E
Eprenz OS
Presentation-ready KPI plan

Eprenz OS KPI and A/B Testing Plan

This version is designed for presentation. It explains what we are testing, why we are testing it, how users are split, what is measured, and how the result turns into a launch decision.

What this page is for

Not a fake results dashboard. This is the testing plan before beta data exists. After beta starts, the same structure becomes the actual KPI dashboard.

Primary beta test

The first test is a controlled 2x2 landing page experiment.

We split 800 invited users into four groups. This lets the team test video and message separately instead of mixing all results together.

Group
What user sees
Message being tested
Audience split
Success metric
Variant AVideo plus ecosystem message.
Landing page with video at top.
Apply to join the Eprenz ecosystem.
200 users.
Application starts, completions, OS activations.
Variant BNo video plus ecosystem message.
Landing page without video.
Apply to join the Eprenz ecosystem.
200 users.
Application starts, completions, OS activations.
Variant CVideo plus OS message.
Landing page with video at top.
Get access to the OS for small business owners.
200 users.
Application starts, completions, OS activations.
Variant DNo video plus OS message.
Landing page without video.
Get access to the OS for small business owners.
200 users.
Application starts, completions, OS activations.
Independent variables

Each variable gets judged separately.

This prevents the team from saying “the page worked” without knowing whether the result came from the video, the message, the sender, the follow-up, the audience, or the UI.

Test area
Variable
Options
Primary measure
Why it matters
Compass landing pageThe page before the application.
Video vs no video.
VideoNo video
CTA click rate and application start rate.
Shows whether video helps or slows conversion.
Compass languageWhat we call the process.
Application vs Survey.
Compass ApplicationSurvey
Application completion rate.
Shows whether application framing increases perceived value.
HeadlineMain page promise.
Problem framing.
Know what to do nextAccess OSJoin ecosystem
CTA click and completion.
Shows which promise founders understand fastest.
Invitation messageBefore they arrive.
Message personalization.
GenericCustomLocation-basedStage-based
Invite click-through rate.
Shows whether personal context improves response.
Follow-upAfter no response.
Timing and owner.
No follow-up24h72hKevinDeepuDavid
Start lift after follow-up.
Shows whether conversion came from the page or human follow-up.
UI/UXForm and page experience.
Friction level.
Short formFull formSingle pageMulti-stepProgress bar
Completion rate and time to complete.
Shows whether users quit because the experience feels too long or unclear.
OS activationAfter Compass.
Onboarding path.
Dashboard firstProfile firstRecommended pathCommunity prompt
OS activation and first action.
Shows whether Compass leads into actual OS usage.
Measurement journey

Track the full user path.

Every test should show how many people moved from one step to the next.

01
Invite sentTrack source, sender, message, personalization, location, stage, and follow-up plan.
02
Page viewedTrack landing variant, video, headline, CTA, page length, and device.
03
Compass startedTrack application start, time to start, and first drop-off point.
04
Compass completedTrack completion rate, time to complete, stage classification, and answer quality.
05
OS activatedTrack profile completion, dashboard entry, content use, community use, vendor use, and AI interest.
Presentation readout

How to explain the results later.

When beta results come in, this is the format that makes the data presentable.

Question 1

Did the landing page convert?

Look atCTA clicks, Compass starts, and completion rate.
DecisionKeep the page only if starts turn into completed applications.
Question 2

Did video help?

Look atVideo play rate, video completion, and start rate vs no-video page.
DecisionUse video only if it improves qualified starts or completions.
Question 3

Did application language work?

Look atApplication vs Survey completion and answer quality.
DecisionUse the wording that creates more serious completed applications.
Question 4

Did follow-up distort the result?

Look atFollow-up owner, follow-up timing, and lift after follow-up.
DecisionSeparate page performance from human follow-up performance.
Question 5

Did users activate in the OS?

Look atProfile completion, dashboard entry, course, community, event, and vendor usage.
DecisionDo not scale if people complete Compass but do not use the OS.
Question 6

Did the system hold up?

Look atFailed submissions, tracking gaps, security issues, support tickets.
DecisionFix system issues before broader launch.
Decision logic

The winner is not the highest click rate.

The winning version is the one that creates qualified completions, OS activation, useful engagement, and clean data. A page with fewer clicks may still win if it brings better users.

ScaleUse when one variant clearly wins on completion, activation, and user quality.
RetestUse when results are close, sample quality is uneven, or follow-up changed the outcome.
FixUse when users click but do not complete, complete but do not activate, or activate but do not engage.