Similarweb

Trust-First Activation

Testing whether contextual permission requests
could improve user quality and retention.

Introduction

BrandBee’s activation flow had a business-critical goal: users needed to enable Accessibility permission. This permission allowed Similarweb to collect behavioral data even if the user did not return to the app immediately.

Internally, this made permission opt-in a key definition of active usage. From a business perspective, that logic made sense. If a user enabled the permission during onboarding, BrandBee could immediately start creating value for Similarweb’s data model.

But from a product experience perspective, the definition was incomplete. A user who grants permission and never returns is not necessarily a healthy activated user.

Problem Statement

Rewards apps already trigger suspicion. Users wonder whether the app is legitimate, whether they will really get paid, and why sensitive access is needed.

Asking for Accessibility permission too early created a trust problem. Users were asked for a heavy commitment before they fully understood the product, the value exchange, or why Similarweb needed that access.

The question became: could BrandBee move from permission-first activation to trust-first activation without losing the business-critical permission opt-in?

Research

The hypothesis was based on several signals.

Onboarding funnel metrics showed user drop-off around sensitive activation moments. Public reviews and web feedback showed that users were anxious about legitimacy, payout, privacy, and whether the app was worth their time.

The team needed to validate whether a clearer value proposition and better timing would improve permission quality and retention.

The core hypothesis was: if users understand BrandBee’s value proposition before being asked for sensitive access, they may enable permissions with higher intent and retain better over time.

Design Process

We tested two onboarding models with 50 new users.

  • Control group: 25 users were prompted for Accessibility permission early in onboarding.

  • Variant group: 25 users completed a faster onboarding first and were asked for permission later, in context of their actions.

The goal was not only to compare immediate permission opt-in. The more important question was user quality: would users who received the permission request later, after gaining context, retain better?

Solution

The variant reduced upfront friction and reframed the permission request around user understanding.

The design approach included faster onboarding, clearer microcopy, a more transparent explanation of BrandBee’s value exchange, a better explanation of Similarweb’s role, and permission prompts connected to user actions rather than appearing as an abstract onboarding gate.

The goal was not to hide the sensitive ask. It was to earn the right moment to ask it.

Impact

The results supported the trust-first hypothesis.

  • The control group enabled permission earlier, but retained worse.

  • The variant group retained better.

  • The variant group still enabled permissions later at a meaningful rate.

This suggested that immediate permission capture was not always the healthiest activation signal. Context and timing could improve user quality, even when the business-critical permission was not asked for immediately.