Le-Ko — Weekly product health snapshot

Focus: what improved vs what needs attention (current week vs prior week)

Current period: 03 Aug – 09 Aug 2026 · Comparison: 27 Jul – 02 Aug 2026
Report generated: 10 August 2026 · Source: PostHog live queries · Production (106150) · https://eu.posthog.com

Active Users Product - Active Users

761
Active users (period)
-11.3% vs prior week
57.8%
Opened app only 1 day
+3.4% vs prior week
2.06
Avg active days per user
-1.4% vs prior week

Fewer people are opening the app

Active users fell from 858 to 761 (-11.3%). New users: 506 → 308 (-39.1%). Returning users: 352 → 453 (+28.7%). Lowest day this period: 2026-08-08 with 183 daily actives (prior low: 194 on 2026-07-27).

SegmentThis weekPrior weekChange
New users (first active in range)308506-39.1%
Returning users453352+28.7%

Onboarding v2 Product - Onboarding v2

Matches dashboard Funnel - Onboarding v2 complete flow: SplashAnimationStarted with empty event user_id (test accounts excluded), login = LoginPhoneSubmitted, ordered path. Includes people who never become signups — those are Stage 1 drop-offs. OTP verified = LoginOtpSucceeded. Privacy step in the live funnel is Privacy accepted (Continue) (separate privacy screen step may be absent).

289
Finished (Let's go)
-43.4% vs prior week
20
App installs
+42.9% vs prior week
54.2%
Re-open after onboarding
-14.9% vs prior week
9.9h
Avg hours to next app open
-64.6% vs prior week

Fewer truly new users finish Onboarding v2

Finished (Let's go) 511 → 289 (-43.4%) — same count as the final funnel step below. Splash starts: 710 → 455. Installs: 14 → 20.

Install tracking looks broken vs onboarding volume

Only 20 app installs vs 289 finished onboardings (455 splash starts). Install events are not a reliable acquisition proxy this week — treat Finished / splash as the real volume signal. Funnel still includes returning re-logins.

Finishing signup and coming back are still two different challenges

54.2% of completers opened the app again (was 63.7%). Weakest day: 2026-08-09 — only 36.8% of completers that day came back (19 completers). Average return after 9.9h (was 28.0h) — faster return is good even when coverage fell.

Onboarding v2 funnel — this week vs prior week

Same definition as dashboard Funnel - Onboarding v2 complete flow: ordered aggregate_funnel, empty splash user_id, login = LoginPhoneSubmitted, OTP = LoginOtpSucceeded, Privacy = Continue accept. Early droppers who never reach signup are included (no signup_completed cohort gate). A few already-onboarded returners may still appear. Each bar = users who reached that step and every prior step in sequence.

Stage 1 — Splash → OTP verified

Stage 1 (splash → OTP): 455 → 338 (74.3% conversion, 117 dropped). Biggest leak: Splash finished → Login screen (61 users, 14.2%). Prior week conversion was 76.3% (biggest leak then: Splash finished → Login screen, 97 / 14.1%).

TransitionFromToDroppedDrop %
Splash started → Splash finished455430255.5%
Splash finished → Login screen4303696114.2%
Login screen → SMS sent36936271.9%
SMS sent → OTP verified362338246.6%

Stage 2 — OTP verified → Finished

Stage 2 (OTP → Finished): 338 → 289 (85.5% conversion, 49 dropped). Biggest leak: OTP verified → Privacy accepted (Continue) (33 users, 9.8%). OTP → Privacy this week: 338 → 305 (33 dropped). Prior week Stage 2 conversion was 94.3%; OTP → Privacy was 542 → 521 (−21).
Slowest median step this week: Sports → Survey (0.33 min).

TransitionFromToDroppedDrop %
OTP verified → Privacy accepted (Continue)338305339.8%
Privacy accepted (Continue) → Name screen30530500.0%
Name screen → Birthdate screen305292134.3%
Birthdate screen → Gender screen29229200.0%
Gender screen → Sports screen29229200.0%
Sports screen → Survey screen29229110.3%
Survey screen → Notifications screen29128920.7%
Notifications screen → Welcome screen28928900.0%
Welcome screen → Finished (Let's go)28928900.0%

Median minutes between steps (this week vs prior)

One metric only — median time. High “upper” outliers are intentionally omitted.

Step transitionMedian this week (min)Median prior week (min)Change
Splash → Login0.180.25-28.0%
Login → OTP0.220.20+10.0%
OTP → Privacy0.020.020.0%
Privacy → Name0.100.100.0%
Name → Birthdate0.180.22-18.2%
Birthdate → Gender0.220.220.0%
Gender → Sports0.030.030.0%
Sports → Survey0.330.32+3.1%
Survey → Notifications0.200.200.0%
Notifications → Welcome0.070.070.0%
Welcome → Finished0.030.030.0%

App installs by day of period

Re-open coverage by day of period

Hours until next app open by day of period

Funnel checkpointThis weekPrior weekChange
01 — Splash started455710-35.9%
03 — Login screen369593-37.8%
05 — OTP verified338542-37.6%
Privacy accepted (Continue)305521-41.5%
Finished (Let's go)289511-43.4%

Onboarding analysis — where users drop

Stage 1 — SMS sent → OTP verified (−24, 6.6%): Of 362 users who got an SMS, only 338 accepted the code. The rest reach the OTP screen and mostly leave (typical next event: app backgrounded) without LoginOtpSucceeded. Secondary Stage 1 leak: Login → SMS (−7, 1.9%). Biggest Stage 1 leak overall: Splash finished → Login (61 users).

Stage 2 — after OTP: OTP → Privacy accepted is 33 drop (338 → 305). Continue → Name: 0 (305 → 305). After Name, Stage 2 is tight through Finished (289). Prior week OTP → Privacy was 542 → 521 (−21).

Bottom line: funnel volume (splash 710 → 455; finished 511 → 289). Biggest Stage 1 leak is Splash finished → Login; OTP accept and SMS send failures still matter. Post-OTP conversion is 85.5% this week. Watch re-open coverage separately (63.7% → 54.2%).

Feed Product - Feed

Core engagement surface. KPIs match the Production Feed dashboard. Feed visitors = Feed pageview: custom PageFeed or $screen with $screen_name = 'PageFeed' (tab open / refresh / return). All-time onboarded = distinct people with person.properties.user_id who finished Onboarding v2 (onboarding_v2_welcome_continue) or the old Onboarding last screen (PageOnboardingLocation). Avg feed triggers / user ≈ scrolls/refreshes (each pageview loads a post page). Posts published = feed_post_create_post.

710
Feed visitors
-18.4% vs prior weekheld up vs fewer onboarders
270
Engaged users
-21.7% vs prior weekheld up vs fewer onboarders
38.0%
Engagement rate
-4.1% vs prior week
326
Posts published
-9.2% vs prior weekheld up vs fewer onboarders

Period onboarded context

Finished onboardings this week: 289 vs prior 511 (-43.4%). Engagement metrics below are checked against this growth — if users rose a lot but engagement did not, that is called out.

Proportional check — Feed visitors held up vs fewer onboarders

Onboarded fell -43.4% (511 → 289), but Feed visitors was -18.4% (870 → 710). Engagement declined less than the drop in new finishes — relatively resilient.

Proportional check — Feed engaged users held up vs fewer onboarders

Onboarded fell -43.4% (511 → 289), but Feed engaged users was -21.7% (345 → 270). Engagement declined less than the drop in new finishes — relatively resilient.

Proportional check — Posts published held up vs fewer onboarders

Onboarded fell -43.4% (511 → 289), but Posts published was -9.2% (359 → 326). Engagement declined less than the drop in new finishes — relatively resilient.

2,616
All-time onboarded (w/ user_id)
stock · v2 + old
27.1%
Feed visitors ÷ onboarded
-18.4% vs prior week
4.98
Avg feed triggers / user
+9.2% vs prior week
3,538
Feed pageview events
-10.9% vs prior week

How active is the onboarded base on Feed?

All-time onboarded (v2 Finished or old Location, with user_id): 2,616. This week 710 distinct people hit Feed (27.1% of that stock; prior week 870 / 33.3%). Avg Feed pageview triggers per visitor: 4.98 (was 4.56) from 3,538 pageview events — a proxy for how many Feed loads / “scrolls” each visitor did.

Engagement fell — driven mainly by likes and comments

Visitors 870 → 710 (-18.4%); engaged users 345 → 270 (-21.7%); engagement rate 39.7% → 38.0% (-4.1%).

Likes: 6,369 → 4,956 events (209 → 164 users). Comments: 435 → 323 events (131 → 113 users). Post opens: 650 → 469 (144 → 123 users). Shares 19→35; reposts 17→15; follow-from-feed 1→1.

Posts published fell

Publish confirms (feed_post_create_post): 359 → 326 (-9.2%) from 150 → 125 unique creators. Negative actions (hide/report/unfollow): 6 → 1.

Composition (share of published posts with each field_filled attachment — a post can count in more than one): photo 266 → 218 (74.1% → 66.9%); video 97 → 110 (27.0% → 33.7%); location 29 → 24 (8.1% → 7.4%); no media 28 → 29 (7.8% → 8.9%).

Feed visitors by day of period

Posts published by day

Count of feed_post_create_post — user confirmed/published a post on Feed.

Post composition — photo / video / location by day

Daily published posts vs how many included each field_filled attachment (overlaps allowed).

What drove engagement — events and unique users

ActionEvents this wkEvents priorΔ events Users this wkUsers priorΔ users
Like4,9566,369-22.2%164209-21.5%
Open post469650-27.8%123144-14.6%
Create post326359-9.2%125150-16.7%
Comment323435-25.7%113131-13.7%
Share3519+84.2%1713+30.8%
Repost1517-11.8%97+28.6%
Follow user110.0%110.0%

Groups Product - Groups

373
Community tab visitors
-19.6% vs prior week
135
Joins
-57.9% vs prior week
8
Groups created
-50.0% vs prior week
1
Leaves
-83.3% vs prior week

Groups membership fell

Joins 321 → 135; leaves 6 → 1; net 315 → 134. Tab visitors 464 → 373; engaged inside groups 155 → 102; creates 16 → 8.

Proportional check — Groups tab visitors held up vs fewer onboarders

Onboarded fell -43.4% (511 → 289), but Groups tab visitors was -19.6% (464 → 373). Engagement declined less than the drop in new finishes — relatively resilient.

Proportional check — Group joins fell harder than onboarders

Onboarded -43.4%; Group joins -57.9%. Drop is steeper than the onboarded decline — not explained by fewer new users alone.

Community tab visitors by day

Joins vs leaves by day

Profile actions Product - Profile actions

Own-profile clicks vs visitor actions on someone else’s profile — same triggers as the Production dashboard.

301
Own-profile viewers
-18.0% vs prior week
136
Visitor-action users
-12.3% vs prior week
122
Follows (from profile)
-21.8% vs prior week
21
Messages
+90.9% vs prior week

Profile social actions fell

Own-profile viewers 367 → 301; visitor-action users 155 → 136. Follows from profile 156 → 122 (visitor Follow taps in mix: 156 → 122). Messages 11 → 21.

Most clicked on own profile this week: Web card. Most clicked as a visitor: Gallery.

Proportional check — Profile visitor-action users held up vs fewer onboarders

Onboarded fell -43.4% (511 → 289), but Profile visitor-action users was -12.3% (155 → 136). Engagement declined less than the drop in new finishes — relatively resilient.

Proportional check — Profile follows held up vs fewer onboarders

Onboarded fell -43.4% (511 → 289), but Profile follows was -21.8% (156 → 122). Engagement declined less than the drop in new finishes — relatively resilient.

Own-profile clicks (top actions)

ActionThis weekPriorChange
Web card111185-40.0%
Repost7792-16.3%
Add awards4568-33.8%
QR code3751-27.5%
Feed3544-20.5%
Share3361-45.9%
Click service1511+36.4%
Add first service12120.0%
Become coach880.0%
Add work experience56-16.7%

Visitor actions on others’ profiles

ActionThis weekPriorChange
Gallery183179+2.2%
Follow122156-21.8%
Repost3452-34.6%
Feed2330-23.3%
Message2111+90.9%
Service1012-16.7%
Work experience7n/a
Awards4n/a
Certifications3n/a
Schedule32+50.0%

Workouts Product - Workouts

5
Workouts created
-37.5% vs prior weektracks onboarders (-38% / -43%)
8
Create give-ups
-38.5% vs prior week
38.5%
Create success rate
+1.0% vs prior week
0
Joins
n/a
2.0
Avg minutes to create
+82.4% vs prior week
1.6
Median minutes to create
+80.2% vs prior week
5
Paired start→confirm
for timing

Workout creation fell

Created 8 → 5; give-ups 13 → 8; success rate 38.1% → 38.5%; joins 0 → 0.

Time to create (start workout_group_create → confirm, same person within 24h): avg 1.1 → 2.0 min; median 0.9 → 1.6 min (n=5 paired this week).

Proportional check — Workouts created moved with onboarders

Onboarded -43.4%; Workouts created -37.5%. Roughly tracks the change in new finishes.

Created vs give-up by day

Service creation Product - Service creation

4
Services created
-76.5% vs prior weekfell harder than onboarders (-76% vs -43%)
0
Give-ups
n/a
100.0%
Success rate
flat vs prior week
16
Create starts
-36.0% vs prior week
8.2
Avg minutes to create
+178.2% vs prior week
2.0
Median minutes to create
-22.4% vs prior week
4
Paired start→confirm
for timing

Service creation fell

Created 17 → 4; give-ups 0 → 0; success 100.0% → 100.0%; starts 25 → 16.

Time to create (entry tap → service_creation_create_confirm, same person within 24h): avg 2.9 → 8.2 min; median 2.5 → 2.0 min (n=4 paired this week).

Proportional check — Services created fell harder than onboarders

Onboarded -43.4%; Services created -76.5%. Drop is steeper than the onboarded decline — not explained by fewer new users alone.

Created vs give-up by day

Engagement depth Level 1 vs Level 2

Level 1 = main entrance tabs we put in front of users: Feed, Explorer, Groups tab. Level 2 = deeper actions beyond those surfaces: visit profile, explorer swipe/click, join/create groups, start/create service or workout, etc. Question: are people only consuming the main screens, or going deeper into the product?

710
L1 users (main tabs)
-18.4% vs prior weekheld up vs fewer onboarders
426
L2 users (deeper)
-26.0% vs prior weekheld up vs fewer onboarders
60.0%
L2 of L1 users
-9.4% vs prior week
284
L1 only (no L2)
-3.4% vs prior week

How deep did users go?

710 people hit a main tab (L1); of those, 426 (60.0%) also did a deeper L2 action. 284 stayed on main surfaces only (prior week: 294 L1-only; L2-of-L1 66.2%).

Prior week L1 870 → L2 576. Onboarded this week 289 (-43.4%).

Proportional check — Level-2 (deeper) users held up vs fewer onboarders

Onboarded fell -43.4% (511 → 289), but Level-2 (deeper) users was -26.0% (576 → 426). Engagement declined less than the drop in new finishes — relatively resilient.

Retention Product - Retention

In plain English: we put people into a group by the week (or day) they first opened the app. Then we ask: what % of that group came back later? Higher % = better. A 0% on a very new group usually means “too early to measure,” not “everyone left.”

Needs attention

The last finished weekly group (w2, 292 people who first opened the app that week) came back at 6.2% one week later. That matches the headline number for week ending 2026-08-03: 6.2%. Vs the week before (w3 at 29.5%), that is down (-23.3 points). The newest group (w1, 13 people) still shows 0% — that is normal: not enough time has passed for a “came back after 1 week” number.

What’s good

  • Best finished week in the table: w11 at 39.5% (43 people).
  • Several recent bigger groups are holding up: w3, w4, w5.

What’s bad / what to watch

  • Week-1 return for w2 is only 6.2% (below the ~25% bar).
  • Weakest bigger group: w2 at 6.2% (292 people).

Daily chart — how to read it

Each date = people who first opened that day. Green (day 1) = % who came back the next day. Red (day 7) = % who came back a week later. Recent days can look low on day 7 until a full week has passed — prefer finished days like 2026-07-30 (d1 27.5%, d7 12.8%). Across this chart, day 1 ran about 17.2%–36.8% and day 7 about 0.0%–26.3%.

Weekly table — how to read it

w1 (top row) = newest week of first opens. Rows go older as the number goes up (w2, w3, …). Columns w1 / w2 / w4… = “came back after 1 / 2 / 4 weeks.” 0% on the right usually means that week has not happened yet for that group. Focus on the first finished row with a real week-1 number (right now: w2 at 6.2%).

CohortUsersAfter 1 wkAfter 2 wkAfter 4 wkAfter 8 wkAfter 12 wk
w1130.0%0.0%0.0%0.0%0.0%
w22926.2%0.0%0.0%0.0%0.0%
w351629.5%4.8%0.0%0.0%0.0%
w423129.9%21.6%0.0%0.0%0.0%
w531830.5%21.4%3.1%0.0%0.0%
w642631.2%23.2%17.8%0.0%0.0%
w711032.7%19.1%10.0%0.0%0.0%
w829426.2%18.7%8.5%0.0%0.0%

Priority list

Critical — investigate first

Medium — dig deeper

Cross-dashboard threads