How the AI growth funnel is changing: acquisition and onboarding are blending

Forcing users to sign up first could be hurting your growth. Here’s what to do instead.

Get users in the door first, show them value after.

In growth, we’ve long operated with this playbook. We even built the AARRR funnel (Acquisition, Activation, Retention, Referral, Revenue) around this logic. Acquisition precedes activation, and the “aha moment” is gated by a sign-up flow; only then does activation happen.

But that leads to the question: What does activation really mean?

During my years working on growth at Meta, I spent months trying to figure that out. We ran complex models, debated metrics, and eventually, this became a political battle of true data signal vs. representative business definition.

But AI is changing the rules, fast. The fastest-growing AI products don’t wait to show value, they deliver it right away, often before users even sign up. As a result, activation now comes before acquisition in many cases.

Here’s what that looks like in practice:

• Perplexity and ChatGPT let you ask questions before signing up.

• Talkie.ai allows you to start chatting instantly.

• Lovable and other AI vibe code generators lets you start the creation process before asking for credentials.

• Finch and Tolans, among other voice-first productivity tools, personalize onboarding through conversation, blending it with the sign-up experience.

In this scenario, forcing users to fill out a form before experiencing any value is an approach that will inevitably fall short. Below, I’ll detail the trends I’m seeing in the AI growth funnel and list three concrete measures you can implement to ensure your company doesn’t lag behind.

What has changed?

After consulting with dozens of companies in the last few years, the main trend I’m seeing is a funnel that compresses itself to the point where acquisition and activation often happen simultaneously. In this scenario:

1. Users expect value immediately

Delivering value to match intent used to be a best practice for SEO landing pages, but now, it’s expected across all channels. Whether traffic comes from search, social, or direct, users want to start engaging immediately.

Growth from top-of-funnel to mid-funnel happens in hyper speed, and Time To Value (TTV), or how long it takes your new customers to get value from your product, gets compressed dramatically as users can get to the “aha moment” as soon as they start interacting with a product.

Example: Rows’ homepage lets you interact with the product right away.

2. Sign-up has to be earned

Users won’t give away their email just because you asked nicely. If they do, their expectations are much higher and activation has to happen fast. Now, sign-up has to feel like the logical next step in the user’s journey. A few ways to design for this include delaying it, making it optional, or tying it directly to something valuable, like saving progress or accessing specific features.

Example: Perplexity’s in-line sign-up flow

3. Onboarding no longer gates activation

The long onboarding tasks that once delayed activation are now being compressed or eliminated entirely. AI products are doing more of the heavy lifting by using AI interfaces that allows users to use zero shot or few-shot prompting to get an outcome.

Zero-shot prompting means the user can just ask AI for what they want without providing examples, like “Create a presentation on the history of photography for a high school audience”. The AI interprets the user’s request, generates the output, and handles the setup behind the scenes, removing the need for manual configuration. This allows users to experience highly personalized outcomes even while skipping long to-do checklists, extensive tutorials, and gets them to value faster.

Example: Gamma zero-shot prompt for slides

4. Pre-login behavior is now critical

Since users engage with the product before creating an account, their anonymous sessions hold critical data about intent and engagement. You need to factor this data into your analytics setup so that you can associate it with the same users and accounts later on to avoid losing key behavioral data or attribution. This extends the conversion window which strains a traditional measurement funnel, especially for traditional product-led sign up flows.

5. Landing pages are now product surfaces

For PLG companies, a single page now needs to serve multiple jobs: educate, demonstrate, engage, and convert. AI-first onboarding means the landing page is part of the product experience rather than just a marketing surface Landing pages with the ability to convert intent into action let users interact with the core value of the product immediately. The page becomes the start of the user journey, and the acquisition and activation metrics will become blurred.

Example: ElevenLabs AI voice generator demo

How growth teams can design the new funnel

So how do you design your product onboarding for this new funnel, where activation comes first and sign-up follows?

1. Define what is part of the preview

Start by deciding what users can access before creating an account.

Allow them to see what's possible, without giving away unlimited repetitive value over time. The goal is to reach the "aha moment” fast while, at the same time, establishing the need to return and build habit formation.

Take the example of products like ChatGPT or Perplexity. They let users run queries without logging in, but there are limits. If you want to save your history or use better models, you need to create an account. The same goes for Talkie.ai - you can talk to one bot for free up to a few interactions, but continuing to chat, accessing more characters, or building your own, requires signing up.

Example: Talkie AI sign-up starts after the 4th interaction with an AI chat bot.

Previewing helps users understand your value, but also sets expectations about what’s free and what’s not. And once you get that balance right, people will want to create an account, instead of feeling forced to.

2. Build mini product flows into acquisition

We are all familiar with the traditional landing page structure: top fold with value proposition, feature modules, social proof, and finally a CTA trying to get users to sign up. But once they click it, they're hit with a registration wall… before experiencing any value.

That approach no longer works.

AI products reimagine this by integrating product functionality directly into the acquisition flow. So, don’t tell users what your product does — let them experience it right away.

Think of code generation tools like Lovable. Instead of showing screenshots of what the dashboard looks like, they let users engage with the interface right there on the landing page. Users can see example prompts, enter their own, and start creating — only then does the sign-up flow appear.

This turns acquisition into part of the activation flow, and sign-up into part of the onboarding journey.

3. Test the impact on retention

Give away too much value and users never sign up. Give away too little and they never return.

It’s a fine balance to strike, and a dilemma close to the freemium model, creating a complex mix of users: serious prospects, AI tourists, and opportunistic freeloaders. The right approach is found only by testing.

Perplexity tested extensively and found that the ideal moment for a sign-up gate was after a few searches. The sign-up appears within the chat interface, creating a natural transition when users understand the value but still want more.

You’ll need to run experiments to find the answers to questions like:

• How long should users explore before hitting a gate?

• When does sign-up feel like a step forward instead of a wall?

• What data do you really need right away, and what can wait?

The goal is to make sign-up feel like a natural progression rather than an arbitrary barrier. Experiment in both directions until you find the right balance between acquisition and retention.

Conclusion: earn sign-ups through activation

In the AI era, the best products earn sign-ups by giving users a glimpse of activation first. Products that hide their "aha moment" behind registration walls will increasingly struggle to compete with more frictionless alternatives.

The new growth funnel doesn't follow the linear AARRR sequence. Instead, it's becoming a fluid experience where users can experience value, create accounts, and deepen engagement in whatever order makes the most sense for their journey.

The best growth teams are already adapting. If you're exploring how to apply this in your own product and still struggling, feel free to reach out - I'd be happy to help.