Every report says it’s working. The customers say otherwise.

The problem

Why it happens.

Campaigns are live, sends go out, dashboards are green. Yet new users stall before they activate, trials don’t convert, and customers go quiet without anyone noticing.

The reports measure the moments: sent, delivered, opened. They don’t show which transition is breaking, for whom, or why.

What Perennus does

The work.

1. Map the lifecycle as states and transitions

   Signup, activation, conversion, retention, recovery and expansion, defined from the data your workspace actually holds.

2. Read the account, not the dashboard

   Triggers and entry and exit conditions, segments, deliverability and authentication, and who owns which relationship across your tools.

3. Classify every finding by its evidence

   Each finding is marked observed, derived, inferred or unknown, so you can see what is known and what is still a hypothesis.

4. Quantify only when the evidence supports it

   Where the data supports a figure, we state it and show the assumptions. Where it doesn’t, we say what’s missing instead of estimating.

5. Prioritise the fixes

   Ranked by leverage and by the strength of the evidence, with the next step for each.

Who it’s for

A good fit when…

• SaaS and subscription businesses with a lifecycle already running that isn’t performing
• Teams deciding what to build next, who want the decision made on evidence
• Leaders who suspect their reporting is flattering a system that isn’t working

The engagement

What it looks like.

1. 1Call
2. 2Read access
3. 3Investigation
4. 4Readout

A fixed-scope engagement. It starts with a call, then read access to your workspace, then the investigation and a readout.

Timing depends on the complexity of the workspace and how much evidence is available in it, and is agreed before we start.

Deliverables

What you get.

• A lifecycle map of your states and transitions
• A findings report, with every finding classified by its evidence
• An infrastructure check: deliverability, authentication, list and data ownership
• A prioritised plan of interventions
• A readout session with your team

Evidence

Where we’ve done this.

• 60.2%AI companion subscription appReading the trigger logic found one entry condition that had queued around 174,000 sends against a 66,000-contact list. It appeared in no campaign report.Read the case study
• 78,708RealworldThe audit came before the build: 85 product events available, 13 used by the lifecycle, and two systems addressing one audience with no ownership rule.Read the case study
• 4 → 0Notion · ResearchThe method in public: four signals collected at signup, none used in the lifecycle.Read the study

What happens after

Afterwards.

• Take the plan to your own team: the Diagnostic stands on its own.
• Or move into a Lifecycle Build and have the Studio fix what it found.