Tips & Tricks

If you're ever feeling stuck or have a question about using Amplitude, you're just a few clicks away from real-time help—thanks to our new AI-powered assistant. Step 1:Click the (?) icon in the upper right corner of your Amplitude workspace. This opens our Digital Customer Success Center. Step 2:Explore everything the Digital Customer Success Center has to offer, including: Submitting feature requests Browsing help docs, the Amplitude Community, and Amplitude Academy Voting on integrations you'd like prioritized in our roadmap Checking operational status And most importantly… chatting with our AI agent! The AI assistant pulls from all our documentation to give you a synthesized, context-aware response—24/7. Step 3:Start chatting with the Amplitude Data Monster (yes, that’s what we call it!).It’s fast, responsive, and a great first stop for self-serve support. Keep in mind: AI might not always nail it on the first try—but with a specific question and a bit of feedback, it can ge

 Self-serve analytics is the goal for many data teams. But Saish Redkar, Senior Product Manager at DataRobot, cautions that “Without proper data governance self serve product analytics is a myth.” Explaining that, “Anything which keeps your data more organized and data quality consistent, be it any tools or processes comes under data governance. But not everyone is willing to take the data governance first approach in their implementations.So teams are always looking to get to their MVP of event tracking and create dashboards for their stakeholders ASAP. And it's tempting to roll out data governance on the fly, but the earlier you implement it, the richer the rewards it will give you.”This is a snippet from 7 Lessons from 7 Years using Amplitude in which Saish shares his insights from using Amplitude and from leading over 1300 discussions on data governance and product analytics on the forum.Related Tips: What is Data Governance? Insights Are Only As Good As Your Schema and Instrumen

 Changing your event naming convention can be a lot of work. But according to Thora Gudfinnsdottir, Product at Avo, “It is usually worth it if you have a really messy, confusing taxonomy, or the traffic plan is really difficult to understand.” But how do you go about migrating successfully? Thora says:“Start with a few events and a few properties that are important to you, and then gradually migrate to a better naming convention. You need to do it both on the taxonomy side and the implementation side. So it is a collaboration between different parts of the organization. And get reviews to make sure everything that you are doing makes sense. And of course, first, commit to and document what naming convention you are migrating to.”This is a snippet from Tracking Plan Tips: Event and Property Recommendations an Office Hours session with Thora Gudfinnsdottir and Anuj Sharma, discussing data design, data governance, and tracking plan best practices.Related Tips: To Track What Matters: Start

 Amplitude comes with many different chart types and ways to analyze data, but that doesn’t mean you need to use them all. Saish Redkar, Senior Product Manager at DataRobot, has found that “80 percent of your product insights will be always derived from 20 percent of the chart types.”But you shouldn’t feel pressure to use them all at any one time. “Sometimes I have seen that users try to force and extract insights from so many different chart types, but end up not having the best of experience. And that is totally acceptable - it's okay to not leverage all the charts out there in Amplitude.”This is a snippet from 7 Lessons from 7 Years using Amplitude in which Saish shares his insights from using Amplitude and from leading over 1300 discussions on data governance and product analytics on the forum.Related Tips: Insights Are Only As Good As Your Schema and Instrumentation Event Tracking is Not Product Analytics 3 Ingredients to Data Governance Success

 Both tags and categories can be helpful in making events more discoverable within Amplitude, but which should you use? Anuj Sharma, Sr. Customer Success Architect at Amplitude, shares this distinction:“Categories kind of restrict you to adding one category to each one of your events, but you can add multiple tags to those events.So your tags could be based on your development team, could be based on sprint release cycles, or releases, a number of different tags that you can add, right? So using tags on top of your category makes it easier to filter those events.You can filter by a category, add an add filter to it, filter by tags, and keep adding those filters on top of it. So categories and tags would be the best solution to finding the right event you need.“This is a snippet from Tracking Plan Office Hours with Thora Gudfinnsdottir (Avo) and Anuj Sharma (Amplitude). They share best practices in event data design to help you get the data you need without blowing up your event volume.

 Amplitude is really useful during fundraising, as a source of insights and for communicating your growth story. However, while it may be tempting, Jamie Dunbar Smyth Chief Growth Officer at COMIS, advises against giving investors access to Amplitude or using it live during fundraising pitches. Here’s why:“You can either use Amplitude to showcase it to your investors, or you can build slides. I recommend slides because if you open up Amplitude to savvy investors, they will start asking you questions on the spot.And if you don't really understand how to best answer those questions, it's not a great experience. So control the experience, take the data, put it into slides and be ready to answer those questions. In Amplitude, you have your metrics that you can set up, define your metrics ahead of time, make sure everyone speaks the same language.So when you speak to investors, you know, all of you can almost answer the same question and you all are speaking the same thing.” This is a snipp

 Deciding whether something should be a user or event property can be tricky. Things can get a little blurry. Thankfully, Anuj Sharma, Sr. Customer Success Architect at Amplitude, has some tips to help guide you to the right choice:“Anything that adds more context to a user is a user property, anything that adds more context to the actual action, that's the event property.But another way to look at this, anything that should maintain state with multiple events in a session that can also be considered as a user property.So things that are maintaining state would be user property as compared to things that would change within the session that would be an event property.”This is a snippet from Tracking Plan Tips: Event and Property Recommendations an Office Hours session with Ampliteer Anuj Sharma and Avo’s Thora Gudfinnsdottir who discuss data design, data governance, and tracking plans.Related Tips:Boolean Properties? Just Don't. Generic Vs. Specific Events Event Tracking is Not Prod

 Wondering what’s the best way to integrate Amplitude? Anuj Sharma, Sr. Customer Success Architect at Amplitude, suggests it’s a trade-off between time and control.“If you're in a time crunch, use the tag manager or the templates that we have in tag management. If you have the dev resources and you can spend time in building out a taxonomy for Amplitude, use the Amplitude SDKs, get events directly through the SDK and have a customized taxonomy.”Ideally, when you start breaking down your business questions, you'll identify the events and properties you need, and that will build the basis for your taxonomy, you can “then reach out to your devs to get that data into Amplitude.” However, “when you're using the GTM template, on the other hand, it's the other way around - you're implementing the template. If you're building a taxonomy, that's much easier to understand for your end users. That's the way to go. But if you're in a time crunch, if you don't have the dev resources, GTM template w

 According to Saish Redkar, Senior Product Manager at DataRobot, “You can only get so much from Amplitude if you're tracking plan isn't designed well enough.” If you don’t nail that, then your insights will be limited. To overcome this, he has the following tips:“Spend time on the tracking plan, give it the love it deserves. A good start to creating your tracking plan would be instead of asking, what do we need to track? You need to ask, what do we need to understand from our product?Keeping the event naming convention as simple as possible. Capture enough context with your event properties. They are the superchargers to your tracked events.You also need to try balancing out the insights to noise ratio in any tracking plan.”This is a snippet from 7 Lessons from 7 Years using Amplitude in which Saish shares his insights from using Amplitude and from leading over 1300 discussions on data governance and product analytics on the forum.Related Tips: How to Simplify Cross-Device Event Tracki

Amplitude users can now select pie charts as a visualization for most segmentation measures!(Note: You can have up to two segments in this visualization.) You can find more platform release highlights from September here:  

 If you’re wondering how to implement event tracking across devices, Anuj Sharma, Sr. Customer Success Architect at Amplitude, shares this reminder:“You'd have to use  three different events if you were looking at iOS, web and Android, as compared to if you had a single event with an event property for platform. So absolutely a single event, for different platforms. And if you were using any of the Amplitude SDKs, they by default, track platform property. So you don't have to go to the extent of instrumenting that separately. It's out of the box.“This is a snippet from Tracking Plan Office Hours with Thora Gudfinnsdottir (Avo) and Anuj Sharma (Amplitude). They share best practices in event data design to help you get the data you need without blowing up your event volume.Related Tips: The Power of Detailed Event Descriptions 4 Elements of a Good Event Name Generic Vs. Specific Events

 Implementing event tracking is a big, important step in your data analytics journey. But as Saish Redkar, Senior Product Manager at DataRobot, highlights - there’s more to it than that.“I've often seen that product analytics setups take a big pause after event tracking. I've tried to break down successful product analytic strategies, which I've come across into three major buckets: Right questions Right data Right tool setupSo you need to know the why behind your product instrumentation process and start asking better questions. It's important to chalk out the key questions you wish to answer about your product beforehand and try to iterate upon them as you progress.And the right data  The questions you need to be asking is, is your product infrastructure able to capture the right event data, which will help you inform and answer your questions?And the third one the right tool setup. A bad tool setup will eventually cost your teams and analytics efforts, and you won't be justifying th

 If you’re wondering what does Data Governance really mean, Jennifer Rudin, Sr. Data Strategist at Amplitude offers up this explanation:“Data Governance is a lot of things. But essentially it's a way of setting up rules and processes to label, organize, safeguard, and ensure high quality and useful data.  The goal is to break down data silos and build up data trust.It's to bring some form of data accountability to ensure the most relevant and consistent data is produced for every team to benefit from.  Everyone essentially is responsible in some capacity for data quality. And Data Governance never ends. It should be iteratively practiced as frequently as your product is iterating and as frequently as your data use cases analyses are expanding.”This is a snippet from Data Governance: Lessons From the Field in which four brave Ampliteers step forward and share their stories of data mishaps inside our Amplitude setup and across a range of customer use cases.Related Tips: 4 Stages of Data

 Anuj Sharma, Sr. Customer Success Architect at Amplitude, says there are 3 important areas to a Data Governance function: “One would be Education. Second would be Instrumentation and third would be Maintenance.”So you need to “Make sure you have enough documentation to educate your users, be it in terms of what your events mean, be it in terms of what your taxonomy naming conventions are.There is instrumentation, which is making sure your devs understand those conventions and are ready to instrument that and follow the same standard across the entire instrumentation.And finally maintenance, which is making sure that once you've instrumented your data, you're maintaining that taxonomy. You do need a regular cadence where you meet, you look at your entire taxonomy, make sure the events that you have are usable. If there are duplicates, you're cleaning those up. If there are events that no longer are used, you're hiding those from your taxonomy to reduce end user confusion.Maintenance is

 When it comes to tracking, it can be tempting to track everything. But Thora Gudfinnsdottir, Product at Avo, recommends you don’t do that: “That's going to make you end up with a bunch of events that you'll never use and going to just inflate your volume and that's going to increase cost. You want to track what matters.”So how do you do that?“Start with the why. So instead of just when you're creating a new feature, just trying to figure out all the events that you can. Start with the questions like, what are your goals? What are the metrics that you need to be able to measure your goals? And then in the end, what are the events that you need to construct the metrics?”This is a snippet from Tracking Plan Tips: Event and Property Recommendations an Office Hours session with Ampliteer Anuj Sharma and Avo’s Thora Gudfinnsdottir who discuss data design, data governance, and tracking plans.Related Tips: 4 Stages of Data Governance Maturity Event Tracking is Not Product Analytics The Po

 When thinking about whether to use a generic or specific event, Anuj Sharma, Sr. Customer Success Architect at Amplitude, says the important question to ask is, how critical is it to your user flow?“Is it a part of your critical user journey? If yes, you need that event individually.  If no, and you just need to build a specific user path, it could be a generic event. We definitely don't want to be in a situation where you're trying to build a funnel and all that funnel contains is ‘element click, element click, element click’, or ‘page view, page view, page view’. That wouldn't really make any sense, right? So if there are critical user parts in your application, that's where you should be pulling your main events or the critical events from.”This is a snippet from Tracking Plan and Event Schema an Amplitude Tracking Plan Office Hours session with Ampliteers Jennifer Rudin and Anuj Sharma, who answer users’ burning data governance questions, covering when to use events or user proper

 Shoin Wolfe, Founder at Shoka Sonjuku, says that “good dashboards are GAS.”What does he mean by that?“So good dashboards are Governed Actionable Storytelling.Governed would mean that you have someone has to be in charge of the metric choice. The quality of each chart, the chart type is adjusted according to changing needs. So it's iterable and dynamic.Actionable is pretty easy. I think the golden rule if it's a metric that doesn't change your behavior, it's a bad metric, right? If that's true for metrics, it's true for dashboards too. I would add to actionable that the charts are interactable. You can do something to the chart. You can look into it. You can segment it. You can deep dive into it.Storytelling is very important. Does something have a narrative? Is it easily digestible?”This is a snippet from Good Dashboards vs. Bad Dashboards an How I Amplitude session with Shoin Wolfe, where he talks through patterns he's seen across growth teams and what makes some dashboards drive act

 What makes for a good event name? Thora Gudfinnsdottir, Product at Avo, says there are 4 key elements to consider:“The first one is the object-action structure. Usually event names consist of an object and consist of an action. It's like  'sign up completed', or 'game started'.Then we have the casing. Just pick one casing and stick to it. I really like title case because it's human readable.Then there's the tense, past tense, present tense, whatever you pick, stick to it. I like past tense. When you're analyzing data, you were looking at actions that were performed in the past. So it just kind of makes sense. Then there's the language. So don't try to be fancy or use industry standard words, just use what your company uses. Some examples of event names: 'signup started', and 'profile configured'.”This is a snippet from Tracking Plan Tips: Event and Property Recommendations an Office Hours session with Ampliteer Anuj Sharma and Avo’s Thora Gudfinnsdottir who discuss data design, data g

 Data Governance is a constant work in progress. However, Stefania Olafsdottir, CEO at Avo, has identified 4 stages of maturity:“I think we, as organizations typically go through stages of, data quality and data governance, we go from winging it where there is no data because we're just starting a product or there's no tracking or something into the wild west.Where it's basically every team for themselves. So we have a lot of data and variants of all of the versions of the data   some humans are able to use their own data, hopefully, but probably in most cases not, use other people's data. And from the wild west, we go into something like a centralized governance.And in that state, we typically have a person or a team that ends up owning data governance - they saw that it's completely dysfunctional to have no standards if they are going to help teams do that. And then they graduate from that, centralized governance state to a, the utopian state, I would say, like the ultimate one is se

 Activation in software takes a customer from initial sign-up all the way to establishing a habit around the product. But from an instrumentation perspective, where does that start and end? Claudia Canamas-Donnelly, Senior Product Analyst at Appfire, says there are four key events to track:Signed up Setup moment Aha moment Habit momentHere’s how it works:“The Signup moment, someone creating an account on an app for the first time.  Followed by the Setup moment, which is the user has done things that sets them up for the aha moment. Usually means they have entered additional information or configured settings that they need in order to be able to get value from the app. And we have the Aha! Moment. It's the first time that the user has experienced the core value proposition of the app .And lastly, the Habit moment, this is the point where they've gotten that aha moment enough times that they've established a habit around getting value from our app, and we are pretty confident that we're

 Creating a good dashboard can be tricky, but according to Shoin Wolfe, Founder at Shoka Sonjuku, there are three common mistakes you need to avoid while creating them:“Three things that make dashboards bad in general. I think the biggest  thing is data overload, poor data visualization in which you can't make actual actions from the data you look at, and lack of customization. A lot of dashboards they exist as dashboards and you're not allowed to customize them.”This is a snippet from Good Dashboards vs. Bad Dashboards an How I Amplitude session with Shoin Wolfe, where he talks through patterns he's seen across growth teams and what makes some dashboards drive action, and others leave teams in a reactive state.Related Tips: Good Dashboards are GAS To Track What Matters: Start with Why Insights Are Only As Good As Your Schema and Instrumentation

 Boolean properties are good: True or false? According to Thora Gudfinnsdottir, Product at Avo, they’re best avoided. “If anyone ever creates a boolean property in our tracking plan, I will not be happy. Maybe even mad.”Here’s why:“When you start a product you maybe only have one authentication method and that's email and then you don't need a property for which method it is. Then you add facebook and then you add the property used facebook: true or false. And then you add google and then you have to add used google. And then you add github and then you have to add used github. So this doesn't scale. Then you have so many properties that are describing the same thing.”This is a snippet from Tracking Plan Tips: Event and Property Recommendations an Office Hours session with Thora Gudfinnsdottir and Anuj Sharma, discussing data design, data governance, and tracking plan best practices.

 As Saish Redkar, Senior Product Manager at DataRobot, says “Product analytics is a team sport. If you don't pass and share, no one wins.”To that end, he suggests 3 things have helped him to boost internal Amplitude adoption:“Notebooks are a powerful feature. I use them for defining and sharing our onboarding guide, the FAQs, and data storytelling. I think data storytelling is the major aspect of how you are sharing your insights and how we are communicating to the stakeholders.Slack integration has been powerful. I've seen some impact after enabling the unfurling of chat links in Slack. People are getting to see the first hand view of the chart inside Slack only.Team specific office hours were a huge bonus for us.  A very team specific or focused hours, like just having engineering versus customer success gives more value because the questions you can answer them are tailor made to that.”This is a snippet from 7 Lessons from 7 Years using Amplitude in which Saish shares his insights f

 Why bother adding event descriptions in Amplitude? Anuj Sharma, Sr. Customer Success Architect here at Amplitude, explains:“Descriptions make it really easy to understand what that event is for. If there's a new user who's onboarding to Amplitude, onboarding to your taxonomy, descriptions make it a lot easier for them to follow along and understand what that event really means. So you should definitely add descriptions to your event, event properties, user properties, and make it available, in your event dictionary within Amplitude, wherever possible.This is a snippet from Tracking Plan Tips: Event and Property Recommendations an Office Hours session with Thora Gudfinnsdottir and Anuj Sharma, discussing data design, data governance, and tracking plan best practices.