Sprint April 7 to April 18

[MichaelKutsch-ph]

Sprint planning

List of incidents + any action items that haven't been scheduled yet

Team sprint planning

For your team sprint planning copy this template into a comment below for each team.

# Team ___

**Support hero:** ___

## Q2 objectives

1. Objective + status (red/yellow/green), action points if yellow/red

## Retro

<!-- Grab the high and low priority items from last time and add whether that item was completed or not -->
- 

### High priority

-

### Low priority / side quests

-

[MichaelKutsch-ph]

Team InfraInDaHouse

Support:

• Infra: @danielxnj @MichaelKutsch-ph
• CH TBD

Retro:

• 🔴 Move non-capture steps to Warpstream on all prods @frankh
  • migration failed again, need to rethink our strategy here
  • run smaller topics one by one and test it properly the next time
  • tiered storage -> AWS TAM and let's take it from there
• ✅ (reprio) toolbox cli for un-distruptable jumpboxes
• 🟡 Plan and execute first steps for prop-defs rds migration @MichaelKutsch-ph - planning in progress. AWS talk lined up to get support and best practices -> dev setup is there for testing,
• ✅ Finish rollout new subnets / nodepools on prod @bciaraldi - ip exhaustion averted, but we need to adjust those again to resolve CIDR conflicts
• ✅ argo deployment flow is stable
• 🟡 migrating all apps to stable argo flow @MichaelKutsch-ph @danielxnj
• 🟢 remove nginx-ingress (deprecated and zero day)
• 🟢 RDS param groups: performance tweaks @bciaraldi
• Cost spike in March: +120k

Infra:

CH:

• Retro

• 🟡 Moving US cluster to US account @Daesgar

  • delays due to unavailable instances in AWS
  • switch over done - cleanup WIP

• ⚪ Makeing clickhouse parts smaller @Daesgar

• Focus on deduplication @tkaemming

• 😢 Pawel moving off to team DW

OKR Infra

🏃Improve build times for posthog apps
💽 Splitting RDS
🔒 AWS Account and networking re-work

OKR CH

• Query Observability and Performance Improvements
• Make ClickHouse ops easy
• Use Altinity Antalya in production for events

High priority

• Finalize move to argo flow @MichaelKutsch-ph @danielxnj
  • Move ingestion services to argo/new chart with Eli @danielxnj
• Analyze cost spike in March @Daesgar @MichaelKutsch-ph
• Test S3 tiered storage on MSK @frankh
• Review and extend runbooks w/ team infra @Daesgar
• Hand off deduplication @tkaemming
• -> Prop defs to CH -> RFC @tkaemming @bciaraldi
  • AWS DMS for property def data migration or logical replication @bciaraldi
• Pair incidents with infra/ch @Daesgar @PostHog/team-infra
• Dump events to iceberg @Daesgar
• Get started with upgrade clickhouse version @tkaemming

Low priority / side quests

• investigation/RFC on moving persons off postgres @frankh
• Makeing clickhouse parts smaller @Daesgar
• investigate: what grinds query exec. to a halt sometimes?
• Fix missing events in livestream @orian

✅=finished 🟡=in progress 🔴=won't finish ⚪=not started

[lucasheriques]

Team Surveys

• Surveys support hero: @marandaneto
• SDKs support hero: @ioannisj then @lucasheriques
• 2 day PTO - @lucasheriques
• 1 day PTO - @ioannisj

🟢=finished 🟡=in progress 🔴=won't finish ⚪=not started

Retro

Retro - Team Notes

What could've been better

• Should have updated docs for the new getSurveyResponse much sooner when I first implemented it last week @lucasheriques
  • My plan changed a lot due to this, didn't have a chance to work on my main feature nor continue working on partial responses
• Still struggling a bit with keeping up with support hero duty - definitely an area to improve @ioannisj

What went well

• Grateful I got the time to rest up when I got sick @lucasheriques
• PR reviews have been great, quality comments and fast approvals @lucasheriques
• Caching issue finally solved and survey API requests finally back to normal @marandaneto
• Attending a conference in Paris @marandaneto
• Bad start of the sprint with some personal issues first week, all well now! @ioannisj
• Very constructive 1:1 with Manoel @ioannisj

Retro - Quarterly Goals (skippable since we talked about those in q2 planning)

• 🟡 Improve quality
  • 🟢 Reduce SLA to fix bugs from months to weeks
  • 🔴 Reduce paper cut issues by 50% (currently 17/30) so almost there - very close but prob won't make it due to other issues on slack/my todo list
• 🟡 Mobile Surveys
  • 🟢 RN
  • 🟢 iOS
  • 🟡 Android
  • 🟡 Flutter
• 🟡 Talk to customers
  • 🟢 5 user interviews
  • ⚪ Reference customers
  • ⚪ Case study
• 🟡 Finish up WIP features
  • 🟢 AI summaries
  • 🔴 Partial responses
  • ⚪ Adaptive sampling
  • ⚪ Survey actions

Retro - Tasks

High priority

• 🟡 iOS Surveys beta release + docs - early in the sprint @ioannisj
• 🟡 Start work on Flutter surveys (Running on iOS for now) @ioannisj
• 🟢 Android surveys - all display logic and start with the UI @marandaneto
• 🟡 [Not planned] Better query format for handling survey responses (getSurveyResponse)
  • Implemented the function as a hogQL function
  • Updating docs for using the new function and also replacing $survey_response with $survey_response_{questionId}
• 🟡 [Not planned] New endpoint for survey stats @lucasheriques
  • Necessary for fixing ticket for key customer
• 🔴 Partial responses @lucasheriques
• ⚪ Assist @lucasheriques with adaptive sampling @marandaneto
• ⚪ Adaptive sampling for surveys (deprioritized) @lucasheriques
• ⚪ 🐛 Don't calculate NPS score for non-NPS questions @lucasheriques
• ⚪ Survey feedback widget events can be more accurate replaced by the issue below @lucasheriques
• 🟢 Survey open-choice question not auto-selected on text input @lucasheriques (higher prio since it's causing low NPS scores)

Low priority / Side quest

• 🟢 👟 Surveys snapshot throttle - capture snapshots only on view layouts @ioannisj
• 🟡 🐛 SwiftUI masked items zindex on top of everything @ioannisj
• 🟢 🐛 SDK Compilation Failures on VisionOS @ioannisj
  • Already a community PR with the fix there but slow progress.
• ⚪ Surveys SDK part is getting very big. [Already around 80kb per survey partial responses PR] @lucasheriques (chore: send survey partial responses event posthog-js#1788). So I want to slowly refactor some of the code, because there's a lot of repetition and the logic is hard to follow, with a lot of callbacks. Want to take a look on the survey sending logic this time

Mobile replay and Mobile SDKs

• Mobile SDKs already use the new remote config API @marandaneto @ioannisj

Goals

Click to expand quarterly goals

• 🟡 Mobile surveys, Everywhere

  • Android
  • Flutter

• ⚪ Level up survey analysis

  • Global analytics view
  • A cleaner, more useful survey responses page
  • Be able to see surveys result breakdown by week
  • Better results analysis with metric based templates
  • Don't calculate NPS score for non-NPS questions
  • Extra polish and UX improvements across the board, pair up with the Design team to get some ideas

• ⚪ Survey customization

  • Better wait period controls
  • Alignment customization here and here
  • Inline surveys
  • Add icons
  • Templated strings
  • Single-response submit

• Templates, Templates, Templates

  • Save custom templates
  • Duplicate surveys across projects/environments
  • Add mobile-specific templates
  • Explore using AI to create survey templates

• Make Key Customers Super Happy

  • Be lightning-fast at resolving their issues
  • Build a case study with a power user

• Hire a Cracked Product Engineer

  • Hire and onboard our next teammate
  • Make onboarding smooth and fast so they can hit the ground running

High priority

• Survey partial responses @lucasheriques
  • Figure out a way to allow continue filling a survey in the same session if it closes due to refreshes/url changes
  • Update all other charts and data table to consider partial responses (consider only one response per response id)
  • Update surveys usage to only consider one event per response id.
• 🐛 Don't calculate NPS score for non-NPS questions @lucasheriques
• Android surveys - UI views and beta release/docs @marandaneto
• Flutter surveys - This depends on the output of the next couple of days but minimum goal is to integrate with iOS SDK and show a dummy view when a survey is activated @ioannisj

Low priority / Side quest

• Improvements for survey multiple question visualization @lucasheriques
• MVP for global analytics view - can leverage the new /surveys/stats endpoint for an initial version that shows that data @lucasheriques
• Option to reuse the anonymous user id between user changes @ioannisj
• Stop all integrations like session replay and autocapture on opt-out @ioannisj

[pauldambra]

Team Team

Support hero:

@pauldambra week 1
@veryayskiy week 2

@pauldambra off all of week 2 - and no laptop in Italy with the kids

Retro

• nobody got arrested or lost in Budapest
• i loved my dream of 10 hours in Genova airport
• persons as a separate thing instead of as a product analytics feature hit us again @pauldambra @veryayskiy
  • https://github.com/PostHog/product-internal/pull/725#discussion_r2023623886
• excited by "find the needle in the haystack" @pauldambra

## what we shipped

High priority

• 👩‍🔬 Goal 2: Churn and retention exploration
  TBC
• 🟡 🎤 Goal 3: Watch less and get more value making filters better
  TBC
• 💀 Goal 4: Squash distractions @pauldambra
  • 🟡 move E2E tests from Cypress to Playwright - @pauldambra
  • 🟡 testing our own rrweb with our own recordings

OKRs 🟡

1. OKR, status (red/yellow/green) and action points if yellow/red

• 🚀 Goal 1: Get Alex 2 rocketing 2

  • it's great to have Alex 2 on the team... let's make it a success

• 👩‍🔬 Goal 2: Find the needle in the haystack

  • Many of our customers have lots of recordings, but struggle to find the "useful" ones. We want to test and iterate on a bunch of different ideas to get customers to value faster.
  • Understand the needles that people are looking for using user interviews, data analysis and more
  • Better options for users to watch recordings from where they are (e.g. in funnels or experiments)
  • Some people are scared of filtering, some people have no patience for it - give them ways to avoid it
  • Make it super easy to see what filters have results to watch - "show me, don’t make me figure it out"

#30460

High priority

• Many of our customers have lots of recordings, but struggle to find the "useful" ones.

  • summaries for single sessions with click to visit the point in time @sortafreel

• Make it super easy to see what filters have results to watch - "show me, don’t make me figure it out"

  • timeline for recordings into the UI @pauldambra
    • where does it go - filters? what to watch? taxonomic filter?
    • how to test it - just "amount of filters set"

• Understand the needles that people are looking for

  • ✅ get the unsubscribe survey updated and react to new responses @veryayskiy

• watch recordings from where they are

  • a lot of people use persons as the entry point for recordings, want to improve that @veryayskiy
  • better "not found" state for recordings @veryayskiy
    • e.g. TTL passed, or recording might be sampled, etc etc

Side quests and lower priority

* add-on: research store recordings longer paid feature @annikaschmid 
* RN - session timeout - stop of recordings (we have it in web) @veryayskiy 

[daibhin]

Team Error Tracking

Retro

🟢 Much more progress on pricing and prepped most everything for launch
🟢 Switched to Rust based ingestion pipeline to reduce Kafka hops. Cost go 📉
🟢 Split up issue query for faster loading
🟠 Switched from PostHog focused features to building integrations for Linear & GitHub
🟢 Olly and Hugues did stellar on support!

OKR

🟢 Launch with pricing

Be the default choice for the languages we support

⚪️ Make it easier to integrate
🟢 Improve product quality
🟢 Build advanced features for larger customers

High priority

Launch @daibhin

• Merge PRs
• make sure the rollout goes smoothly

Improve product quality @hpouillot

• Add platform display and filter to issues
• Stack trace ordering
• Add 'kaboom' flag to frame in SDKs
• Ensure ordering is correct for all SDKs
• Order correctly in cymbal or frontent

Build advanced features for larger customers

@daibhin

• Linear & GitHub integrations
• Filter alerts based on assignee

@oliverb123

• Build HogVM
• Custom server side issue grouping

[mariusandra]

Team DevEx

OKR

• Build the "Product OS"
  • New layout and "file-based" project view
    • Nested nodes within nodes
  • Every product is able to export operations (backend) and widgets (components) => AI, notebooks etc.
    • Shortcuts and suggestions
  • Permissioning
  • Keep iterating on shared UI components
• Improve dev tooling in PostHog
  • Roll out "product folders" for all teams
  • Convert E2E tests to playwright
  • Tests take 5-10min max
  • Standardise dev tooling (vscode, flox, docker, self hosted)

High priority

Marius off April 10-18

• Marius
  • Definitely get the feature preview live
  • Sorting (recent files), pinning
• Adam
  • Product tree table view
  • Finish the new layout design
• Cory
  • "Products" to "Shortcuts"
  • Finesse

Low priority

• AI tree organizer with human in the loop (modal: current tree, user input, diff preview, save)

[meikelmosby]

Team CDP

Support hero: @meikelmosby
Onboarding Week for @benjackwhite & Toronto Environments work with Zack

OKR

1. OKR, status (red/yellow/green) and action points if yellow/red

😍 OMG so many destinations 🟢
💖 Data pipelines UI sparks joy 🟢
🤑 Pricing thats works for everyone 🟢

Retro

- Add legacy plugins comparison logic to validate it is all 1-1 so we can move off of @benjackwhite 🟡 - Migrate all existing Zapier hooks to hog functions @benjackwhite 🟡 - Release Hog based transformations @meikelmosby 🟡 - still missing hog watcher to get it released - Work with docs team to get transforms and destinations good for release 🟢 - Try to hook up Segment destinations to PostHog @MarconLP 🔴 - TikTok Ads destination @MarconLP 🟢

High priority

• Release Hog based transformations @meikelmosby
• Start using HogTemplate code instead of copying it into each hog function @meikelmosby
• Figure out what to build for community Templates @annaszell @meikelmosby
• moving ad destinations to node and convert them to use mappings @MarconLP
• figure out a way to test pixels in node @MarconLP
• Try to hook up Segment destinations to PostHog @MarconLP
• CDP pricing @annaszell
  • should be able to take a decision to do something about pricing this quarter or next quarter

Low priority / side quests

• rolling out site_destinations out publicly @MarconLP
• getting slack app verified @MarconLP

[pl]

Team Ingestion

• Support: @pl

Retro

High priority

• 🟡 Complete and roll out recent improvements to cyclotron and propdefs @eli-r-ph
• 🟢 Set up a mirror propdefs deployment (including the db) for experimentation @eli-r-ph
• 🟢 Investigate propdefs caching improvements @pl with support from @eli-r-ph
• 🟢 Implement long term storage for mr blobby v2 @pl
• 🟢 Review the data between blobbies v1 and v2 @pl
• 🟢 Decide on the approach for session recording deletion @pl

Low priority

• ⚪ If we get propdefs into a stable state, @eli-r-ph will return to implementing the propdefs API

Discussion

• Couldn't roll out the cyclotron changes, as we had to wait for CDP
• Waiting for Tim's proposal about using Clickhouse for propdefs
• We had fewer incidents / issues this iteration

OKRs

• 🟢 Fix property definitions service
• ⚪ Finish moving to capture-rs
• ⚪ Develop ownership of ingestion incidents
• 🟢 Improve infrastructure for deployments

High priority

• Onboarding Nick in Vancouver 7th-11th of April @pl @eli-r-ph
• Continue rollout of Blobby V2 @pl
• Roll out batch write changes to propdefs @eli-r-ph
• Land the caching changes @pl
• Isolate the database in the propdefs-v2 deployment @eli-r-ph

Low priority

• Investigate data migration options @eli-r-ph

[rafaeelaudibert]

Team Revenue Analytics

• Support: @rafaeelaudibert

Retro

First sprint planning, nothing to retro about

OKRs

• 🟠 Web-Analytics-like dashboard for metrics tracking
• 🟠 Expose streamlined revenue DW view to allow customers to improve on top of it
• ⚪ Streamline Stripe connection

High priority

• Initial views in the database
• Decide on how we can decide whether we use charges or invoices
• Initial UI for Revenue Analytics with basic MRR/ARR

Low priority

• Migrate the config to a separate table

[Twixes]

Team Max AI

Q2 objectives

WIP

Retro

✅ Offsite and charting a new course

High priority

• Proxy LLM calls @kappa90
• Port static analysis to new extension @kappa90
• Simplify code agent @kappa90
• New query planning architecture, incl. evals for SQL fallback @Twixes
• Run evals in CI @Twixes 

• Integrate PostHog UI (probs iframe)
@Twixes
• Codebase indexing MVP @skoob13
• Top AI requests in product analytics @HamedMP
  1. Add insight to dashboard from prompt
  2. Create dashboard from prompt
  3. Tweak current insight being viewed
  4. Add specific filter in taxonomic selector
• Smooth creation of personal API key @joshsny (guest)

[dmarticus]

Team Flying, Leaping, and Galloping (Flags)

Retro

Theme: tuning `/flags` performance and rolling out new API response to all SDKs.

What went well

• @haacked

  • Shipped the new flags API response to other things

• @dmarticus

  • fixed the Int -> BigInt issue for all the scary tables and enhanced the runbook about it
  • posthog.com is now powered by /flags, but the p99 for /flags is worse than /decide. Dug into it, found a perf bug, PR is drafted and will be shipped 04/07
  • built a ton of o11y around new flags to figure this out.

What could've gone better

• Didn't make progress rolling new flags out to more customers because of the perf issue we found with /flags – didn't see the projected 5x speed boost. Anxious to see how that goes this week.

This sprint

Themes: Roll out new flags everywhere, finish /decide?v=4 support for all clients, start on behavioral targeting, and anything else on the docket.

High Priority

• @dmarticus

  • Roll out new flags to a meaningful percentage of production traffic. Continue to tune performance.
  • Finish behavioral cohort targeting RFC

• @haacked

  • Rolling v4 to the all the client SDKs (React Native, PHP, maybe Mobile)
  • We might need to resolve the send_feature_flags issue for all SDKs, too

Low Priority

• Long tail of client SDK v4 responses

Support

• 04/07 - 04/10 @haacked (short week due to vacay)
• 04/10 - 04/18 @dmarticus

Time Off

@haacked from 04/10 to 04/14

Q2 2025 Objectives

(Details here)

• 🟢 Flag Stability, Consistency, and Resiliency
• ⚪ Behavioral Cohort Targeting
• ⚪ AI IDE Integration

[aspicer]

Team Product Analytics (3/31 - 4/14)

• Anirudh

  • Support
    • Custom names for formula series (lot of users/sales asking for it)
    • (stretch) insight edit mode
  • retention docs
  • slack notifications for alerts (integrate with cdp destinations)

• Thomas

  • Support follow up: Exporting “all columns” from the Activity page is broken for some customers as the query OOMs. Turns out this is mainly because the events are ordered by timestamp, but the table is ordered by toDate(timestamp) . There is a workaround by first sorting an inner query by toDate(timestamp) and then sorting the outer query by timestamp . Want to implement this in the events query runner or as a general HogQL modifier.
  • Support follow up: Add missing taxonomic property definitions to the backend and/or create them from a single source.
  • Finish up dashboard colors feature: Need to change the schema of the customized breakdowns so that it can also take into account the breakdown type to support cohort breakdowns. feat(dashboards): allow customizing breakdown colors #30291
  • Finish up the initial version of the paths v2 feature feat(paths-v2): add a separate paths v2 query #29364

• Sandy

  • Optional Funnels Steps
    • Scope
    • Design
    • Coding
  • Support

[robbie-c]

Team Web Analytics

Support hero: @lricoy

@robbie-c on vacation 21st to 25th

OKRs

• 🟡 Finish cookieless mode
• Build marketing analytics
  • 🟢 Add revenue data to web analytics
  • 🟡 Get ad spend data into data warehouse and use it for per-channel analytics
• Build per-page reports
  • 🟢 Core web vitals

Retro

• @lricoy add feature preview for page reports 🟢
• @lricoy iterate on feedback from feature preview 🟢
• @lricoy exploratory query performance work (in particular, explore aggregated data) 🟡
• @robbie-c start defining some of the web analytics <> AI stuff (e.g. max controlling web analytics filters) 🔴 (deprioritised)
• @robbie-c backfill the events.session_id_uuid column 🟡 (it's running in the background)

Low prio:

• @robbie-c look at google ads data warehouse connector 🔴

Next sprint

High priority

• @lricoy page report iteration based on post-release feedback (released 7th April)
• @robbie-c onboarding!

Low priority

• @lricoy query perf

[andehen]

Team Experiments

Support hero: @andehen 1st week, @rodrigoi 2nd week

@andehen on vacation 14th to 21th

Retro

Q2 OKR's set (link).

Anders:

• Onboarding Rodrigo
• Support
  • Less than it used to be 👍
• New experimentation engine work
  • Refactoring to improve abstractions and code ergonomics
  • Multi-step funnel support
  • Bug fixes

Rodrigo:

• Onboarding and onboarding tasks

Next sprint

• Release the new experimentation engine as a "Feature preview"
• Have 10 distinct organizations try out the new experimentation engine
• @andehen Support Winsorization
• @rodrigoi Support traffic estimation in running time calculator
• Many minor bugfixes and UI tweaks
• Refactoring to make UI changes easier
• Get Rodrigo up-to-speed on support

[hoppybunny]

Hello @pauldambra ! A big fan of posthog here and I'm wondering if there's anywhere I can check the status of this particular feature? It's a big pain point I have - I've got huge volumes of user recordings and it's unrealistic to watch everyone of them even if they're at 5x speed, and would love to find a way to identify recordings that are worth watching. Happy to jump in to test anything out or provide feedback too :) Thank you!

High priority

• Many of our customers have lots of recordings, but struggle to find the "useful" ones.
  • summaries for single sessions with click to visit the point in time @sortafreel

[sortafreel]

Hey @hoppybunny 👋 Would be awesome, pinged you on LI to schedule something.