Schedule rides in advance

Introducing App Analytics in the Uber Developer Dashboard

February 8, 2017 / Global

Many of you have asked for better visibility into your apps’ performance on the Uber Platform. We heard you, and today we’re excited to announce the first version of App Analytics in the Developer Dashboard.

Thanks to this update you can now track HTTP traffic between your app and the Uber API as well as the number of clicks on deeplinks — all in real time.

This is just the beginning. We plan to add new charts in the coming months to help you track your integration’s system and business metrics.

The Details

Log into the Uber Developer Dashboard, select your app, and open the new Analytics tab to see three real-time charts:

One tracks all of the HTTP responses sent from your app to the Uber API, and another shows only the HTTP error responses (HTTP 4XX — 5XX). The Deeplink Impressions chart displays the number of deeplinks into the Uber rider app that have been attributed to your app using a client_id parameter.

You can use the date range selector in the top right corner to specify the dates you are interested in.

*Changing the date range in the Deeplink Impressions chart*

You can also use this new feature for debugging. After you add your first few deeplinks, use the Deeplink Impressions chart to confirm that you are receiving proper attribution. (For extra assurance that your deeplinks were created properly, check out our recently introduced Deeplink Generator.)

Analytics in Action

As an example, let’s look at how ACME, our amazing (and imaginary) integration, would use App Analytics throughout the development lifecycle.

ACME helps developers meet up for hack sessions at nearby coffee shops. The mobile UI includes a button that opens the Uber rider app through a deeplink, so that users can request an Uber to the cafe. But how does ACME ensure that Uber attributed the ride to its app? If the query parameter was formed correctly, the Deeplink Impressions chart should jump from 0 to 1. This tells the folks at ACME that users are able to request rides through their app, and they can be confident that Uber will correctly attribute those taps.

Fast-forward a few months. ACME has been gaining traction and one day it lands on the front page popular tech news site. This leads to 20 times more signups than yesterday. But how often are these new users actually attending the hack sessions? To find out, ACME’s team sets the date range to the last three days. This way, they can measure the engagement of their latest users and compare those numbers to the early adopters.

Looking Forward

We hope you’re as excited as we are to try out this new tool! This is our first iteration of the product, so please keep an eye out for additional charts and improvements in the near future. Also, if there are any other metrics that you think would help you build a better API integration, please let us know in the comments section or share your thoughts with us via our Twitter handle @UberDevelopers.

Things to keep in mind

To get the most use out of App Analytics, please keep the following details in mind:

On average, events show up within 30 seconds. Worst case, you might have to wait up to two minutes.
The time ranges are based on the time zone set by your browser.
Events are grouped by hour. If a deeplink is triggered at 3:59 p.m., it will appear in the 3:00 p.m. slot for the same day.
You can query back to January 20th, 2017 (UTC).
An event in the Deeplink Impressions chart indicates that your integration successfully opened Uber’s rider app and your client ID was recognized as the referrer. It does not guarantee that the deeplink resulted in a user taking any particular action while in the app, such as hailing a ride with Uber.