Spend the context twice

Welcome to Product Cocktail, where the takes are as polarizing as a shot of Fernet—but the insights come together like a perfectly crafted daiquiri.

A customer walks into your bar and tells you they like a refreshing, not-too-sweet summer cocktail with prosecco. You're recommending an Aperol Spritz before they finish the sentence.

Linktree asks you to start a free trial promising "the full power of Linktree" — to someone who's never seen any of it.

On step two of the sign-up flow — after providing an email and verification code — you're presented with a "Claim a 7-day Pro trial" screen.

You’ve never built a Linktree profile before. You don’t know what Pro means. You don’t even know what Pro costs.

The five star “Trusted by 70M+ creators” line attempts to substitute social proof for product proof, but it disappears into the abyss right above the CTA. (I genuinely didn’t notice it until ChatGPT pointed it out.)

It's a well-built paywall (10 points to Gryffindor for transparent trial terms), it's just in the wrong place. This is commitment before comprehension.

Three months ago, a recruiter pitched me this role:

Linktree already knows the first half of the answer. "Generate an optimized profile from public signals" is right there in the job description. I built the half they left out.

AI onboarding's first job is to create value faster. Its second is to use what it learned to make the paid value concrete.

Generated Linktree would use public creator signals to produce a credible first version before asking the user to configure anything. That does more than shorten onboarding. It reverses the order of the funnel.

Instead of learning a new tool before seeing its value, creators would begin with a populated profile they could refine. That gets them to value faster — and closer to the behaviors Linktree ultimately monetizes, including shops, courses and digital products.

Arguably, they can go even further.

Generated Linktree could go even further by using the customer’s data to improve how they merchandise their paid plans. The price does not need to change, but the explanation should.

Remember our customer earlier? Once Generated Linktree scrapes her socials, it knows she likes prosecco — and a generic paywall makes her read the whole menu anyway.

If you’re taking on the cost to build something for the customer for free, you might as well use that data to sell them your product more effectively. There’s no need for your paywall to sell a (probably) complicated abstraction with a one-size-fits-all comparison table that’s been A/B tested to death.

If your customer goes through the trouble of pulling in their website and social tags before they’ve even finished your acquisition funnel, they’d be handing you a personalized merchandising strategy on a silver platter.

The same social media scrape that tells you their color scheme and top links also tells you that they’re a mid-tier travel content creator. From there you can infer their goals and recommend the best possible plan to help achieve those.

Their visual identity makes Pro’s enhanced themes immediately demonstrable. Their audience size and outbound links suggest that advanced analytics and mailing-list integrations may be relevant, as opposed to the free tier’s basic functionality or Enterprise’s overkill — showing exactly what they stand to gain by converting to that pricing tier.

The recommendation should also show its work, explaining to the user why we landed on that particular plan. (Loyal readers will remember this classic AI trust and transparency beat.)

Linktree isn't the first to the party with AI onboarding. Others beat them to the punch, but no one has built the monetization payoff.

Intelligent onboarding flows are beginning to replace questionnaires or generic templates with inferred or generated initial states. There are a few different flavors.

The first type of intelligent onboarding is what Shopify launched in mid-2025 with “AI Store Builder”: declared context. This feature lets would-be customers describe and generate their store in a single prompt. (It broke when I tried to test it.)

The second is a structured questioning style that feels much less static than a traditional onboarding Q&A. Wix uses an AI-fueled interactive survey to build a website — but then drops you onto a boring, unpersonalized task list. (If you can find your way to their web editor, you’ll find a pre-populated prompt you can run to build an initial site based on the info you’ve provided).

The third flavor is closest to what Linktree is proposing: inferred context. Linktree direct competitor, Beacons, already uses an AI onboarding flow, pulling social data to preload everything for a v1 of your link in bio page. They even give you multiple (fairly mid) design options to choose from.

The same pattern is appearing in B2B, usually through imported company context. HubSpot and Intercom can use a customer’s website and existing materials to prepopulate a workspace, turning onboarding from construction into correction.

Each one solves the blank canvas problem, but dumps you onto a generic paywall. The missing second half of AI onboarding is universal.

Most products probably won’t do this. They’ll implement generated onboarding and a generic plan grid or free trial interstitial.

That still improves activation, but it squanders your best context at the one moment it matters most: when they're deciding whether to hand over a card.

A paywall is an intentional choice. Start treating it like one. You’re asking the customer to value you, to trust you, and to commit to you. Placing a free trial paywall early in your acquisition flow may increase trial starts while producing customers who understand the product less clearly, making the top of funnel look better without improving trial-to-paid conversion, early retention, or lifetime value.

There is adjacent evidence that intent-matched paywalls outperform generic ones. Adapty assessed the impact of Apple Ads’ paywall personalization feature across 8,000 apps and saw a 24%+ lift in trial-to-paid conversion when using tailored paywalls.

Different mechanism, same principle. Match the offer to intent and it converts better.

Organizations who do this right will reuse that context to personalize how the product is sold to customers, anticipating objections and meeting them with a show-don’t-tell approach to merchandising.

Generated Linktree solves the blank-canvas problem. I built the missing second half: a paywall that reuses the same context to match the right plan to the right customer.

Here’s the demo: https://tobrien88.github.io/generated-linktree/

Compare Linktree's existing paywall to mine. Generic grid-based paywall with the same copy for every customer vs. personalized, context-aware merchandising that demonstrates the subscription value on top of your own content.

For the demo travel creator, Pro is not presented as a bundle of “advanced customization” and “analytics.” The paywall shows how her existing travel photography could become the profile theme, why her audience size makes deeper analytics useful, and which capabilities disappear on the cheaper tier.

The prototype keeps Linktree’s plans and prices intact, but it changes the positioning to clearly demonstrate what the creator would gain or lose at each tier.

The paywall stops being a dense feature matrix and becomes a mirror, reflecting back what Pro tier can do for this profile.

If I was the Staff AI PM at Linktree, right now I’d be running a test with customers. I’d create three variants of my prototype and see what customers responded to (holding constant prices, trial terms, plan availability, and paywall timing for scenarios 2 and 3):

Comparing the second and third variants isolates the core question: does reusing onboarding context at the paywall improve conversion beyond the activation benefit of generation itself?

Most teams will use AI onboarding to shortcut setup work, then send every customer to the same generic pricing grid. That is a missed opportunity. Once onboarding knows enough to build the product, the paywall should be able to explain why it’s worth paying for.

Your move, Linktree.

Send me questions, feedback, and cocktail recipes:
[email protected].

_{Icons made by Icongeek26 from www.flaticon.com.}