+ of free users to paid, paid growth loops become viable
  • Canva's initial engine was long-tail SEO on 'make' and 'template' keywords

“we're talking like in the hundreds of dollars most consumer subscription companies uh that are like five to seven dollars a month their ltds typically…”

Yuriy Timen · 11:00
#subscription#growth#ltv#unit-economics#paid-acquisition
Explainer15:30

What Network Effects Actually Mean for Growth

Yuriy defines network effects simply: each user's utility rises as more users join. He adds a nuance for marketplaces and collaboration tools, where it's not raw user count that matters but the density of users in the specific market, team, or company you care about.

  • Network effect = each user's utility increases with more users on the platform
  • For marketplaces, what matters is users in the specific markets you care about, not users in the abstract
  • For collaboration tools, it's users within your team/company, not total users
  • Network effects are largely present or absent from inception and very hard to manufacture later

“in the case of collaboration tools it's not the more users in abstract terms it's the more users within your team”

Yuriy Timen · 16:00
#network-effects#virality#collaboration#marketplaces
Explainer18:00

The 2-of-3 Test for Whether to Invest in SEO

Before committing to SEO, Yuriy runs a back-of-the-envelope audit across three angles: a unique editorial angle, a unique programmatic angle (Canva's templates, Zillow's listings), and a unique data angle (a Mint-style tool with connected-account spending data). If you can check two of the three, you may be in good shape — then focus on lowering the cost of experimentation and time-boxing to three months.

  • Three SEO angles: unique editorial, unique programmatic, unique data
  • Programmatic examples: Canva templates, Zillow / real estate listings
  • Data angle example: Monarch Money turning connected-account spending data into search experiences
  • Checking 2 of 3 boxes is a signal you may be in good shape
  • SEO has a long return horizon — earliest results ~6 months, so time-box exploration to three months

“if you can check two of two of three of those boxes right as like a back of the envelope framework you may be in…”

Yuriy Timen · 19:30
#seo#organic-growth#programmatic#experimentation
Explainer49:00

Conversion Benchmarks for Prosumer Freemium Products

Yuriy shares category-dependent conversion benchmarks for prosumer freemium tools (Grammarly, Canva, Whimsical style): a healthy site-visit-to-free-signup rate is roughly 20-35%, higher (40-50%) for a narrow strong-PMF audience but asymptoting to 25-30% as you broaden. On free-to-paid, anything under 5% won't sustain an independent company long-term — you want north of 5%, ideally north of 7%.

  • Benchmarks are category-dependent, not case-by-case
  • Visit-to-free-signup: healthy ~20-35% for prosumer freemium
  • Narrow strong-PMF audiences can hit 40-50%, asymptoting to 25-30% as you broaden
  • Free-to-paid under 5% won't work long-term regardless of top-of-funnel size
  • Target north of 5% free-to-paid, ideally north of 7%

“it's probably like you got 20 to 35 range from landing on the site to signing up”

Yuriy Timen · 49:30

“i think anything under five percent is not gonna it's not gonna work long term regardless of how big your top of funnel is”

Yuriy Timen · 50:00
#conversion#benchmarks#freemium#prosumer#metrics
Explainer50:30

Why Onboarding Is Almost Always a Big Opportunity

For robust prosumer products (Airtable, Whimsical, Canva Video) that democratize professional-grade work, the danger is dropping a new user straight into a complex editor. Yuriy says onboarding that gathers relevant info up front and customizes the first experience is almost always high-ROI — early-stage teams can 2-4x activation, and even later-stage (Series B+) can still get 20-30% lift.

  • Complex products risk losing users who land in a bare editor
  • Good onboarding collects up-front info and customizes the first experience
  • Democratizing pro-grade work means hiding the product's full robustness at first
  • Early-stage teams can 2-4x activation rates through onboarding
  • Later-stage (Series B+) can still get 20-30% activation lift

“i mean you can 2 to 4x activation rates easily through onboarding”

Yuriy Timen · 53:00
#onboarding#activation#product#prosumer

Tool· 3

Tool21:30

Buy a Cheap SEO Audit Before You Commit (and Who Yuriy Rates)

SEO is a specialized dark art where the basics are known but the winners are 'determined between the lines' via algorithm-tracking nuance. Yuriy's advice: get an inexpensive outside audit ($5-10k) that would take you many human hours to replicate, then decide whether to hire help. He names Ethan Smith of Graphite as a first-principles SEO thinker, though it's a high-end shop not suited to a Series A.

  • Basics (best content wins, no shady backlinks, clean on-page and crawlability) are table stakes
  • Winners are decided by keeping up with nuanced algorithm changes — outside specialists help
  • A boutique audit runs ~$5-10k and saves many human hours
  • Ethan Smith / Graphite recommended as a top SEO thinker (high-end; not for early-stage)
  • Graphite is automating a lot of its work into tools/betas

“for 5 to 10k would take you many many human hours to try to pull together yourself”

Yuriy Timen · 22:00

“one of the most innovative disciplined first principles seo thinkers and i've met is ethan smith from graphite”

Yuriy Timen · 22:30
#seo#tools#agencies#graphite
Tool28:00

Why Media Mix Modeling and Incrementality Are Making a Comeback

With iOS 14 breaking click-based tracking, Yuriy says 1950s-era media mix modeling is returning for the digital age, led by Recast — useful if you spend $100k+/month across 3+ channels. He argues click/cookie attribution only ever showed correlation, never causation, and that true incrementality testing (turning a channel off in a market) is the only way to prove causal lift — pointing to Measured and INCRMNTAL.

  • Click/cookie/URL-parameter attribution only ever proved correlation, not causation
  • Media mix modeling is returning as channel portfolios go multi-channel and less trackable
  • Recast leads the modern media-mix-modeling approach — best for $100k+/mo spend across 3+ channels
  • Only randomized incrementality testing proves causal lift
  • Measured (measured.com) and INCRMNTAL address incrementality testing

“click based and url parameter based attribution it never demonstrated a causal relationship between our media spend and business results it was only good for…”

Yuriy Timen · 30:30
#attribution#incrementality#media-mix-modeling#tools#paid-acquisition
Tool59:30

Yuriy's Book Recommendations

In the lightning round, Yuriy names three books that stuck with him: Essentialism (on cutting noise and finding singular focus — he bought copies for 40+ people at Grammarly), Man's Search for Meaning (you can't control what happens, only your reaction), and The Splendid and the Vile by Erik Larson (Churchill in 1941, which he read for its parallels to the war in Ukraine).

  • Essentialism — cutting out noise and finding a singular focus; a career game-changer for him at Grammarly
  • He bought Essentialism for 40+ people on his team
  • Man's Search for Meaning — you can't control events, only your reaction to them
  • The Splendid and the Vile (Erik Larson) — Churchill in 1941, with parallels to Ukraine

“it's a book called essentialism i forget the author's name i think his last name is mccown or something”

Yuriy Timen · 1:00:00
#books#recommendations#essentialism#leadership

Takeaway· 3

Takeaway13:30

When Virality Is Worth Betting On

Yuriy's first test for virality is whether inherent product network effects exist, since he believes they're nearly impossible to engineer after the fact (Airbnb, Dropbox, Airtable have them; Grammarly didn't). But he notes referral loops can still work without network effects if the product and brand are genuinely beloved, citing an Australian fresh dog food subscription running a successful give-one-get-one program.

  • First look for inherent product network effects — they're there or they aren't from inception
  • Grammarly lacked network effects because written communication isn't inherently multiplayer
  • With network effects, layer on referral and viral loops (Dropbox file-sharing is the iconic example)
  • A beloved premium brand can run incentivized referrals even without network effects

“the first thing you look for is that is there inherent product network effects it's something that it's either there or it isn't from inception”

Yuriy Timen · 13:30
#virality#referrals#network-effects#growth-loops
Takeaway32:30

The Thing Worse Than a Channel Failing: Never Giving It a Fair Shot

Asked whether to focus on one channel or try many, Yuriy prescribes focus paired with rapid iteration and guardrails so you know exactly when to move on. His recurring warning: companies constantly declare a channel 'doesn't work' when they never set it up to have a real shot in the first place.

  • Neither extreme focus nor spreading thin is right — focus with guardrails
  • Extreme focus risks wasting scarce runway if the one bet is wrong
  • An overloaded growth person tries five things and does none of them fully
  • Companies routinely misdiagnose a channel as broken when it was never set up to succeed
  • Set explicit guardrails so you know when it's genuinely time to move on

“the only thing that's worse than a channel or a tactic that you tried not working the only thing that's worse than that is when…”

Yuriy Timen · 34:00
#prioritization#channels#experimentation#focus
Takeaway58:00

Offline Channels Are Getting a Second Look as Digital Attribution Erodes

Yuriy sees out-of-home, direct mail, and podcasts gaining traction. The old knock on offline — poor attribution versus digital — has narrowed from both sides: offline channels improved their measurement while online attribution deteriorated (iOS 14). He calls podcasts especially performant, and cites an investor anecdote of a 10:1 ROI on an at-home campaign.

  • Out-of-home, direct mail, and podcasts are underrated and gaining traction
  • Offline was historically dismissed over weak attribution
  • That gap narrowed: offline measurement improved as online attribution deteriorated
  • Podcasts are very performant for a lot of brands
  • An at-home campaign delivered ~10:1 ROI in one cited example

“and podcasts especially are actually very very performant uh for a lot of a lot of brands”

Yuriy Timen · 58:30
#offline#out-of-home#podcasts#direct-mail#attribution