At Eyk, we are constantly working on making the analytics experience for our users a bit better every day. Having the right tools helps tremendously. Read more below about our search for a suitable embedded analytics solution and why we chose Embeddable as our technology partner.

But first, let me show one of the features we've shipped since the start of the partnership: the customizable E-commerce cockpit. Since release we've had many users reach out to share how it's transformed the way they start their workday, and some even have it on full-screen mode in their office now! 🔥

Read along to learn how Embeddable helps us ship better features faster 🚢

What is embedded analytics?

When we started building Eyk, we knew we had to build a truly remarkable analytics experience if we wanted to continuously deliver insights that actually move the needle. For this, we had to find a way to solve a hard technical problem (serving data visuals based on large amounts of data in a fast and secure way is not easy) while allowing for fast iteration and prototyping together with our users to get to the core of the business problems we are solving.

So where does embedded analytics come in? In the words of Rodel van Rooijen (read more about embedded analytics on his excellent blog here):

Beyond traditional analytics tools, embedded analytics seamlessly integrates data analysis and visualisation directly into the core applications, empowering SaaS B2B companies with the ability to offer their users a uniquely customised experience.

After comparing and trying out different solutions in the market, we found the technology that fits our needs.

What we were looking for at Eyk

When we started our search we jotted down the following topics that we would compare solutions on:

• Performance: we needed sub-second response times to ensure Eyk would not be the limiting / slowing factor in the analytics workflow of our users
• Full control over functionality and look & feel: we knew to build a remarkable analytics experience we could not be limited to a pre-defined set of charts and controls
• Fast iteration together with our customers:to support a workflow of quickly acting on customer feedback and testing new ideas
• Security: we needed a solution that could handle our single tenant architecture (>> a dedicated Google BigQuery data warehouse per customer, for strong data separation)
• Natively Embedded: to support complex interactions between the embedded data visuals and other components in our app

In our search, we considered the following solutions:

• Apache superset
• Looker
• Vizzly
• Embeddable
• Building ourselves, with Cube.js as a caching layer

Why we chose Embeddable

During the comparison, we were well aware that the decision we were making was a long-term one - as it would be hard / time-consuming to transition to a different solution later on.

After extensive comparison, trying out different solutions and prototyping, Embeddable came out as the clear winner for the following reasons:

• Full control & fast iteration: Embeddable strikes an ideal combination between having full control (by allowing to build your own components), and enabling quick iteration through the visual builder
• Security: they supported our single-tenant architecture out of the box
• Natively embedded: Embeddables are embedded in our app as HTML Web Components, which has many benefits over embedding through an iframe (which some of the other solutions offer exclusively)

Note: while part of the initial selection criteria, performance did not play a large role in the final decision because it turned out impractical / too time consuming to thoroughly test this across all solutions. What gave us confidence that Embeddable would be performant were the underlying technology choices: building on proven open-core caching project Cube.js.

What is Embeddable?

So what is Embeddable exactly?

Embeddable is a SaaS solution that allows you to build a snappy and fully customized analytics experience. From my view, the most important components are:

• A no-code builder where you can import custom data visualisation components and use them to build dashboard that embed natively into your app
• An API that allows you to set-up and maintain secure data connections
• A caching layer that queries your data warehouses, caches the results, and provides lightning fast responses to your web application

Or in their own words:

Embeddable is a developer toolkit for embedding lightning-fast, fully-native dashboards in your app. Embeddable takes a headless approach to give you full control over the UX and UI in code, while saving your developers time by handling things like interactions, security and performance for you.

Looking back

Now that we have been live with embedded analytics in Eyk for a couple of months we've launched many impactful features (examples below 👇) and feedback from our users has been immensely positive. I can happily conclude we made the right decision by partnering with Embeddable.

Lastly, while not part of the selection process and criteria at the time, there are two more clear reasons why working with Embeddable has been great:

• In our workflow we've seen the Embeddable become a platform for different disciplines on our team (eg data & frontend engineering) to work together toward a release
• The Embeddable team is very capable and friendly - they are truly committed to helping us succeed and provide outstanding support

For anyone looking to embed analytics into their application: Embeddable deserves your consideration. Check them out here.

---

A small selection of features we've launched in the last 3 months, since the start of the partnership:

A cohorts dashboard for analyzing Customer Lifetime Value in different groups (shown with fake data)

‍And performance dashboards like paid performance, email performance, product performance, and many others... (shown with fake data)