Account Structure - Tracing Symptoms to Their Source

Context

Covenant Eyes provides an accountability service that helps people quit pornography through monitoring, reporting, and trusted relationships.

Built for families, but the audience shifted

Our account structure was set in 2013 and optimized for families — supporting multiple users per account, shared subscriptions, and parent and child roles. Since then, our strategic audience has shifted to individuals. Families represent roughly 10% of subscribers, but the structure still carries all the complexity of that original design.

The cost of not fixing it is rising

Signup confusion, billing issues, and duplicate accounts have persisted for years. But we’re seeking to grow our user-base, and can’t have these problems scale linearly as we do.

We are also replacing our free tier with a 7-day trial, which means time-to-value now directly affects conversion. If people get stuck setting up their account, and aren’t 100% convinced to keep the service by day 7, they’ll leave.

Frame

Same root cause, different symptoms

Signup, onboarding, billing, and various support issues often traced back to the same thing: how users, accounts, and subscriptions relate to each other — structural relationships nobody had reexamined in over a decade.

Starting from problems people already cared about

My early attempts to comprehensively audit roles and permissions weren’t very successful. Without a clearly framed problem, the team couldn’t align.

So I shifted my approach, starting from current problems that were clearly backed by member feedback, then tracing those symptoms back to their structural causes.

That reframing changed everything.

Crystallize

Email non-uniqueness

Most services require one email per account. We don’t, because young children don’t always have an email address, so we let parents reuse theirs.

That was reasonable at the time (~2013), but the cascading effect has been significant:

• duplicate accounts
• misdirected communications
• compounding confusion

The problem is not yet solved, but it’s now clearly scoped.

Children as authenticated users

Parents must set up each child as a fully authenticated user — username, password, security questions, email — regardless of age.

This creates administrative overhead and is the direct cause of the email reuse problem described before.

With the help of our CEO, I traced this decision to its origin and challenged whether it still made sense. We now have traction around an alternative where devices, not children, become first-class objects in the system.

Subscription sharing contradicted our direction

We wanted one user per account, but subscription sharing would require consolidating users onto shared accounts.

By modeling the structure first, the team could see the contradiction before anyone built anything.

We pivoted to simplifying roles and permissions instead.

Sustain

Sequencing the path forward

I worked with PM to propose a sequenced approach: role simplification first, subscription changes later, preserving family-focused functionality in the interim. This gave us a way to make real progress without requiring the entire future state to be resolved at once.

A North Star to orient around

I presented a vision — one user, one account, with devices taking the place of child users — to Member Care leadership and the full department.

I did not present it as a finished plan, but as a way to contrast our current and future states. People shared their concerns, questions, and ideas for how to make it better, which was the entire point.

This work benefits from transparency.

Results

Role simplification became the team’s priority instead of subscription sharing — a direct result of structural analysis revealing the contradiction.

Product management arrived at a sequenced, grounded direction.

And we built organizational belief in a future state by involving people in the progression from the start.

We’re not done. But we have real traction for the first time in years.