Each tile shows Friday 3 Oct against Thursday 2 Oct. The bars are the last 7 days, Saturday to Friday, hover for the value.
Sat 27 Sep to Fri 3 Oct against Sat 20 Sep to Fri 26 Sep. Counts are unique people unless stated. Production only.
| Measure | This week | Prior week | Change |
|---|---|---|---|
| Marketing unique visitors | 307 | 101 | ×3.0 |
| Marketing pageviews | 682 | 296 | ×2.3 |
| Marketing sessions | 344 | 135 | ×2.5 |
| Trial links requested | 11 | 0 | new |
| Contact forms submitted | 2 | 0 | +2 |
| Signups completed | 31 | 6 | ×5.2 |
| Wizard completed | 22 | 4 | ×5.5 |
| First event created | 18 | 5 | ×3.6 |
| First arrangement added | 10 | 4 | ×2.5 |
| First proposal previewed | 9 | 3 | ×3.0 |
| First proposal sent | 1 | 3 | −2 |
| Checkout started | 4 | 3 | +1 |
| Trial converted | 2 | 2 | 0 |
| Trial expired | 5 | 1 | +4 |
| Florists active in the dashboard | 200 | 111 | ×1.8 |
| Dashboard pageviews | 3,850 | 3,117 | ×1.2 |
| Public client page visitors | 929 | 776 | ×1.2 |
| Client enquiries submitted | 9 | 5 | +4 |
The 31 people who signed up between Sat 27 Sep and Fri 3 Oct, and how far each milestone reached by Saturday morning. Trial conversions on 27 and 28 Sep were older accounts.
Unique visitors to bloomboard.io by referring domain and campaign tag. Instagram includes the link-in-bio traffic (utm_source=ig, utm_medium=social).
| Source | Fri 3 Oct | Thu 2 Oct | 7 days |
|---|---|---|---|
| Google search | 19 | 15 | 72 |
| Direct / untracked | 17 | 8 | 103 |
| Instagram (l.instagram.com) | 4 | 10 | 130 |
| bloomboard.io (internal links) | 4 | 4 | 33 |
| ChatGPT | 0 | 1 | 2 |
| Other | 2 | 1 | 3 |
| Tag | Fri 3 Oct | 7 days |
|---|---|---|
| ig / social (link in bio) | 4 | 53 |
| chatgpt.com | 0 | 4 |
| No tag | 36 | 264 |
| Country | Fri 3 Oct | Thu 2 Oct | 7 days |
|---|---|---|---|
| United States | 30 | 16 | 195 |
| Australia | 0 | 2 | 31 |
| United Kingdom | 4 | 9 | 26 |
| South Africa | 3 | 4 | 21 |
| Canada | 2 | 1 | 9 |
| New Zealand | 0 | 0 | 4 |
| Thailand, France | 0 | 0 | 3 each |
| 14 other countries | 1 | 2 | 1 to 2 each |
| Device | Fri 3 Oct | 7 days |
|---|---|---|
| Mobile | 18 | 195 |
| Desktop | 22 | 111 |
| Tablet | 0 | 1 |
Instagram carried the Tuesday and Wednesday peak (59 and 86 unique visitors on 29 and 30 Sep) and has faded since: 10 visitors on Thursday, 4 on Friday. Friday's traffic was Google search and direct, and more desktop than the week's mobile-heavy mix.
| Page | Views Fri | Visitors Fri | Views 7d | Visitors 7d |
|---|---|---|---|---|
| / (home) | 54 | 39 | 447 | 276 |
| /features/proposals | 6 | 6 | 47 | 37 |
| /faq | 4 | 3 | 26 | 16 |
| /features/recipes | 3 | 3 | 23 | 22 |
| /florist-proposal-software | 0 | 0 | 20 | 6 |
| /florist-software | 1 | 1 | 19 | 19 |
| /changelog | 2 | 2 | 14 | 11 |
| /contact | 1 | 1 | 13 | 11 |
| /switching | 0 | 0 | 11 | 8 |
A signup is "stitched" when the same PostHog person also has a bloomboard.io pageview in the last 45 days. Since 2 Oct the dashboard identifies signed-in users by their user id and merges the anonymous history.
Daily stitched / signups: Sun 28 Sep 2/2 · Mon 29 Sep 6/6 · Tue 30 Sep 9/11 · Wed 1 Oct 1/3 · Thu 2 Oct 2/3 · Fri 3 Oct 5/6. The 6 unstitched people most likely signed up on a different device from the one that browsed the site, or arrived directly at app.bloomboard.io/register (66 unique visitors to /register this week).
What happened. One florist's public enquiry link received 587 unique visitors in 7 days, about 160 a day, almost all South African, on mobile, arriving from m.facebook.com or with no referrer (the Facebook app strips it). The link is masked as /e/:token in analytics since Thursday, so the per-link count stops at 2 Oct, but the daily volume is unchanged (162 unique public-page visitors on Friday).
Why it matters. The enquiry form funnel for that traffic is 394 opened → 0 started → 0 submitted from m.facebook.com, and 226 opened → 2 started → 0 submitted from the no-referrer South African mobile group. Over the same 7 days, South African visitors arriving from Instagram went 12 opened → 10 started → 2 submitted, and visitors arriving from florists' own websites went 16 opened → 11 started → 4 submitted. Only about 5 of 150 daily enquiry-page sessions record any click at all. 3 rage clicks were recorded on enquiry pages this week.
Two readings. Either the Facebook post draws people who only want to look (a competition or a share with no intent), or the page does not work in the Facebook in-app browser on Android (Chrome WebView, which is what these visitors use). The zero-start count across 394 people is extreme enough that the second reading needs to be ruled out by hand: open an enquiry link from a Facebook post on an Android phone and check the form renders and the start event fires.
Feature flags and experiments. The PostHog project has 0 feature flags and 0 experiments defined (active or archived). Nothing is being A/B tested in PostHog. Product gating lives in the app, not in PostHog.
Test accounts filtered. No person in the project carries the internal-or-test marker in the last 7 days. Staging (app.staging.bloomboard.io, 99 pageviews, 10 people) and one Vercel preview host (4 pageviews) are excluded from every number above. Custom events are filtered to app_environment = production. The production database shows 0 of the 31 signups this week use an internal @bloomboard.io address.
Method. All counts come from PostHog SQL over the events table (project BloomBoard, EU), with timestamps converted to Africa/Johannesburg and bucketed by calendar day. Unique visitors are unique PostHog persons. "Florists active in the dashboard" excludes the public paths /e (enquiry), /p (sent proposal) and /i (invoice); "public client pages" is those three paths. Marketing is $host bloomboard.io or www.bloomboard.io.
Caveat. Headless browsers are dropped by the PostHog bot filter, so these are human-ish visitors, but the Facebook figures above have not been checked against a bot classifier.