SolutionsGrowth
Growth. Run your funnel as an enforced process.
A gate before every send and spend, and a journal that shows why a number moved. Campaign, funnel, and cohort work run as enforced processes across your martech stack.
Runs as a service today.
01Capabilities
The funnel, run as one process.
Growth teams run many tools and many channels. Growth runs them as one enforced process, with a gate before anything that sends or spends.
- Campaign orchestration
- Campaigns run across channels as one enforced process. UTM discipline is part of that process: a link that breaks the tagging convention fails its gate before it ships, so attribution never depends on someone remembering the taxonomy.
- Funnel-stage automation
- Each funnel stage is a step with a gate, and consent state is a gate input. A contact without the consent a channel requires does not advance to a send, whatever a timer or a list says.
- Cohort and retention analysis
- Cohorts are built and re-built as runs, with the attribution window recorded alongside the cohort definition. A retention number keeps its meaning because the window it was measured in travels with it.
- Experiment execution
- Experiments run with evidenced outcomes: the variant, the audience, and the measurement window are journaled with the result, fixed when the hypothesis is recorded rather than after the numbers arrive.
- Martech integration
- CRM, email and lifecycle, analytics, and ad platforms connect through Adapters. Provider limits, such as an ESP's send rate, are written into the process as policy the run obeys, not lore an operator carries.
- A funnel memory that holds
- Why a number moved is a signed, append-only record: which campaign, which cohort, which window. Ask as of any date and the answer is what was believed then.
02Walkthrough
One experiment, journaled end to end.
This is what an experiment looks like as an enforced process: hypothesis first, a gate before the send, and a journal that replays the whole decision later.
hypothesis.recorded
A shorter onboarding email lifts activation for the new-signup cohort. Recorded before anything runs.
variant.drafted
One variant, one control, drafted inside the process. Both journaled.
breakpoint.holding
The send is a consequential action. The run stops and states what it wants to send, and to whom.
breakpoint.approved
A growth lead reviews the audience and the copy, then approves.
send.executed
Through the connected email system, exactly as approved.
outcome.measured
Activation measured against the cohort as defined. The definition is stored with the result.
replay.completed
Months later, the whole decision replays from the journal: hypothesis, approval, send, result.
When someone asks later why the number moved, the journal already holds the answer. An unapproved send costs trust and an unapproved spend costs budget; the gate is cheaper than either.
03Calibration
Calibrated to your growth evals.
The harness (agent, skills, tools, process, memory) is tuned against evals defined on your own outcomes. No invented numbers; the target is whatever your instrumentation measures.
- Evals on growth outcomes
- Evals are defined on the outcomes your team already tracks: activation, conversion, retention lift, as your instrumentation measures them.
- Process and skills tuned
- The process definition and the channel skills are adjusted and run against the evals until runs hit the target score.
- Held, then re-checked
- Once calibrated, the harness holds. The evals re-run, so a drift in a channel or an audience surfaces early, as a regression in the scores.
04The console
What the console shows.
- A campaign board across every connected channel.
- Funnel and cohort views with stage conversion.
- Experiment runs, each stopping at an authorization gate before spend.
- Martech integration status for every connected system.
You can run the board, its gates, and the journal yourself; they are open source. Calibrating them to your funnel, and operating the loop with your team, is the engagement.
When a number moves, Growth can show what moved it, and who approved the move.
Bring one experiment; watch it clear its gates and land in the record.
A Growth demo walks one experiment through its gates: hypothesis, approval, send, replay.
Or emailhello@a5c.ai
The intro call is 30 minutes and free.