My main use case for
Amplitude is to set up analytics dashboards for monitoring and understanding user behavior and spotting issues and bugs when we release software.
A quick specific example of how I used Amplitude for monitoring user behavior or spotting issues is when we want to release feature A and have the user stories that incorporate events, which pop up on the Amplitude dashboard. From that perspective, I could see the adoption, see if expected users have expected features, and if we see some spikes or drops, then we understand something went wrong. From the technical perspective, we monitor our API endpoints, seeing both successful and error responses, and we can set up alerts if something goes wrong to track down details and figure out what went wrong.
My main use case is very useful for exploratory testing, understanding user behavior, and feeding back into testing techniques and the software development life cycle. Things that we could miss during planning, things we are not aware of, we feed back, implement them, and see how it goes moving forward.
The best features Amplitude offers are useful dashboards that I set up easily from a user journey perspective, as well as a single point to understand one event. I appreciate how it integrates with other systems and all the insights you can drill into, giving you a lot of information for one of the users. You can see not only how they use the product but also which platforms they are on and which areas they are using. All those details help a lot to understand users from a holistic point of view.
The dashboard setup is easy to discover for a newbie; although sometimes over time, it gets complicated. Last year's release or UI redesign got a bit more confusing on how to make one or the other dashboard. In general, you have user paths, so there are two or three main paths to go through, which was good to understand. For integrations, we integrate with tools such as Slack and other monitoring tools as well as our SDK for our apps, allowing us to send events from our applications to Amplitude.
Amplitude offers a lot of features. I'm not an analyst to use all of them; I'm a quality software engineer, so I'm sure I didn't use all of them. For me, it was enough. From my perspective, the simpler, the better, so I can quickly leverage the features. Analysts might do more complicated things, and it was also useful; I remember you can use different queries if you need to find information very quickly, which was useful.
Amplitude has impacted my organization positively in a huge way. Everyone uses Amplitude if they want to have some learnings before implementing new features. It is used during software implementation to gain insights into understanding user behavior, which is massive for debugging issues. You can find the user, see where they're coming from, and what went wrong. So it is absolutely the same as on monitoring. Everyone in the business adopts data-driven development, and Amplitude is one way to understand that data and user behavior, which backs up everyone's work. This leads to a massive positive change, positive attitudes, lots of learning, and encouragement to use Amplitude in daily work for everyone in the company, not just analysts.
A specific outcome and metric that show Amplitude's impact is when I have a dashboard monitoring users logging into an iOS app for a particular release. I want to monitor, hypothetically, and it is useful if the chart is going smooth and the adoption of the new version is going up, which means things are going well for users. The version is out there, downloaded, and people can log into the app. If we can see previous app behavior, it indicates that if people downloaded the new app version, the previous one is going down. If things aren't going well, such as people can't log in or aren't adopting the version much, we drill down to understand what happened. Mostly, we monitor new feature adoption or existing core parts of the business functionality on a daily basis, having Slack channels notify us if anything goes beyond the threshold. This helps us spot lots of bugs, such as personalization that we thought some segments of people would get, but they were not getting it, allowing us to fix our personalization because we had insights and understood what segment of people did not get what we expected.
I would improve Amplitude by making it as clear and easy to use as possible. The feature discoverability could be simpler. I think AI could help in building queries and dashboards based on data, assisting people in learning about the tool, whether they're new or not analysts, helping them discover what they can do more. Additionally, insights on monitoring, identifying potential issues, and suggesting areas to look at to better understand trends on that dashboard would be appreciated. I realize that every context matters, but making it less confusing would be my main approach.
I choose nine because, lately, when I was using it, the more people wanted to improve, the more complex things got—not in a good way, but confusing. If I were using it right now without prior knowledge, I wouldn't know where to start. It matters when an expert is in the room and when new users want to do some things. I wouldn't know where to start; that would be my main concern—how you help people who are not data analysts or who want to discover their data to get into Amplitude.
I have been using Amplitude in my previous job for about two or three years.
My advice to others looking into using Amplitude is to understand what you want to get, what your main goal is because Amplitude can offer a lot.
I have a very positive experience with Amplitude. I enjoyed it a lot in my work. I don't use it now, but I hope to have a chance to use it in the future. I would rate my overall experience with Amplitude a nine out of ten.