NorthlightScientific Growth Dashboard
Text to get started

For teams running growth experiments

Turn every growth idea into a real test ranked by RICE.

Text your agent a growth idea. It lands on your board, and the agent asks the questions that make it an experiment: the hypothesis, the metric, when to kill it. Reach, impact, confidence and effort become one RICE score, so you know what to run first.

Text this app to get started

Or

Sign in with Northlight

Ideas found in your meetings and Slack wait in an inbox until you accept or dismiss them.

Acme Robotics, referral credit ideaExample run

Where it is

0 of 5 steps

  1. Saved your text as an idea in Backlog

    Waiting

  2. Checked it for a hypothesis, a metric and kill criteria

    Waiting

  3. Asked you the first question, by text

    Waiting

  4. Filled in reach, impact, confidence and effort from your answers

    Waiting

  5. Scored it with RICE

    Waiting

A demo, not your data.

Every idea, ready to test

What you text becomes an experiment on a board, with what is still missing spelled out.

A RICE score

Reach × Impact × Confidence ÷ Effort, worked out once all four are known, so the board sorts by what to run first.

Referral credit for annual plans: RICE 1,200.

The next question

The coach finds what keeps an idea from being a test, such as no metric, no baseline or no kill criteria, and asks for one thing at a time.

What single metric decides whether this worked, and where will you read it?

A board by stage

Ideas move through Parked, Backlog, Up Next, In Progress and Shipped.

Referral credit for annual plans: Backlog, Outbound.

What it keeps doing

You keep texting ideas. The app keeps the board honest.

Finds ideas in your meetings and Slack

When the workspace owner turns scans on, a daily scan reads recent meetings and Slack for growth ideas and puts them in your inbox.

Test a referral credit for annual plans.From a team call on Oct 9

Checks what others already tried

An idea can carry a precedent: who tried it, what happened, and a verdict of supports, mixed or against, with links to the pages.

Mixed: a competitor kept referral credits for annual plans only.From 2 pages the agent read

Logs the result and the call

When a test ends you record the number against the baseline, the outcome and the decision to ship, iterate or kill.

Baseline 2.1%, target 3%, actual 2.4%. Iterate.From your experiment log, Nov 4

You decide what becomes an experiment

Ideas the scan or an agent finds wait in your inbox. Accept puts one in Backlog. Dismiss drops it for good, and later scans will not propose it again.

Test a referral credit for annual plans

Team call, Oct 9: What if annual customers got a credit for every referral?

Text this app to get started

Send one message, then text your first growth idea.

Text this app to get started

Or

Sign in with Northlight