Data Instrumentation

I have a property like “app_version” and want to delete all event where “app_version” is less than “2.0.0”. I read about data filters, but I need to delete existing data to prevent duplications.

Hello,I encountered an issue in adding additional data source from S3.the error I get:   for context I have another source from the same bucket that is currently working. When I looked into the source of the problem I took a look at amplitude’s docs:amplitude docs - S3 import guideIn the policy that is there it seems that the S3 integration need the following: s3:ListObjects but s3:ListObjects is deprecated, the error from the policy editor in AWS: I would like to know if this is the cause of the problem? could it be on amplitude’s end?the currently running source is working fine. Thanks for any help in advance!  

Our team is trying to understand how our events interact with the Identify API over time.  We understand that the identify API intercepts events from the given image in the doc:What we’d like to understand is the sync user properties step works, specifically how the Identify API handles user properties for events that we backfill.Imagine the following scenario for a given user (all times in universal standard time): 1. at 4PM 01/01/2024 : We send Event A for User 1: `{ time: 4PM 01/01/2024  }`2. at 5PM 01/01/2024 : We set the Identity for User 1 `{ foo: “Bar” }`3. at 6PM 01/01/2024 : We send Event B for User 1: `{ time: 4PM 01/01/2024 (BEFORE we set the identity)  }`4. at 7PM 01/01/2024 : We send Event C for User 1: `{ time: 7PM 01/01/2024 (AFTER we set the identity)  }`Our goal is that Event A and Event B should NOT have a user property `foo: “Bar”` set for the user, and Event C should.From the documentation here, We think this second sentence may be incomplete (there is no following

Hi!I want to remove a user (user_456) that I have in a group “organization_123” with type “Organization” but I can make it work via the API (Keep getting “missing event” 400 error). Can someone advice how it should be done? Endpoint: https://api2.amplitude.com/groupidentifyHeaders: Authorization: Basic XXXXXXX Content-Type: application/jsonBody:{ "user_id": "user_456", "group_type": "organization", "group_name": "organization_123", "group_properties": {}} Thanks in advance!

This has got to be something simple, but I can’t figure out what it is. I’m successfully tracking events with the browser SDK. I’m now trying to bounce certain events to my Node.js server and I’m getting the error “Request missing required field”. This is a console log of the error in the response from the `track` function:result { event: { time: 1723243786017, insert_id: 'beba5469-9fad-4e38-a9f9-b24de353cead', plan: undefined, event_type: 'Change Selected Legal Entity', event_properties: { user_id: '8d024873-5018-46c3-9520-9b288d0cc5b7', device_id: '68f3b49c-6854-47a9-bed2-5011470376b0', screenResolution: '2560x1440', appVersion: '0.0.0', isInternalUser: true }, event_id: 0, library: 'amplitude-node-ts/1.3.6' }, code: 400, message: 'Request missing required field'} My call to `track` seems correct given the docs:track(data.event, eventProperties)Event properties is a `Record` type with the properties you see in the error response. I

Hi, there!I’ve tried to connect with Adjust, but I couldn’t.In Data > Source page, Adjust doesn’t appear even if there are successful Attribution API histories in Ingestion Debugger(Data > Sources > Ingestion Debugger tab).I requested responses of Amplitude Attribution API for Adjust, and I received the log which means Adjust properly requested Amplitude Attribution API and got 200 from Amplitude.How could I connect with Adjust?

I’ve noticed a strange behaviour - we’ve started to receive lots of events (around 40K per day) from IP’s belonging to Yandex (russian search engine), and it looks like that their spiders run javascript code and trigger a corresponding events. And it’s ruining our analytics :(This happens only with Yandex spider. Do you have any idea why this happened and how can we fix it PS I know, that I can ban events coming from a certain IPs, but I think that the ignoring search bots should be done automatically. 

In our service, non-logged-in users can perform certain actions, and for the purpose of tracking metrics such as conversion rates, we need to uniquely identify these non-logged-in users.Previously, we identified these non-logged-in users as individual users through Device IDs. However, the problem is that these Device IDs frequently change due to reasons such as app updates, app/cache deletion, etc. This has resulted in instances where a single actual user is assigned multiple Device IDs, causing them to be recognized as multiple users.Fortunately, in Amplitude, even non-logged-in users with multiple Device IDs (but not assigned a User ID) can be uniquely identified as individual users through a single Amplitude ID.The issue arises when analyzing log data via queries in Google BigQuery, as Amplitude IDs are not separately stored. Therefore, users must be identified based on Device ID or User ID, and since non-logged-in users can only be identified by Device ID, the above-mentioned issu

We’re testing some changes on a squarespace shop page. We have 2 versions of the page made, the control and treatment. The experiment script snippet has been successfully inserted, and both of these urls have been defined. The experiment page says that the site is ready for testing.The issue is that when we try to look at the versions o the test from here, both versions just link back to the current, ‘control’ version of the site. This happens even when swapping the pages so that the ‘control’ is the B version and the ‘treatment’ is the A version - both links just go to the currently used page. On the treatment version, the url gets an address like:www.ourproduct.com/product?testing-feature=treatment&PREVIEW=trueBut it still goes to the control version of the page. I’m not sure why we’re not able to successfully see both versions of the test. Even tried enabling the experiment and opening links to the control page via private browser windows, etc - still getting the control version

I’ve followed the instructions to use an event stream to push conversion events from Amplitude to Google Ads which can then be used to optimize my campaigns by following the instructions here: https://www.docs.developers.amplitude.com/data/destinations/google-ads-event-streaming/. However, I cannot get the settings to pass the validation when I test the connection.I’m getting tripped up by the Google Cloud Service Account field and what to put in there. The instructions say “Enter your Base64 encoded Google Cloud Service Account.” and the link references the service account you need to create in Google Cloud Console. I’ve created this account and downloaded the appropriate json file, but it’s not clear what needs to be input into the Google Cloud Service Account field. I tried encoding the json file in Base64 and inputting that and I’ve tried the private key information (both encoded and not). None of which work. I get two kinds of errors...When I input the entire base64 encoded json f

After bulk loading backfill data, I cannot check this page.

Hi everyone :)Props on improving on every release! It’s looking very slick!!! Loving how it’s becoming a tad bit easier to use Ampiltude overall!I’ve found out over the course of the last few weeks that the performance in Amplitude has been impacted in Firefox.Selections take a long time. I’ve tried using Firefox in Private mode with no extensions enabled (ublock is disabled for Amplitude). This doesn’t happen in edge or webkit/blink-chromium browsers. Here’s a video:YouTube short videoNote:This is apparent in the entire site. Just for context:Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:128.0) Gecko/20100101 Firefox/128.0Macbook Pro M3 Pro (36GB/12core)

After creating a destination to Snowflake, I found out there is only one table EVENTS_ in Snowflake and missing Merged User Table 

Hello. I'm having a problem where some countries are showing up as "none" when I set the "country" filter. How can I fix this so that "none" is treated as a specific country? 

Hello. I'm having a problem where some countries are showing up as "none" when I set the "country" filter. How can I fix this so that "none" is treated as a specific country? 

Hi Community! My team is currently developing a new feature to show our professional users view and contact statistics. Currently we are sending view and contact events to Amplitude and need to access these in order to calculate the statistics, on demand, for each user and serve/present them in our App & web.My understanding is that there is no endpoint we can use in the Amplitude API that can return a selection of events, for a single user, for a configurable time-frame. Is this correct?If the understanding above is correct, as an alternative, we were thinking to integrate Amplitude with out ElasticSearch cluster using a Webhook.Additionally, looking at the Webhook configurations, it seems that the Webhook would be called for every event received. Is this correct? Is there a way to batch the calls to the Webhook (e.g.: Event Streaming configuration)? Also, is a request volume limitation (request/minute)?I would be must grateful if you could help us find the best solution for this

Hello! Do I understand correctly that in order for the number_of_projects to increase by one each time the event is triggered, I just need to send `identifyEvent.add('number_of_projects', 1);`? Or I'm wrong? Because I read that the data in amplitude is immutable. Or should every time you send an event about a user joining a project just send a numerical value corresponding to the number of projects at the time the event was sent?

In Segment, I have two JS sources for our Test and Production environments. And the respective destinations to Amplitude Actions.Both destinations share the same configuration. Nevertheless, I see sessions in the Test environment but not in the Production environment.Page calls & Track calls Mappings for Session ID in both destinations are configured as so:If I look at the raw events in the debugger, page and track events from both the Test & Production Environments come with an “integrations” object:"integrations": { "Actions Amplitude": { "session_id": 1644244386047 } } In Amplitude, Test & Production projects share the same session definition:Nevertheless, in Amplitude I can see sessions in session charts for Test project but not for the Production project.Any ideas why this is not working on Production?

Hi! I am trying to sync cohorts to Klaviyo, but the sync is not working. I keep receiving the following message:“The target of the sync at the destination has been deleted. Create a copy of this cohort and try syncing the new cohort, then contact your CSM if the issue persists.”I have already done that.I also deleted the Destination and created a new one using other API Keys. My current Destination has a “Connected” status.Can you please help me solve this issue?Best,Manuela

On our chatbot app, users can do the following for each query/response set:1.    Add Comments2.    Give Thumbs up3.    Give Thumbs down4.    Give Thumbs up with comments5.    Give Thumbs down with comments The app is triggering the following separate events in Amplitude. These are the event properties:Commentsname: commentvalue.comments: TRUEvalue.correct: NULLvalue.EventId: “unique identifier assigned to the chat interaction”Thumbs upname: feedbackvalue.comments: FALSEvalue.correct: YESvalue.EventId: “unique identifier assigned to the chat interaction”Thumbs downname: feedbackvalue.comments: FALSEvalue.correct: NO​​​​​​​value.EventId: “unique identifier assigned to the chat interaction” There are no unique events for thumbs up/down with a comment, and we cannot change that right now.In other words, when a user provides Thumbs down with a comment, the events come through separately (A Comments event and a Feedback event with value.correct=No). This is tied to the user interface, which

Hi everyone,I have an event total and a property which is a number as presented in the following screenshot:How do I measure the average weighted arithmetic mean each week? for example in the photo value 24 appeared 64982 times so I multiply them and so on and the do the weighted average of that week.Thanks! 

Hello Amplitude Community,We're encountering a challenge with tracking web push notification clicks using Amplitude, particularly when dealing with anonymous users who haven't yet interacted with our website. Here's an overview of our issue:We utilize OneSignal for web push notifications. When a new user clicks on a web push notification, they are directed to our website. However, since they are anonymous initially (because they are new, we don’t get an event for the click. As a result, we can't track these clicks of WebPush notifications in Amplitude. Amplitude assigns an external ID to users only after they interact with our website, complicating the tracking of initial web push interactions. Do you know about an efficient solution for this issue - tracking WebPush click for anonymous users?Best regards,Dor

Hi everyone,I’m currently implementing Amplitude in my Expo app (v51.0.14) using the latest version of @amplitude/analytics-react-native and Ampli.Everything is working except the logout part when I need to anonymize my user when he logout / delete his account. I tried to use the ampli identify function and providing an undefined userId and then tried the reset function but its not working the events are still being sent under the same userId.I’ve enabled logLevel to Debug and it log nothing when I’m calling the reset function, is it excepted or something going wrong ? Thank you :)

Hi,I’ve been trying to set Channel Classifier but can make it work properly.I created a new classifier from scratch and added one row with condition utm_source = Klaviyo. utm_source is the only column in this classifier.In my data I have ads event with the utm_source = Klaviyo but it isn’t classified as Klaviyo channel. Instead it’s classified as none.There is no way I can make this classification work. Need help. 

Hello, We are new to Amplitude hence we would greatly appreciate any help with our issues:We see some discrepencies in data & user analytics, any way to validate coming data etc. or how do we know it’s setup correctly? I understand that adblock can block some etc. but we see rather big discrepencies. How to properly create funnel for purchase 3 step purchase process? step 1 -> xx.pl/../anonymous/1/step1 step 2 -> xx.pl/../anonymous/1/step2 in between stripe payment that is redirecting to their website step 3 -> xx.pl/../anonymous/1/step3 but it turns out that those step3 etc urls can be revisited after purchase that I think is influencing funnels as someone that bought item is entering from email again the link with step3 which is counted many times. Can we create idk. specific events on each page to be triggered or what is the best approach to catch users in this funnel?We launched new language and entered new market which before URL has de. and for past 2 days it’s no