Deep dive in the dashboards [Episode #2]: Dashbird
Posted On 13/09/2018
Last month, we had a closer look at the dashboards inside the Google Ads system. For this next edition of ‘Deep dive in the dashboards’, we had a chat with Dashbird, a scaling platform to monitor & debug your serverless apps. In this post, you’ll learn how Dashbird set up the dashboards, how their customers use the dashboards, and why it’s a core feature of their platform.
What is Dashbird?
Dashbird is a useful platform for product teams who are building apps, using AWS Lambda. As a software provider, the first most important thing is to have a healthy app, with stable performance. When there’s a failure, error or memory outage, you need to be able to act fast. But if you don’t know there’s a problem, how can you act upon it?
Especially for serverless applications, it’s often a hassle to gain observability into the health of your application. What are the problems, errors, and the costs that go along with it? That’s where Dashbird steps in. Their app is designed specifically for monitoring & alerting on failures, specifically within serverless applications using AWS Lambda.
Dashbird contains a whole range of insightful dashboards, which show you at a glance what’s happening with your app.
Dashboards adapted to the customer’s needs
Let’s have a closer look at a couple of dashboards from the Dashbird app. Which extra insights can they provide to their users, thanks to the dashboards?
The app contains multiple dashboards for multiple purposes, instead of having a ‘1-fits-all’ dashboard. This is great for the user experience. For example, a software developer will want to have an overview of specific functions, or dive into error logging when there is a failure. The dashboard below shows the data for 1 function:
Now, a CTO might be more interested in a high-level overview of the app’s health. He will look at the overall billed duration, number of invocations, memory usage,… to spot where issues might arise. As you can see, the dashboard below gives a more general overview:
This is an important aspect when you’re creating dashboards for your platform. Much of the success rate will depend on how well you’re solving your user’s needs. So before creating a dashboard, ask yourself the question: “Which people are viewing our dashboards, and which information is crucial to them?”
From that end, Dashbird is a perfect example. The dashboards are a good fit for different user profiles. And what’s more, they even further optimize the dashboards, based on the customer’s feedback:
We have made a lot of changes in the dashboards in the past couple of months, to figure out what’s most useful to our customers. In one of our updates, we deleted the cost KPI from our dashboard, as we felt that it was of less importance to our users. However, we received lots of feedback from our customers, which showed that the feature was actually used quite frequently. So we simply added the element again to the dashboard.
Good dashboards are a living product. This is clearly also the case for Dashbird, and we literally mean”a living product”. Apart from the regular updates and changes, the dashboard allows for interactivity to zoom in or filter on what you need exactly. And on top of that, there is the possibility to view your performance in real time, so you can track issues and see changes as they happen.
How Dashbird built the dashboards themselves
Usually, when adding dashboards to your platform, you have 2 options: either you develop the dashboards yourself, or you integrate an existing analytics platform into your own software. In most cases, an integrated solution will be the fastest & most cost efficient in the long term.
However, there can be multiple reasons why you would decide to develop it yourself. In this case, the Dashbird team developed the dashboards fully in-house.
If analytics are rather an ‘additional’ feature to your platform, you don’t want to lose focus. Your developers should keep focus on your core business. However, it’s clear that for Dashbird, the dashboards áre at the core of their platform. Then, it’s of course more interesting to have full control & ownership over that software, so you can completely customize the analytics as needed. In addition, the company has the right skills ánd resources in-house to develop such features.
But of course, building & maintaining it yourself asks for more development time. In the long run, that also implies a higher cost. We asked Dashbird about their experiences with a build-it-yourself approach. They clarified the impact on the team:
Scaling is always a difficult aspect. For example, we’ve recently added a new pricing model to our app. It includes a free tier, which led to a much higher usage. So, our platform needed to scale urgently. Luckily, we have a great team with the right skill set to solve this. However, if your company doesn’t have the time or resources, an integration would be much more beneficial. Then, your integration vendor will handle the technical complexity and scaling.
Key take-aways from Dashbird
Are you offering dashboards to your own product users? Or considering to do so? Here are some key learnings we gained from Dashbird’s analytics:
Listen to your users: When creating the dashboards, try to think from a customer point-of-view, or even ask them for feedback. The better you fit their needs, the more value you’ll offer them.
Interactivity is key: The interactive mode is one of the most used features of the Dashbird platform. Allow your users to play around with the data.
Build vs. buy: When adding dashboards, always keep in mind your product focus. Do you have the time & resources to develop analytics? In Dashbird’s case, they did, but usually an integration can speed up your time-to-market.
If you have questions about integrating an analytics module, we’re happy to help you ahead. For more examples of customer-facing dashboards, watch out for the next edition of ‘Deep dive in the dashboards’!