The Subscription Value Loop: A Formula for Growth – Phil Carter, Elemental Growth

Talk to customers and dig through analytics with the data team to find what really drives growth: at Quizlet, six key user actions and four core growth loops, with SEO the most important. Build the simplest model that captures them (a 'minimum viable model'): ~100 variables covering acquisition, retention and monetization assumptions, split by students vs teachers and a few other segments. Keep a one-image blueprint of the actions and loops for company-wide alignment. Start US-only, then replicate per market by passing a country code into a Python script to compare growth across regions.

Frame the org around value creation vs value capture: core product teams focus on creating core value with retention as their primary metric; growth product owns value capture plus new-user onboarding and anything that connects users to core value. Later add 'value delivery' to coordinate with marketing (including lifecycle marketing around onboarding) on bringing in free users and subscribers.

Every flashcard set created by students and teachers becomes indexable content, so Quizlet wins long-tail search queries while other edtech companies fight over head terms. Pair the organic user-generated content with a distinct SEO strategy; paid ads only as a supplement.

Run for the first many years with no subscription and, for a long time, no ads; after introducing a subscription, keep prices quite low. Free users generate the content and word of mouth (a cross-side network effect between teachers and students) that drive compounding growth.

Adopt the Blinkist 'ethical subscription design' pattern in the purchase flow: be as clear and transparent as possible about how the free trial works.

Quizlet, historically focused on repetition-based learning and exam prep, acquired Slader and turned it into Quizlet Explanations (textbook explanations) to move into homework help and getting unstuck.

'Immersive onboarding': steep the user in the product while they go through the setup and aha moments (show, don't tell), rather than repeating App Store screenshots. Guest cites Duolingo (a first lesson in onboarding) and Rise (sleep debt calculated in ~2 minutes) as other examples. No detail on Quizlet's own flow.

The reading app launched Readable Podcasts: convert a podcast into a text transcript, and double-tap AirPods to pause and create a highlight from the audio transcript. It expanded the value proposition to a segment of users who wanted more.

Track paywall view rate: the share of installers who see the paywall at least once in their first session, first week and first month. Place the paywall more aggressively and optimize its copy, visuals and a number of other elements, following Superwall-style best practices.

Starting point: two tiers (Lite, Pro) each with monthly and annual plans; Pro's extra value was licensed publisher ebooks. Instead of A/B testing prices first, run a Van Westendorp survey (at what price is it too expensive, getting expensive, a bargain, too cheap to trust) to get the optimal price range, cut by tier; then a conjoint survey where each user makes 5-10 choices between 2-4 packages of attributes to get marginal willingness to pay per feature. Findings: little extra willingness to pay for Pro over Lite or for annual over monthly, and publisher ebooks had among the lowest willingness to pay. Action: collapse to one tier (Udocz Plus), drop the licensed ebooks, and lower the annual price. Guest's rule: early consumer apps should offer one plan (annual only, or monthly + annual).

Trigger a paywall in every user's first session (right after sign-up or at the end of onboarding, depending on the product), because many users don't know a premium version exists. Pair it with customized paywalls tied to your best premium value props and meter them: give a free number of uses, and when the quota runs out show a paywall built around that feature. Avoid showing the paywall on every open, which churns free users.

Find the time after install by which 90% of your subscribers typically convert (e.g. within the first 24 hours). Users who haven't subscribed by then get a discount offer; their inactivity signals lower willingness to pay. Guest also recommends geography- or demographic-based pricing and promotions with the app's own personality rather than generic Black Friday 'act now' deals.

Word-of-mouth coefficient: in a period (week or month), count new users arriving through direct, branded search or social, and divide by returning users plus new users from non-organic channels (paid ads, non-branded search). Track how it moves over time. At Quizlet and edtech clients it drops in summer and holidays, spikes in August-September back-to-school, after holidays and around exams; during COVID it fell off a cliff and then recovered differently by country. Use the spikes to time virality pushes matched to the product's type of word of mouth.

Ask customers how they would feel if they could no longer use the product (very / somewhat / not disappointed); 40% 'very disappointed' marks strong product-market fit. Re-segment the answers to find the personas most likely to say 'very disappointed', interview them about what would delight them more, and build a roadmap that fills those gaps. Segment by persona, not by feature.

Layer in promotions and sales later, when a growth team has capacity, since running them is labor-intensive; with zero marginal cost, selling at more than one price point captures demand otherwise left on the table.