Audience-Building Playbooks
Building an audience before, during, and after launch — the platforms that compounded, the posting cadences that worked, and how founders turned followers into first customers.
198 tactics · page 5 of 7
“Once we learned to actually be like 'no — why are you here, what are your aspirations?' and view things through that lens, that's been one of the biggest unlocks.”
Segment by user aspiration, not just behaviour — blended metrics hide your best persona
For years Rapchat looked at retention and engagement across their entire user base without distinguishing between someone who opened the app for fun and a serious aspiring artist. Once Seth Miller's team started asking 'why are you here?' at onboarding and segmenting by declared aspiration, they discovered their core monetisable persona had retention metrics far above the blended average. That clarity focused the whole product roadmap.
“Super users go one step beyond great customers: they're evangelists who bring in other members, give you feedback on your products and services, and are willing to help onboard new members.”
Super users invest their own resources to strengthen your model — find them early
Beyond high-LTV customers, a subset actively recruits for you — making referrals, answering questions in user groups, and putting their own time into the community. Robbie observed this with Crossfit members who set up equipment in cul-de-sacs when their gym closed. Designing referral programs, community features, and feedback loops around these people compounds their value exponentially.
“You look at the group of customers who stay the longest and spend the most — develop hypotheses about what this group shares — and then you tell your marketing team to go get lookalikes.”
Segment by post-signup behavior to find your highest-LTV customers — then market only to lookalikes
Start by listing your best and worst customers and identifying what separates them: acquisition source, signup cohort, onboarding path, usage frequency in week one. Once you have a hypothesis, build marketing to attract only the high-value archetype even if it means a smaller TAM. Disciplined focus on the right customer is a competitive advantage.
“The challenge with with that segment is that there's always these you know really esoteric and extreme product requirements... a lot of them are kind of living the van life... they don't want to really pay you any money either it's like this isn't a good growth segment.”
Your Loudest Users Are Not Your Largest Market — Target The Hesitant Majority Instead
AllTrails' 2015 rebrand moved away from hardcore through-hikers (who demanded extreme features and rarely paid) toward the 'my wife' persona — someone who enjoys outdoor time with family but lacks confidence to go alone. That mainstream-hesitant segment is hundreds of millions of people vs. hundreds of thousands of van-lifers. Vocal power users can misdirect your roadmap toward a segment that will never drive revenue.
“With the custom product features you can have up to 35 versions of your app store uh which means you're like if you're a you know wellness app uh if let's just say you're like a meditation app that has uh meditations for sleep or anxiety and how to meditate you can have a separate landing page so to speak on the app store for sleep anxiety meditation right.”
iOS 15 Custom Product Pages Enable Audience-Specific App Store Experiences
iOS 15's custom product pages give apps up to 35 unique App Store URLs with distinct screenshots, preview videos, and promotional text. A meditation app can send sleep-ad traffic to a sleep-focused App Store page and anxiety-ad traffic to an anxiety-focused page — dramatically improving message-match between ad creative and the store experience. This partially recreates the targeting precision lost to ATT by matching the store presentation to the audience's intent.
“The next centric circle outside of dyslexia was ADHD and then low vision and autism and concussion and anxiety and second language learners. Today that's about 25% of our user base — 75% are neurotypical people: doctors, lawyers, accountants, people in the military, executives.”
Concentric-Circle Audience Expansion — Start Niche, Grow Outward in Rings
Speechify started with dyslexia, then expanded ring by ring to adjacent groups with similar needs — ADHD, low vision, second-language learners — before eventually reaching neurotypical professionals. Each new ring was already primed by word-of-mouth from the previous one. This planned concentric expansion let Cliff keep deep product focus while systematically broadening the addressable market without losing the origin story.
“Every single person who comes on a call with me has already binge-watched the YouTube channel. It's as if it's our third call. When your time is very valuable and constrained, everything you do has to be serving — and this is about hiring.”
YouTube Channel as Recruiting Funnel — Candidates Arrive Pre-Sold at "3rd Call" Level
Cliff's personal YouTube channel wasn't built for marketing — it was built to document company-building philosophy and attract operators who already deeply believe in the mission. By the time a candidate gets on a first call with Cliff, they've watched hours of content and feel like they know him. This compresses the trust-building phase of hiring from months to a first conversation, a compounding advantage that gets more valuable as the company scales.
“As seen on TV is a real thing. As seen on Instagram is not a real thing. Five bucks gets me on Instagram. Five bucks does not get me on TV. The barrier to entry was higher — people thought this is legit.”
"As Seen on TV" Trust Signal Can't Be Faked for $5 — Barrier to Entry Creates Credibility
Ryan Beck identified a key difference between broadcast and digital advertising: the high cost and gatekeeping of TV creates an implicit trust signal that cheap digital cannot replicate. For a faith app targeting audiences where trust is everything, that broadcast credibility was worth a premium. Post-IDFA, this logic has become even more relevant — as digital attribution degrades, brand-building channels with inherent trust value become more defensible investments.
“We have historically been a very anonymous-based site — weather really didn't require our user to have a significant value exchange with us for a long time such as providing an email address. As that begins to evolve we're able to lean into improved targeting and understanding how best to package the subscription product.”
Anonymous-First Audiences Limit Subscription Targeting — Solving Identity Unlocks Growth
Weather.com grew for decades without asking users for an account, which meant no first-party identity data for personalization or subscription targeting. Moving toward signed-in experiences — even softly — unlocks better segmentation, targeted paywalls, and lifecycle emails. For any app that acquired large anonymous audiences, solving the identity layer is the high-leverage next step before any paywall optimization.
“Someone made this video before we ever reached out to them. It wasn't about Opal — it was just about the space. They wrote back and said actually I use Opal and I love it. We said can we use this video as an ad and they said yeah sure.”
Mission-Aligned Creators Who Already Use the Product Outperform Agency Creative
Opal's best-performing ad was filmed by a creator who had never been paid by the company. It didn't even show the product — it captured the mission of reclaiming time from your phone. Opal repurposed it with minor tweaks and it dominated their campaigns. Finding creators who already use the app and care about the problem proved more reliable than agency or contractor creative.
“For me Performance Marketing doesn't go without the brand. You will catch some low-hanging fruit, you'll activate two to three acquisition channels and you will see indeed users coming in and they will be instantly profitable — but you won't go very far with that.”
Performance Marketing Always Hits a Ceiling — Brand Is the Only Way Through
Performance marketing yields fast, measurable wins early on, but every channel eventually saturates and CPAs climb. Shireen Khi at Deezer argues the only way to keep scaling is to invest in brand: brand awareness fills the top of funnel so performance campaigns have warm audiences to convert. Without brand investment, you're fighting over the same shrinking pool of high-intent users who already knew you.
“All the campaigns that we do that are more a performance-like campaign optimizing on installs — we knew that it had an impact that was not so easy to measure on our performance campaign but we did not imagine how much. It really confirmed the absolute necessity to keep investing on upper funnel campaigns even though they won't have short-term impact.”
Incrementality Tests Revealed Upper Funnel Had Far More Impact Than Assumed
Deezer ran geolift tests — shutting down specific regions — to measure the true contribution of upper-funnel campaigns to paid subscriptions. The results validated assumptions but at a much larger magnitude than expected. Upper-funnel spend was generating significantly more downstream conversions than last-click attribution credited it with, a common blind spot for apps relying solely on MMP data.
“Every quarter every semester we ask a panel of our audience: does Deezer know music, does Deezer understand me, do I feel close to Deezer — and we especially focus on the paid consideration part. What's actually reliable is that we follow the trend. Maybe it does not measure super precisely the actual level but what we want to see is an upward trend.”
Measure Brand with Paid Consideration Trends, Not Point-in-Time Surveys
Brand measurement at Deezer relies on recurring quantitative panels that track paid consideration — not downloads or impressions. The key metric is directional trend, not absolute score, because the methodology is inherently imprecise. Combine these quarterly panels with incrementality tests and media mix modeling to triangulate whether brand investment is moving the needle on subscription intent.
“Nobody really believed that the purple heart would be the winning version because it was so disruptive — maybe looking more like a dating app. But it came out of the studies as the winning version by far, on all target audiences. And I believe strongly that it really helped my performance campaigns to be again more profitable.”
User-Tested Logo Change Improved Performance Ad Efficiency
Deezer changed its logo from a technical equalizer graphic to a purple heart after user research: they asked panels to draw the old logo from memory, then tested new versions. The study pointed to the emotional heart — not the tech-signals equalizer. Post-launch, ad creative using the new brand identity became measurably more efficient. User-centric brand decisions don't just improve perception; they compound through every downstream marketing touchpoint.
“The concept behind it is that through incrementality testing you are able to calculate sort of a score between what you actually see as direct tracked data and what was more hidden — maybe you over-attribute to search and under-attribute to social ads that come earlier in the funnel. You use that score to modify your statistic model.”
Media Mix Modeling Needs Incrementality Calibration to Be Actionable
Raw media mix models have a known flaw: they assume a static baseline and can't fully account for cross-channel influence. Deezer's approach is to use incrementality tests — geolift, conversion lift, brand lift — to calculate adjustment scores for each channel, then apply those scores to the MMM. The result is a calibrated model that better reflects the true contribution of upper-funnel spend on final conversion events.
“When Spotify Netflix and apple when they double down on specific market for example Germany or then Brazil then several years after people there start to pay for Flo subscription or other subscriptions just because they get in this habit they acceptance that it's okay to pay for digital subscriptions.”
Spotify and Netflix Spending in a Country Trains Its Population to Pay for Subscriptions
Subscription acceptance is not purely a wealth or pricing problem — it's cultural. Countries where Spotify and Netflix have invested heavily in local pricing and marketing see subscription conversion rates rise for all apps several years later. Italy and Poland have decent incomes but historically low digital subscription rates because infrastructure giants haven't normalized the habit there yet.
“like snap it's totally different like the way to approach an app is totally different than facebook the way to approach tic tocs totally different snap right the way to approach outbrain taboola totally different than any of those you know the way to approach youtube is even different”
Post-ATT, Each Ad Platform Needs Its Own Creative Approach — There's No Facebook Template
Facebook's dominance let developers ignore multi-channel expertise because its targeting worked regardless of creative format. Post-ATT, diversifying means genuinely learning each platform's distinct creative and targeting logic — Snap, TikTok, YouTube, and Outbrain each require different creative formats, audience structures, and bid strategies. There is no one-size-fits-all approach.
“Tinder you know really went after the 18 to 24 demo and 24 to 30 but 18 and 24 especially in a way that nobody else really had and so in order to do that you had to kind of invent a couple things.”
Tinder's Early Growth: Target the Demographic Nobody Else Addressed
Match served 35+ users, OKCupid brought that to 30+, but no online dating product had genuinely served 18-24 year olds. Tinder's core product decisions — swipe mechanic, 90-second profile setup, blind double opt-in — weren't arbitrary features; they were purpose-built for a demographic with low patience for friction. The lesson: deep demographic segmentation often reveals underserved users who'll adopt with unusual intensity, and solving for them forces product innovations that become moats.
“download to paid in South Korea for Education apps is over 10%...find your Tailwinds find the places you can go where it will be easy for you to succeed and don't waste time...this can be as much as a negative guide of where not to focus”
Find Your Tailwinds: South Korea Education Apps Convert At 10%+ Download-To-Paid
RevenueCat benchmarks reveal enormous variance by geography and category: South Korea education apps convert downloads to paid at over 10%, versus a global median around 1.7%. The strategic implication is to find where your category already has favorable economics and lean in there first, rather than spreading localisation effort evenly across all markets. The same data is equally useful as a stop-list — some countries will never move the needle at your scale.
“The average realized LTV per download in North America 14 days in is four times the global average at 35 cents compared to 8 cents a multiple that exists both on the App Store as well as on Google Play”
North America LTV Is 4x The Global Average — Focus There First Before Chasing Global Scale
Hard benchmark from RevenueCat's dataset: a North American user is worth four times a global average user in the first 14 days. This shows up consistently across iOS and Android. For any subscription app in early growth, this argues strongly for North America-first before spreading to other markets — the economics are simply more favorable and the data is cleaner. Only after establishing unit economics in NA does geographic expansion typically make sense.
“The campaign is really about enforcing the positioning of our brand — it's not talking about the price, about the trial period, it's not even encouraging you to download the app. It's more really about what it means to be a user of Deezer.”
Brand campaigns should sell identity, not features — no price, no trial, no download CTA
Deezer's TV and out-of-home campaign in France contained zero conversion mechanics — no price mention, no trial offer, no download nudge. The sole objective was making Deezer synonymous with music in users' minds. Brand campaigns are not meant to convert; they are meant to create the emotional context in which conversion campaigns can work later.
“For Deezer, everything is about music and everything is about living the music. We are not positioned on all the audio — we are very centered around music and all our branding is really around this core idea.”
Own one lane, not all of audio — narrow brand positioning beats broad category claims
When Deezer competes with Spotify, Apple Music, and Amazon Music, the instinct is to broaden positioning to cover every audio use case. Shireen Khi argues the opposite: deliberately narrow to a single positioning lane so the brand occupies a clear mental slot. Broad positioning on 'audio' is invisible; owning 'music' is something users can feel.
“Every quarter, every semester we ask a panel of our audience: does Deezer know music, does Deezer understand me, do I feel close to Deezer — and we especially focus on the paid consideration part. What's actually reliable is that we follow the trend.”
Track brand health as a directional trend, not an absolute number — quarterly panels
Deezer measures brand health through quarterly quantitative panels that ask users about awareness, affinity, and paid consideration intent. The panels don't produce precise numbers, but they produce reliable directional signals. An upward trend in paid consideration quarter over quarter validates brand investment even when the absolute number is uncertain.
“One of the first things that I worked on with that client was figuring out how to get them off of the drug of paid acquisition and how to acquire more users through word of mouth or SEO or other non-paid channels.”
Over-reliance on paid acquisition is a drug — benchmarks expose organic deficit before it kills you
A client was best-in-class at Facebook acquisition but nearly all traffic came from paid channels. Blended LTV/CAC looked great but hid a critical fragility. Carter's value delivery benchmarks flagged the organic deficit before it became existential. The fix: invest in word of mouth, SEO, and content before paid efficiency peaks and forces a cold-turkey withdrawal from performance spend.
“Creative generation for ads is probably the least valuable place to apply AI. What you really care about is the concept. What I see people using AI well for is creative prospecting — pulling competitors' ads and using an LLM agent to synthesize what's working. That would have been a full-time job three years ago.”
AI's real marketing value is creative prospecting at scale, not variant generation
Generating 200 variants of the same 10 concepts adds near-zero value — the 190th variant is not meaningfully different from the first. The leverage is in concepting: finding angles that wouldn't emerge from a human brainstorm. AI-powered competitive synthesis (scraping ad libraries, agent interpretation, hypothesis generation) automates what used to require a full-time analyst.
“There's a tool called the word of mouth coefficient that lets you quantify offline word of mouth by looking at how many new users you're getting through channels like direct, branded search, or social in a given period. Subtract your returning users plus new users from non-organic channels — and that coefficient tells you how efficiently you're getting new users through word of mouth from your existing base.”
Word of mouth coefficient: quantify offline referrals by subtracting all known acquisition channels
NPS measures intent to recommend but can't tell you how many actual new users result. The word-of-mouth coefficient backs into a measurable estimate by treating anything that can't be attributed to known channels as WOM-generated. Carter used this at Quizlet to spot seasonal WOM spikes (back-to-school, exam periods) and COVID-driven country divergences — enabling proactive investment in virality during high-coefficient windows.
“We're using the answers on the quiz as conversion events for TikTok. TikTok isn't seeing basically anything happening in the app — they're learning who our user is based on the answers that they're giving us and then learning off what a good user looks like off of those conversions, which are just answering the question.”
Use quiz answers as TikTok pixel conversion events to teach the algorithm your ideal user
Stewart solved the deep-funnel attribution problem on TikTok by routing users from ads to a web quiz, then firing TikTok pixel events on specific quiz answers that correlated with high-quality subscribers. TikTok never needed to see in-app events — the quiz answer became the training signal. This technique bypassed iOS 14 attribution limits, accelerated algorithm learning, and let Stewart iterate ad spend multiple times per day rather than waiting for downstream conversion data.
“I've taken all of our almost 16,000 reviews in the App Store and very manually gone through and created value propositions and problem-solution statements using the words that are coming out of our users' mouths and how they were explaining the product. I don't think anybody in the world understands this customer better than we do.”
Mine your app store reviews to build value propositions in the user's own words
Before writing a single TikTok script, Stewart systematically mined 16,000 App Store reviews to extract the exact language users used to describe their problem and Ladder's value. This verbatim vocabulary informed every hook, every CTA, and every coach's messaging. The result: creative that resonated because it mirrored what target users already believed about their own fitness journey, not marketing language invented in a boardroom.
“The most expensive lead to acquire is casual fitness and everybody in the world is advertising to that person. If I'm attracting a lot of folks with that persona, they're not going to convert and the economics are going to break. We had to figure out how do we get dialed in on getting to the right user.”
Target 'already working out' users, not casual fitness — wrong ICP breaks your economics
Ladder's early paid spend failure came from targeting the wrong ICP: casual fitness users who responded to motivational creative but had no urgency to subscribe. Stewart identified 'already working out' users as the high-converting segment and redesigned the quiz, the creative, and the pixel signals to filter for this audience. Paid economics only work when the ICP is specific enough that creative can credibly address their exact problem.
“When they launched their new AI powered background feature — not only did they see subscriber conversion rates in the product go up significantly but they also found that by leading with those AI features in ad concepts on Meta and TikTok they got much better performance especially when they put it within the first 6 seconds of the ad.”
AI features in ads unlock new markets — PhotoRoom cracked Mexico, Brazil, and Indonesia this way
PhotoRoom had failed to achieve sustainable paid UA in emerging markets because conversion rates were too low for unit economics to work even at low CPMs. Launching an AI-powered background removal feature changed everything: leading with the AI demo in the first 6 seconds caught attention, drove higher CTR, lifted subscription conversion, and fundamentally unlocked new geographies. Carter's lesson: AI features are marketing hooks as much as product features — the wow-moment that drives organic sharing also dramatically improves paid ad performance.