t trove Onboarding brief
Decision brief · July 2026

From first click to a profile that already feels personal.

Today, Trove asks for patience and commitment before it clearly proves the value. The proposal is simpler: let a new player experience Trove, show them something true, then use login to preserve momentum.

The proposed journey, in one line
LandingPromise is immediate
→
2-minute TangleOnboarding through play
→
Real resultValue before account
→
Login to saveReturn to the same result
→
Profile 1/3A meaningful first layer
→
Best next TangleWhy this, why now
36.5%
of production landing sessions clicked the hero CTA
86–89%
of CTA clickers completed the existing play intro
8.3m
median reported completion time for current Tangles
65.7%
of unauthenticated gate exposures reached result data
14.6%
of starters were active on two or more days
01 · Current state

A long runway before the first clear payoff.

The player crosses several ceremonies, explanations, questions, and account asks before the product closes the loop. The exact path varies by Tangle format and auth state; this is the first-time narrative path.

1

Landing ceremony

The chest-and-key interaction runs before the promise and primary CTA become fully available.

Value delayed
2

Everyone goes to Daily

The previous daily-versus-onboarding route has ended. All hero traffic now enters through today’s Tangle.

Current control
3

Another introduction

New devices receive the full cinematic explanation and fingerprint interaction before play.

More ceremony
4

Identity questions

Name, pronouns, and dating preference may be required before the story can begin.

Personal ask
5

Full Tangle

Current completed Tangles report a median of 8.3 minutes, with substantial idle and resume effects.

Core product
6

Rating gate

The player must rate the Tangle before their personalized reading is requested and shown.

Ask before payoff
7

Login before reading

Unauthenticated narrative players are sent through account creation before seeing the reading.

Commitment gate
8

Unclear next chapter

The result, Player Pass, profile, and home surfaces do not tell one simple story about what to do next.

Momentum breaks
02 · What is not working

The problem is not one bad screen. It is the order of value, trust, and momentum.

Initial entry is healthier than repeat behavior. The highest-leverage fix is a coherent journey that turns one result into a reason to create an account—and a reason to return.

01

Too much happens before “I get it.”

The product explains itself through ceremonies and forms before the player has felt the basic loop: choose → be seen → want to know more.

02

The account ask precedes the promise.

Current gate conversion is respectable, but roughly one third of unauthenticated exposures do not reach result data. The ask is doing work before the result has earned trust.

03

Player Pass feels like another onboarding.

Only 51–60% of people who start profile/card onboarding complete it. Most loss happens before the story step is finished.

04

The first result has no runway.

Home, profile, and the next Tangle compete as destinations. A new player needs one clear continuation—not a generic product surface.

05

Repeat play is the real weakness.

Only 14.6% of starters were active on multiple days. Exact D1 return after a first completion was 12.7%; exact W1 return after a start was 5.6%.

06

Some “funnel” signals are not states.

results_viewed means data loaded, starts repeat on remount, and block_abandoned also fires for completed sessions. The new flow needs clean exposure and reveal events.

03 · Proposed flow

Let play create the profile. Let login preserve it.

The first Tangle replaces tutorial and Player Pass setup. It should be a real, two-minute Trove experience—not a demo—and its result becomes the first durable profile signal.

1
Immediate promise What this is, how long it takes, what the player gets.
2
2-minute FTUE Tangle Teach through three meaningful choices and visible consequences.
3
Core result Headline, traits, and one evidence-based paragraph before auth.
4
Login to save “Save this read and start your Trove.” Continuity, not permission.
5
Finish the result Return to the same place, confirm it is saved, then collect the card.
6
Profile 1/3 Show what is known, what is missing, and why two more angles matter.
7
Best next Tangle One curated recommendation, playable now, with a clear reason.
Value first

Do not require an account before the first play, and do not hide the entire result. The player must feel genuinely seen before Trove asks for commitment.

No context switch

OAuth returns to the exact result step. Finish the reading before moving anywhere else. Never dump a newly authenticated player onto generic home.

One next action

The profile becomes the bridge to the second Tangle. The primary CTA is one recommendation; home and “come back tomorrow” remain secondary options.

04 · The key screens

Four screens should feel like one uninterrupted thought.

These are directional content mockups, not final UI. They show the information hierarchy and the handoff between result, account, profile, and the next play.

A · Before loginEarn trust
trove
Your first read
You protect the room before you protect yourself.

When the moment got uncomfortable, you absorbed the tension first—then stepped in once someone else was carrying the cost.

protective measured socially alert
Save this read →
Illustrative result copy
B · Account handoffPreserve momentum
C · After the resultProfile is the destination

Your profile is taking shape.

One real signal, two more perspectives to make the pattern reliable.

1 of 3
What we can say so far
“You notice who is carrying the social cost—and adjust before most people do.”

Based on your choices in your first Tangle. This is a starting read, not a finished label.

PressureFirst signal found ✓
BelongingPlay a social Tangle
AgencyPlay a work or life Tangle
Recommended next

The group chat went quiet

We have seen how you handle pressure. This adds what you do when belonging is at stake.

3 minutesAdds: belongingPlay now →
1 / 3

A meaningful first layer

Show the complete first result, one supported pattern, two missing lenses, and one recommended next Tangle.

2 / 3

An emerging pattern

Show where the two results agree or differ, then recommend the specific final lens needed for Profile v1.

3 / 3

Profile v1 unlocked

Deliver a coherent initial profile. “v1” makes clear that the profile is useful now and continues evolving.

05 · Decisions this flow makes

What we should stop doing—and what replaces it.

Stop

Requiring login before someone has experienced Trove.
Treating Player Pass creation as a separate mandatory onboarding flow.
Sending newly authenticated people to generic home.
Calling a one-Tangle profile empty or “33% incomplete.”
Showing a catalogue when one recommendation would be clearer.

Start

Using the first Tangle as both onboarding and the first real profile input.
Showing a credible core result before asking for an account.
Returning from auth to the same result and confirming it was saved.
Letting the first result automatically create the profile shell.
Explaining exactly why the next Tangle adds a missing perspective.
06 · What to do next

Coordinate the journey before building individual screens.

The immediate deliverable is one shared flow specification that keeps product, design, content, identity, and analytics aligned.

Workstream 1

Lock the experience contract

  • What appears before login
  • What remains after login
  • Hard gate versus dismissible save ask
  • Immediate next play versus tomorrow
Workstream 2

Design one screen family

  • First result reveal
  • Login/save handoff
  • Saved confirmation
  • Profile 1/3 and 2/3
Workstream 3

Define content and progression

  • Two-minute FTUE Tangle
  • Credible first-result schema
  • Two complementary profile lenses
  • Curated next-Tangle pool
Workstream 4

Build continuity and measurement

  • OAuth return-to-result
  • Guest-to-account merge
  • Playthrough and exposure IDs
  • Reveal, profile, and next-play events

How to judge whether it works

Randomize login-before-result against core-result-before-login. Optimize for second-Tangle completion, Profile v1 unlock, and D1/D7 return—not signup alone. Add a true result-revealed event before running the test; the current results_viewed event only proves that data loaded.