Saving development resources & delivering value with embedded analytics: lessons from SupporTrend’s customer story

  • 28 March 2022
  • 0 replies

Userlevel 5

This content, written by Tamara Ford John, was initially posted in Looker Blog on May 20, 2020. The content is subject to limited support.

At SupporTrends, hearing and acting on customer feedback is the cornerstone of their mission, and we’re amazed to see how they use Looker to harness this feedback.

Recently, Oliver Rowen, Founder, sat down with our team to share why he believes customer feedback is a critical — yet underutilized — resource. He also shared why his team decided to build a data product for their customers and the key criteria they considered when evaluating an embedded analytics solution to add to the SupporTrends product.

Prioritizing resources

When first developing their product, Oliver and team thought about their strengths. They knew how to extract customer insights and value from call recordings, surveys, and support tickets. The team wanted to apply that knowledge to their product and provide their customers with quick, real-time value by structuring unstructured data, while still allowing for the flexibility that would be key to their product offering.

However, the SupporTrends team knew they did not have in-house data visualization expertise, and they did not want to waste the time and resources it would take to get it. As their customer base quickly began to grow, Oliver and the team realized it was time to find a solution to their increased demand for visualizations without increasing the demand on their engineering resources.

We were managing both the natural language processing side and the visualization side. Eventually we found that we were spending more time trying to get our visualizations to render than we were building our core product: a fast, automatic natural language processing product. That was the day we committed to working with a company whose job was to make visualizations work well.
— Oliver Rowen, SupporTrends Founder,

From this early stage, they knew their answer to ‘’ would be to find an embedded analytics tool that would fit into their product, allow their team to continue focusing on their strengths, while also allowing their customers’ requests for robust analytics to be addressed.

Picking a solution

Thinking about the SupporTrends product goals vs. the BI product goals helped Oliver identify the most critical components in the search for the ‘best solution’ for their needs.

We learned very quickly that there is no such thing as a ‘best solution’. There is a ‘best solution’ for your particular application, but it is definitely not one size fits all.

As they began to evaluate, test, and deploy embedded analytics solutions, Oliver paid close attention to the following areas that he believes product teams should consider when evaluating embedded analytics solutions:

  1. Deployment time vs. visualization bells and whistles
  2. Engineering time and resource requirements
  3. User experience
  4. Customization opportunities & flexibility

1) Deployment time vs. visualization bells & whistles

When it comes to visualizations, Oliver thinks of this as an analogy for comparing a race car to a family sedan...

It’s pretty clear that the race car is a higher performance machine...but on the other hand, it takes a team of 20 people to get started in the morning and it has to warm up for half an hour. At the end of the day, it’s probably not going to get you to the grocery store and back any faster.

To help SupporTrends decide on what visualization features were most important and which wouldn’t get them to the grocery store — or deliver customer value — any faster, Oliver produced and stuck to a visualization checklist:

  • Do we know what it is and is trying to tell us?
  • Do we have the skills to deploy it?
  • Can we deploy it quickly?
  • Will users know what it is?

It’s easy to get caught up in the beauty of a visualization without staying grounded and knowing that the important things are checked off.

2) Engineering time & resource requirements

Oliver knew the engineering team didn’t have extra time to spend outside of their immediate initiatives, so they needed to leverage their time and resources as best they could.

It’s important to know what your goals are with engineering time. We did not want to reduce engineering time. We did not want to increase engineering time. We wanted to maintain engineering time and to improve productivity immensely given our amount of engineering time.

Thus, minimizing development needs for their engineering team became a critical part of their evaluation criteria for an embedded analytics solution.

3) User experience — such as SSO

Embedded single sign on was also an extremely important part of their embedded analytics criteria.

Our requirement with our analytics platform was not only to sign in once, it was to make it completely seamless.

The end goal was to choose a solution that would make transitioning between platforms as easy, quick, and secure for customers as possible.

4) Customization opportunities & flexibility

Early on, a large enterprise customer started using SupporTrends’ product and overnight, the number of users and data volume skyrocketed.

Knowing that transformational changes like this can happen with new products and initiatives (sometimes overnight), Oliver and his team wanted to continue to provide flexible and customizable options for their customers over time.

As you evaluate platforms, think about whether it has the right combination of off the shelf capabilities as well as infinite customization.

To prepare for this, the team thought about inevitable changes their customers would face, including how their companies will change, how many users they may need to scale to, what their users might want to see from their data in the future, and how their data stack may change overtime. Knowing that these are impossible yet inevitable questions to plan for, they wanted to build in as much flexibility and scalability as possible from the beginning.

Achieving 10x efficiency with Powered by Looker

By focusing on these four factors, SupporTrends found that worked best for their goals and delivering the fastest value to their customers.

Before ending their trial, SupporTrends embedded Looker within their product and took the visualizations they could offer their customers from line graphs to dynamic dashboards.

We improved our efficiency immensely. It took six months to build our platform and half a month to implement Looker visualizations.

Before settling on an embedded analytics solution, one cycle of implementing customer feedback into the SupporTrends product would take anywhere from 1-4 weeks to deploy. Now, the SupporTrends team is able to turn new requests and customer feedback into actions in minutes.

Today, things look quite a bit different,” shares Oliver. “Within 3-4 clicks, we can generate a new takes 8 seconds to design a new visualization that we can show to the customer immediately...Recently, we had a request come in and we were able to show it to the customer in a minute.

To learn more about SupporTrends’ journey from evaluating to deploying an embedded analytics solution and hear more about what they’ve accomplished with Looker so far, check out Oliver’s full webinar here: .

0 replies

Be the first to reply!