ScarlettInvestor Metrics
The AI coworker that lives in Slack, connects to your tools, and turns requests into finished work.
Live · updated Sep 4, 8:27 AM UTC
WorkspacesiDistinct Slack workspaces where Scarlett is installed — external design partners plus our own internal / QA orgs.
60
56 external design partners + internal
People reachediDistinct individuals who have touched Scarlett — anyone who opened the app or appears inside an installed Slack workspace.
331
distinct users across app + installs
Largest deploymentiThe most people active inside one workspace from a single install. Shows Scarlett scales to an entire org, not just one user.
70 seats
org-wide internal adoption — proves it scales to a full team
MRRiList-price monthly recurring revenue of active Stripe subscriptions. Cash collected is actual money in — the two differ while design partners are on comped plans.
$500
$400 cash collected · 4 paying subs
The N+1 engine · single install → whole teamiScarlett's viral loop: one admin installs it, it DMs every teammate with specific help, and each teammate who activates becomes a new seat — so one install expands into the whole team.
One Slack install becomes a multiplayer workspace
1
Admin installs Scarlett
one seat, one OAuth grant
2
Scarlett auto-audits channels
builds a per-teammate read
3
DMs every teammate individually
2–3 specific, named ways to help
4
Teammates activate → new seats
the workspace fills in around the loop
Proven internally: our own org runs entirely on one install — 70 seats, fully multiplayer. The same loop is what expands the 56 external design-partner installs next.
Seats per workspace — every bar is one installiEach bar is one workspace; the length is how many distinct people are active in it. Tall bars = a single install that pulled in the whole team.
Cracked.
70
Test Scarlett
11
HealthRx
9
Muse
6
Fg
6
Personal Scarlett
5
Slavk
4
Taxes BY Erica
3
Modgroup
3
Fine glaze
3
customer internal / dogfood
Cumulative installsiRunning total of Slack workspace installs over time — the top-of-funnel growth curve.
207 installs to date
Sign-ups · weeklyiReal sign_up events from the product, fired server-side at account creation.
20 sign-ups in the last 30 days
Weekly activeiDistinct people active per calendar week in the product. The most recent (lighter) bar is the current, still-partial week.
Active users per week
Revenue · self-serve StripeiSelf-serve credit subscriptions billed monthly through Stripe ($1 = 500 credits). List MRR is subscription face value; cash figures are money actually collected.
$500 MRR
Live from Scarlett's Stripe account.
$400
cash collected · $0 this moiAll-time card revenue actually collected (net of refunds). The honest number next to list MRR.
$200
paying MRR · of $500 listiMRR from subscriptions that have actually paid a non-zero invoice. The gap to list MRR is comped design-partner plans.
$200
set to renewiPaying MRR minus subscriptions already flagged to cancel at period end — the recurring base as of today.
$50
ARPU (paying)iAverage monthly revenue per paying subscription.
10
active subs · 0 new this moiActive subscriptions in Stripe; 'new' counts subs started since the 1st of the current month.
4 + 2
churned + set to canceliSubscriptions fully canceled, plus active ones whose auto-renew is already switched off (they end at period close).
Live plan mixiCurrent active subscriptions grouped by monthly price. Upgrades shift this mix upward — watch for movement into the higher tiers.
$50/mo10
Price book ($1 = 500 credits)
$50/mo · 25k$100/mo · 50k$150/mo · 75k$250/mo · 125k$400/mo · 200k$500/mo · 250k
Billing activityiLive feed from Stripe: new subscriptions, payments, credit top-ups, plan upgrades / downgrades, and cancellations, newest first. No customer details shown.
Latest revenue events
- Payment collected — $50Aug 31
- Subscription canceled — $50/moAug 17
- New subscription — $50/mo (comped)Aug 14
- Auto-renew turned off — $50/mo ends at period closeAug 14
- Payment collected — $50Aug 13
- Subscription canceled — $50/moAug 12
- Payment collected — $50Aug 12
- Subscription canceled — $50/moAug 11
- New subscription — $50/mo (comped)Aug 10
- Auto-renew turned off — $50/mo ends at period closeAug 10
Upgrades, downgrades and credit top-ups appear here automatically as they happen — none yet since launch.
MRR movementiList-price MRR added by new subscriptions vs lost to cancellations, by month. Comped design-partner plans count at face value — the paying split lives in the revenue card.
New vs churned, monthly
Conversion funnels · last 30 days
Visitor → paying workspace
Site → app sign-iniDistinct marketing-site visitors vs distinct app sign-ins in the same 30 days. Identity stitching across site and app is partial, so treat as directional.0%
8,403 tryscarlett.ai visitors4 app sign-ins
Paywall checkout completioniStripe Checkout sessions opened vs completed in the last 30 days. Abandoned sessions are the gap — pricing-page friction shows up here first.75%
4 checkouts started3 completed
Onboarding funnel · all-timeiSign-up flow completion, derived from onboarding page views (channels → connections → autopilot → billing). Percentages are vs step 1. The real sign_up event replaces this once it ships.
Who gets through setup
1. Channels66
2. Connections18(27%)
3. Autopilot19(29%)
4. Billing36(55%)
Derived from app pageviews — some users skip steps or onboard from Slack directly, so treat drops as directional.
Workspace retention cohortsiEach row is the set of workspaces first seen that week. Cells show the % of that cohort active again N weeks later (any event counts). Dark left-to-right rows = workspaces that stick.
Do workspaces come back?
| Install week | Ws | W0 | W1 | W2 | W3 | W4 | W5 | W6 | W7 | W8 |
|---|---|---|---|---|---|---|---|---|---|---|
| Jul 5 | 10 | 100% | 40% | 20% | 20% | 10% | 10% | 0% | 10% | 0% |
| Jul 12 | 6 | 100% | 33% | 0% | 0% | 0% | 0% | 0% | 0% | |
| Jul 26 | 6 | 100% | 33% | 0% | 0% | 0% | 0% | |||
| Aug 2 | 8 | 100% | 38% | 13% | 0% | 0% | ||||
| Aug 9 | 5 | 100% | 20% | 40% | 20% | |||||
| Aug 16 | 1 | 100% | 0% | 0% | ||||||
| Aug 23 | 1 | 100% | 0% | |||||||
| Aug 30 | 2 | 100% |
Seat expansioniAverage cumulative seats per workspace by weeks since its install — the N+1 loop measured. A rising curve means installs keep pulling in new teammates after week 0.
Avg seats by workspace age
Averaged across workspaces old enough for each week; needs 3+ workspaces per point to plot.
Engagement & activation
207
InstallsiTotal times Scarlett has been added to a Slack workspace (includes re-installs).
176
Active seatsiDistinct people who have actually used Scarlett inside a workspace, summed across all workspaces.
30%
activation · 12/40 wsiWorkspaces that reached 2+ active seats within 14 days of install — the N+1 loop actually firing. Workspaces younger than 14 days aren't counted until they mature.
21%
stickiness · WAU/MAU 19/89iWeekly actives as a share of monthly actives. Higher = usage is habitual, not occasional.
19
Active last weekiDistinct people who used Scarlett during the most recent weekly window.
20
sign-ups · 30diReal sign_up events in the last 30 days, fired at account creation.
37
on /tasks · 30diDistinct users who viewed the tasks surface in the app in the last 30 days — the closest live proxy for task activity until the task_completed event ships.
454
Tasks completediThe North Star: finished pieces of work Scarlett delivered. Needs the product to emit task_completed — /tasks traffic proxies it until then.
Tasks completed is the North Star (weekly tasks per workspace). The event ships in the next instrumentation PR and will populate here automatically.
Credit burniAverage credits consumed per active workspace per week, from real credit_used events. Credits are the unit customers pay for ($1 = 500 credits).
Credits per active workspace, weekly
Live credit consumption — the direct precursor to revenue expansion and top-ups.
Workspaces
Who's running on Scarlett
| Workspace | Type | SeatsiDistinct people active in this workspace. | EventsiTotal Scarlett events recorded for this workspace — a proxy for usage depth. | Last active |
|---|---|---|---|---|
| Cracked. | internal | 70 | 6,082 | Sep 4 |
| Test Scarlett | internal | 11 | 419 | Aug 15 |
| HealthRx | internal | 9 | 14 | Jul 17 |
| Muse | internal | 6 | 400 | Aug 23 |
| Fg | customer | 6 | 171 | Jul 17 |
| Personal Scarlett | customer | 5 | 141 | Jul 11 |
| Slavk | customer | 4 | 50 | Jul 11 |
| Taxes BY Erica | customer | 3 | 185 | Aug 30 |
| Modgroup | customer | 3 | 39 | Aug 10 |
| Fine glaze | customer | 3 | 108 | Aug 13 |
| Kikkkx | customer | 2 | 82 | Jul 11 |
| BET ENTERTAINMENT | customer | 2 | 11 | Aug 19 |
Instrumentation & data integrityiWhich events are flowing live versus still being wired up, plus the exact data scope behind every number on this page.
workspace_installed— Slack install callback
$groupidentify (org)— per-workspace attribution
sign_in— web app auth
stripe: subs · upgrades · top-ups— read-only key
sign_up— account creation (server)
task_completed— North Star — live from task workers
credit_used— per-workspace burn
connection_added— 24 connector links
onboarding_completed— setup finished (deduped)
GA4 tag (G-X7SFPFL41Q)— acquisition funnel
All metrics are scoped to the Scarlett app — app.tryscarlett.ai page activity plus Slack workspace events (installs, sign-ins, per-org attribution). Sources: PostHog (live), Stripe (live, read-only key), GA4 property 539983276 (tag instrumenting). Anything marked pending is already instrumented in the product and fills in automatically the moment its first event arrives; proxies are swapped out for the real events at the same time.
Scarlett · confidential — for investor review only