Scarlett's growth bet: one person installs it, it proves useful to their teammates, and the workspace fills in by itself. "Activation" measures exactly that: the share of workspaces that grow beyond the installer within 14 days.

NowActivation 27% (12 of 44 workspaces). Largest deployment: 71 seats from one install. The average workspace reaches ~2.3 seats by week 8 after installing.

The loop is not yet firing reliably: most installs stay single-player. This is the number to fix before spending on acquisition. Caveat: with 44 eligible workspaces, each one moves this percentage a lot. Read direction, not decimals.

Three different "revenue" numbers appear on the dashboard on purpose. List MRR is what subscriptions would bill at face value. Paying MRR counts only customers actually charged money. Cash collected is what has genuinely arrived. Blending them is how early-stage dashboards mislead; separating them is the point.

Now$350 list MRR across 7 active subs · $200 paying · $200 set to renew · $800 collected all-time ($0 this month) · 3 comped partners · 7 canceled, 0 scheduled to cancel.

Comped design partners are the standard early trade: free usage in exchange for feedback, logos and case studies. The number that has to move next is paying MRR.

"Weekly active" counts distinct people using the product in a week. "Stickiness" divides weekly actives by monthly actives: it asks, of everyone who used it this month, how many are back this week? For workplace software, anything around half means routine, not curiosity.

Now38 people active last week · stickiness 100% (40 weekly / 40 monthly) · 1 people visited the tasks surface in 30 days.

100% stickiness says the people who adopt Scarlett keep using it week after week; growth problems are easier to fix than habit problems, and this is a habit signal. Usage depth is currently measured by a proxy (product events per workspace); it switches to real credit burn automatically once the product emits it.

The cohort table groups workspaces by the week they first appeared and asks whether they were active again 1, 2, 3 weeks later. Rows that stay dark from left to right are workspaces that stuck. It is the cleanest early answer to "do people come back?"

NowThe newest readable cohort (7 workspaces) kept 43% after one week. The internal org row holds 100% every week since May.

The pattern so far is normal early SaaS: a share of tire-kickers go quiet after week one, while workspaces that connect their tools stay. The one to watch is week-1 retention on the biggest recent cohorts; if that climbs as onboarding improves, the retention story writes itself. Sample sizes are small, so a single workspace can swing a cell by double digits.

Three gates stand between a stranger and revenue: visiting the site and trying the app (marketing), finishing setup (onboarding), and completing checkout (pricing). Each is measured separately so friction is visible where it actually happens.

Now85 site visitors → 0 app sign-ins in 30 days (0%) · onboarding: 67 start, 57% reach the final step · checkout: 0% of 0 started checkouts complete.

A 0% visitor-to-app rate is unremarkable for cold traffic and fine pre-launch; it becomes the lever the moment paid acquisition starts. The onboarding funnel is the highest-leverage fix available today: everyone who falls out there was already convinced enough to try. One caveat: site and app identities only partially stitch together, so the visitor-to-app number is directional, not exact.

Every number on the metrics tab is real, but a few stand in for events the product does not emit yet. Proxies are labeled on the dashboard and swap out automatically the moment the real events arrive; nothing is restated by hand.

NowLive from source: installs, sign-ins, per-workspace attribution, all revenue figures, sign-ups, credit burn. Proxied today: task activity (1 people on the tasks page).

The North Star is live: 498 tasks completed and counting. Watch tasks per workspace per week become the single number that summarizes the business.

• The multiplayer proof: a single install grew organically to 71 active seats. That is the hardest thing for a workplace tool to fake.
• Stickiness is 100% (40 weekly of 40 monthly users). For a work tool, anything near half means habit, not novelty.
• Revenue is real: 4 subscriptions actually pay ($200/mo) and $800 has been collected. Tiny, but it proves willingness to pay.

• 3 of 7 active subscriptions are comped design partners. Normal at this stage, but list MRR ($350) overstates cash reality ($200 paying) until they convert.
• Of 67 people who start onboarding, 57% reach the last step. Every point recovered here compounds through everything downstream.