Turning AI into revenue: the monetization strategy that boosted profits by 40%

Company
Jusbrasil
Year
2025
Status
Shipped

The problem

Jusbrasil launched a new AI product to establish itself as an AI-first company in the Brazilian legal market.

The challenge was to find a way to monetize AI without breaking user trust or interrupting the value users were getting from it.

The signal

AI created value differently from our subscription, so I analyzed beta testers' behavior and sessions to find where the new AI product added real value.

  1. 01

    Legal drafting

    Users experienced value immediately after generating a useful legal draft.

  2. 02

    Legal research

    Completing a research task was another strong moment of perceived value.

AI delivers the output

Legal drafting
Legal research

Illustrative — each row is one beta session, each mark one tracked event.

Behavioural data and session recordings from beta testers: the value moment landed after the output — in drafting and in research.

The opportunity

We could monetize the moment users understood the value.

Instead of placing AI behind a generic paywall, we explored a contextual model and set usage limits based on how the beta testers behaved.

Generic
User asksblockedPaywallpays on a promiseAI answers
Output-based
User asksgeneratesAI answers + limit usagepays on evidencePaywall
Instead of charging before delivering value, we moved the paywall to the moment users had already experienced the AI's value.

Exploration

I explored how AI products connect value with monetization.

01

Benchmarking research

I benchmarked Notion AI, Grammarly, Claude and ChatGPT to understand how they positioned upgrades around AI usage.

The benchmark board in Figma: four labelled groups of product screenshots — Notion AI, Grammarly, Claude and Chat.gpt — each a captured upgrade flow wired together with red connectors from the first prompt through to the plans page.
The benchmark study was inside Figma, featuring screenshots and buying workflows.
02

Buying journey flows

I then mapped possible buying journeys to explore different upgrade moments and align the strategy with stakeholders.

01At sign-up02Before the answer03After the output limit04On the next taskPaywallpicks a planPlanscheckoutSubscribeddismissedContinue free

Dashed — left on the table. Solid — the one taken forward.

Four candidate triggers, one purchase path — the buying journeys mapped side by side to compare upgrade moments with stakeholders.

The decision

We chose the output-based trigger in a joint internal decision.

The trade-off was to absorb the cost of generating the AI output upfront, in exchange for a more contextual upgrade moment tied to the value users had just experienced.

Iteration

I kept the trigger contextual — and adapted the experience around it.

01

Output-based paywalls

The upgrade appeared immediately after a high-value AI result.

Triggering the paywall after the usage limit is reached.
The Jus IA chat with a due-diligence answer on screen and an inline card below it reading 'Você chegou no limite de respostas', with an 'Assinar agora' button.
The trigger paywall in context.
02

Payment modal

I designed the payment modal to adapt to different use cases, tailoring the experience for trial, upgrade, and winback scenarios.

The payment modal was redesigned to fit new benefits and new tier plans.
03

A new pricing structure

We repositioned Jus AI as a premium feature within the existing subscription tiers, making the model clearer and more scalable.

Jusbrasil's plans page after the repackaging: Essencial, Profissional, and a recommended Premium tier, each including Jus IA, with a monthly/annual toggle.
The plan repackaging — three tiers built around the AI, with Premium set as the recommended plan to lift ARPU.

Results

+40%
ARPU growth
+5,000
new subscriptions in the first weeks
+20k
monthly active subscribers after four months

The strategy turned a new AI capability into a meaningful revenue driver while keeping monetization connected to the value users experienced.