Daily Insights

Thursday 2 October 2026

Covering Wed 1 Oct 2026, 00:00–23:59 SAST ("yesterday") against Tue 30 Sep ("prior") and the 7 days 25 Sep–1 Oct. All times SAST (UTC+2).
Generated Thu 2 Oct 2026, 07:55 SAST
Sources: PostHog project 77139 (bloomboard.io + app.bloomboard.io), production DB (read-only)
Staging and localhost excluded; internal test accounts removed from signups

Quiet Wednesday after Tuesday's peak: 1 real signup, Instagram traffic halved

Tuesday 30 Sep was the best day of the month (11 signups, 86 marketing visitors). Wednesday 1 Oct fell back: 1 external signup (plus 2 internal test accounts), Instagram referrals dropped from 50 to 26 visitors, and the register page saw 6 visitors against 24. A 19-hit crawler burst at 19:15–19:18 SAST inflated the raw marketing count; real visitors were about 58, not 79.

Marketing visitors
58raw 79
▼ 33% vs 86 prior day
bloomboard.io unique visitors; raw minus 19 crawler hits and 2 flagged bots
Marketing pageviews
97
▼ 46% vs 178 prior day
1.3 pages per visit (prior 2.1). 7-day total 663
Production signups
1+2 internal
▼ 10 vs 11 prior day
DB-verified external users. 7-day total 21 (prev 7 days: 6)
Contact forms
0
▼ 1 vs 1 prior day
2 in 7 days: Tue 30 Sep 16:47 and Mon 29 Sep 22:34 SAST, both US
App visitors (logged-in dashboard)
43
▼ 23% vs 56 prior day
425 dashboard pageviews (prior 659). Public client pages: 166 visitors
First event created
1
▼ 3 vs 4 prior day
11 in 7 days (prev 7 days: 5)
First proposal sent
0
▼ 1 vs 1 prior day
1 in 7 days (Tue 30 Sep 01:17 SAST, signup from 28 Sep)
Trial converted
0
= 0 prior day
3 in 7 days (25, 27, 28 Sep). 2 trials expired yesterday

Traffic, 14 days

bloomboard.io unique visitors per day

18 Sep–1 Oct 2026, SAST days. Light segment on 1 Oct = 19 crawler hits + 2 flagged bots.
43 86 0 Thu 18 Sep: 12 visitors Fri 19 Sep: 4 visitors Sat 20 Sep: 5 visitors Sun 21 Sep: 18 visitors Mon 22 Sep: 21 visitors Tue 23 Sep: 23 visitors Wed 24 Sep: 15 visitors Thu 25 Sep: 34 visitors Fri 26 Sep: 22 visitors Sat 27 Sep: 17 visitors Sun 28 Sep: 30 visitors Mon 29 Sep: 59 visitors Tue 30 Sep: 86 visitors Wed 1 Oct: 58 real visitors Wed 1 Oct: 19 crawler hits + 2 flagged bots (raw total 79) 86 59 79 18192021 22232425 26272829 301 Oct Sep
VisitorsCrawler / bot

Production signups per day

External users created, SAST days. Light segment on 1 Oct = 2 internal test accounts. 28 Sep onward from the DB; earlier days from PostHog.
5 10 0 Sun 21 Sep: 1 signup Mon 22 Sep: 2 signups Tue 23 Sep: 2 signups Wed 24 Sep: 1 signup Sun 28 Sep: 2 signups Mon 29 Sep: 7 signups Tue 30 Sep: 11 signups Wed 1 Oct: 1 external signup Wed 1 Oct: 2 internal test accounts 11 7 1 +2 18192021 22232425 26272829 301 Oct Sep
External signupsInternal / test

Yesterday vs prior day

Marketing visitors by hour (SAST)

Unique visitors per hour. The 19:00 spike on 1 Oct is the crawler burst (19:15–19:18).
10 20 0 Prior day (Tue 30 Sep) visitors per hour Yesterday (Wed 1 Oct) visitors per hour 19:00–19:59 SAST: 20 visitors, 19 of them one crawler Tue 30 Sep 04:00–04:59 SAST: 12 visitors (US evening) crawler, 19 00040812162023 Hour of day, SAST
Wed 1 OctTue 30 Sep

Referring domains (marketing)

Unique visitors to bloomboard.io by external referrer.
ReferrerWed 1 OctTue 30 Sep24–29 Sep
Direct / none40 19 crawler1966
l.instagram.com265046
www.google.com121536
chatgpt.com010
www.instagram.com010
block.guard.io001
Device split: Wed 36 mobile / 40 desktop (21 real) / 1 tablet. Tue 68 mobile / 18 desktop. Instagram traffic is almost all mobile.

Countries (marketing)

Unique visitors to bloomboard.io by GeoIP country.
CountryWed 1 OctTue 30 Sep24–29 Sep
United States39 19 crawler5799
Australia14713
United Kingdom1044
South Africa6721
New Zealand310
Canada142
Thailand120
Other (1 each)547
Wed others: Philippines, Singapore, Sweden, Mexico, Cambodia. Tue others: Finland, Indonesia, Netherlands, Mongolia. 24–29 Sep others: France 3, Georgia, Mexico, Italy, Switzerland.

App · app.bloomboard.io

Logged-in dashboard visitors per day

Unique visitors on dashboard routes, 24 Sep–1 Oct (excludes public client pages and login/register).
28 56 0 Wed 24 Sep: 25 Thu 25 Sep: 25 Fri 26 Sep: 13 Sat 27 Sep: 12 Sun 28 Sep: 28 Mon 29 Sep: 41 Tue 30 Sep: 56 Wed 1 Oct: 43 56 43 24 Sep252627 2829301 Oct

App areas, yesterday vs prior

Unique visitors and pageviews by area. "Public" = client-facing proposal, enquiry, brief and payment pages.
AreaWed visitorsWed pageviewsTue visitorsTue pageviews
Dashboard (logged in)4342556659
Public client pages166225142169
Login / register31534963
All app207703207891
Register page visitors: 6 on Wed, 24 on Tue, 15 on Mon. Client-side activity Wed: 3 enquiries submitted, 3 briefs submitted (DB), 0 payments started. Weekly app totals: 1,211 visitors / 4,729 pageviews (prev 7 days 665 / 3,843).

Activation funnel · 7 days, 25 Sep–1 Oct

283
Marketing visitors
deduplicated, bot-filtered; prev 7 days 73
53
Register page visitors
app /register uniques, daily sum
21
Signups (external)
DB users; PostHog saw 20 of them
11
First event created
9 of the 20 tracked signups (45%)
1
First proposal sent
Tue 30 Sep 01:17; prev 7 days 3
3
Trial converted
25 Sep core AU · 27 Sep studio US (since cancelled) · 28 Sep studio US
Other activation events, 7 days: first arrangement added 4, first proposal previewed 4, wizard completed 14, checkout started 11 events (5 people), trial expired 5, trial link requested 11 events, phone sign-in link requested 14 events. Yesterday: wizard completed 4, first proposal previewed 2, checkout started 1, trial expired 2, trial link requested 2 (Brisbane 01:33, Sydney 19:00 SAST).

Signups in detail · 7 days

SAST timeLocationArrived fromDeviceSeen on bloomboard.io firstCreated first eventSent proposal
Wed 1 Oct 02:40Irvine, USbloomboard.ioDesktopyesyes–
Wed 1 Oct 06:15Cape Town, ZAdirectDesktopinternal test account (DB)
Wed 1 Oct 09:25Johannesburg, ZAdirectDesktopinternal test account (DB)
Tue 30 Sep 00:48Los Angeles, USbloomboard.ioMobileyes––
Tue 30 Sep 00:57Delaware, USbloomboard.ioMobileyes––
Tue 30 Sep 01:02Raleigh, USbloomboard.ioDesktopyesyes–
Tue 30 Sep 03:08Brandon, USdirect (trial link via Instagram 03:06)Desktopnoyes–
Tue 30 Sep 03:52La Pine, USbloomboard.ioMobileyes––
Tue 30 Sep 04:12Redondo Beach, USbloomboard.ioDesktopyes––
Tue 30 Sep 04:53Chapel Hill, USbloomboard.ioMobileyes––
Tue 30 Sep 06:39US (city unknown)bloomboard.ioMobileyes––
Tue 30 Sep 06:58Ulan Bator, MNbloomboard.ioMobileyes––
Tue 30 Sep 19:21Savannah, USbloomboard.ioDesktopyesyes–
Tue 30 Sep 19:42Fullerton, USdirectDesktopno––
Mon 29 Sep 01:20Santa Rosa, USbloomboard.ioDesktopyesyes–
Mon 29 Sep 04:38Lake Stevens, USbloomboard.ioDesktopyesyes–
Mon 29 Sep 20:58Los Angeles, USbloomboard.ioMobileyes––
Mon 29 Sep 21:21New Braunfels, USbloomboard.ioMobileyes––
Mon 29 Sep 22:08Chattanooga, USbloomboard.ioMobileyes––
Mon 29 Sep 22:47Alameda, USbloomboard.ioMobileyes––
Sun 28 Sep 00:57Pennsauken, USbloomboard.ioDesktopyesyesyes, 30 Sep 01:17
Sun 28 Sep 23:55Los Angeles, USbloomboard.ioDesktopyesyes–
PostHog recorded 20 of the 21 external signups in the DB (Mon 29 Sep: DB 7, PostHog 6). All 10 mobile signups on Mon and Tue requested a phone sign-in link within one second of signing up, which looks automatic rather than a user action. Today so far (to 07:55 SAST): 2 new users in the DB, 1 tracked signup (Auckland, NZ, 07:17, direct, created first event).

Identity stitch · marketing → app

Can a signup be traced to a marketing visit?

Same anonymous PostHog id seen on bloomboard.io before signing up on app.bloomboard.io (the two hosts share the cookie; identify() is never called by design).
WindowExternal signups trackedStitchedGap
Wed 1 Oct110
Tue 30 Sep1192
7 days, 25 Sep–1 Oct2018 (90%)2
The 2 gaps are both direct arrivals on the register page (Brandon and Fullerton, US). One of them requested a trial link from the Instagram landing page 2 minutes earlier, so the marketing visit happened on another device or in the Instagram in-app browser.

Marketing visitors later seen in the app

Share of each day's bot-filtered bloomboard.io visitors whose id also appears on app.bloomboard.io between 25 Sep and 2 Oct.
DayMarketing visitorsAlso in appShare
Wed 1 Oct771114% 19 crawler in base
Tue 30 Sep862226%
Mon 29 Sep591831%
Sun 28 Sep29828%
Sat 27 Sep17424%
Fri 26 Sep22314%
Thu 25 Sep31413%

Quality signals · flags

Rage clicks in the production app

Rapid repeated clicks in one spot. 78 rage clicks on Wed 1 Oct (prior day 70).
PageWed 1 OctTue 30 SepNote
One event page (/events/d07d950a…)51153 sessions, US desktop Chrome: Tue 01:50–02:44 (15), Wed 01:54–02:19 (27), Wed 03:21–03:27 (24). Target has no text (image or canvas). flag for D2
Other event pages19 6 pages46 8 pagesMostly 2–6 each; Tue had 17 on one page and 9 on another
Template studio (2 proposals)403 untargeted + 1 on "Shown"
Public enquiry / brief / week-of pages35"back", "Continue" buttons
Events list, media library, register14Tue: "Copy link too", "Welcome", "Start free trial"
Replays for the event-page cluster: 01a0f4bc-bc35-77fa-8167-0bdaa23ac9bc, 01a0f4d8-8b89-72df-8a4d-5746d95293ee, 01a0ef67-5de1-7911-8777-0248d152b95e at eu.posthog.com/project/77139/replay/<id>.

Experiments and feature flags

PostHog state as of 07:55 SAST.
ItemState
PostHog feature flags0 defined
PostHog experiments0 defined
Flag values on events ($active_feature_flags)none in the last 2 days
App feature flagsServer-side in Django (feature_flag_definitions), not visible in PostHog

Billing changes outside the trial funnel

Subscription rows created for accounts from March 2026: Tue 30 Sep 11:48 SAST (studio, active) and Wed 1 Oct 16:18 SAST (core, active). These are plan changes by existing customers, not trial conversions, so no PostHog event fires.

Marketing detail

Top pages, bloomboard.io

Unique visitors per page. Wed's deep pages are almost all the crawler (1 visit each).
PathWed 1 OctTue 30 Sep
/ (home)5781
/features/proposals210
/florist-software29
/features/recipes26
/contact15
/switching03
/faq23
/florist-proposal-software (+ templates)59
/vs/curate, /vs/details-flowers31
/changelog21
Bounce (single-page sessions): Wed 68 of 73 sessions (93%; 49 of 54 after removing the crawler), Tue 52 of 84 (62%).

Campaign and link parameters

Unique visitors by UTM and affiliate parameter on bloomboard.io pageviews. Attribution reference only.
ParameterWed 1 OctTue 30 Sep
utm_source=ig, utm_medium=social514
utm_source=chatgpt.com02
Affiliate parameter (fpr=) present2847
No parameters4740
Most Instagram arrivals carry the affiliate parameter rather than UTMs, which is why utm_source=ig undercounts Instagram against the referrer table.

Data notes

  • Windows. Days are SAST calendar days. "Yesterday" = Wed 1 Oct 00:00–23:59 SAST; "prior" = Tue 30 Sep; "7 days" = 25 Sep–1 Oct; "prev 7 days" = 18–24 Sep.
  • Crawler. 19 pageviews between 19:15:01 and 19:18:32 SAST on Wed 1 Oct came from Chrome 106 on Linux at 1920×1080 (US, New Jersey / Virginia / Nebraska data-centre IPs), each a new anonymous id hitting one distinct page, no referrer. PostHog did not flag them as bots; the adjusted figures remove them. This pattern matched 0–1 visitors per day on the previous 6 days.
  • Test accounts. PostHog has no user identity (by decision), so internal accounts are removed by matching the day's signups against the production users table (bloomboard.io, demo and test emails). Staging (app.staging.bloomboard.io, 3 people) and localhost are excluded by host.
  • Marketing uniques count anonymous device ids; the same florist on two devices counts twice. Weekly uniques are deduplicated across the 7 days (283), so they are lower than the sum of daily uniques (327).
  • Signup source of truth is the DB users table (21 external in 7 days). PostHog's signup_completed saw 20 of them; the PostHog cohort percentages use 20 as the base.
  • Trial conversions are PostHog trial_converted events, cross-checked against billing_subscription rows created on the same day.