AWS Business Intelligence Blog

How Amazon QuickSight empowered Treez to build an industry-leading analytics product at a lower cost

This is a guest post by Elling Hofland, Director of Product Management from Treez. 

In this post, we discuss how Treez built an embedded business intelligence (BI) dashboard system in our POS offering for our customers and their own internal staff using Amazon QuickSight. Treez is a leading provider of enterprise point of sale (POS) software for the regulated cannabis retail industry, serving thousands of customers. It’s a young industry with a large proportion of small businesses operating within it, and they work in a complex environment of regulatory scrutiny and change.

Our goal goes beyond providing technical solutions for cannabis retail stores and websites. We feel an enormous responsibility to help them understand their data and business as best we can. Our customers’ success is our success—that’s the approach that we take to working within our industry.

An urgent need for consistent, high-performance BI dashboards

We started our journey with QuickSight because we were being let down by our current BI platform. In short, it was falling apart. Digging into the challenges we faced, we couldn’t establish a stable connection with our data warehouse, were often simply not able to pull data through to our dashboards, and faced regular, large-scale data system corruption. We were facing the risk that our reputation for dependability and experience would go up in smoke.

Our biggest need from our BI platform was the ability to keep a synchronized, stable connection with the hundreds of gigabytes of data in our warehouse, and maintain industry-leading speed no matter what the query and analytical volume required.

Technical challenges and our new approach

Digging into the details, our legacy system regularly couldn’t sync with our Amazon Redshift data warehouse, plus the connection was unstable and often would drop out, impacting our oversight of customers and their ability to analyze their own data. We ended up using seven different Redshift connector approaches to try and solve this issue without any luck.

Even worse, the system had a habit of corrupting, with the connectors linking our BI tools to our data completely breaking. Each time this happened, it would mean we’d be facing a month-long process to manually migrate customer data into new datasets and reengineer our dashboards to pick up information from these. This meant a lot of our own time was spent on trying to fix these issues and minimizing the impact it was having on our customers.

Customer service also played a part in our frustrations. Our legacy provider was evasive with the issues we were having; they didn’t take accountability and support staff promised root cause analyses that they never delivered.

That’s why working with AWS and the QuickSight team was such a breath of fresh air.

We already had a relationship with AWS and knew the standard of attentiveness and expertise we would receive with a move to QuickSight. The team worked seamlessly with us; we collaborated to build a proof of concept system for embedding BI dashboards into our POS offering and never looked back.

The approach from AWS and the impact of QuickSight gave us three important initial impacts:

  • The reliability and speed of the system for delivering and analyzing customer information
  • The level of technical knowledge and proactive support from the QuickSight team
  • The response from our retail customers, who were excited by the efficiency and deep insight that QuickSight put into their hands

Planning, building, and delivering to a new standard

After we engaged with the QuickSight team, things moved fast. In a little over 6 months, we managed to go from proof of concept to a widespread rollout across our hundreds of external customers, not to mention using the pixel-perfect reports and customizable dashboards in QuickSight to analyze customer data internally.

Our overarching timeline, from beginning to full release, went as follows:

  • June 2022 – Conducted a proof of concept in QuickSight
  • July through November 2022 – Built the MVP of Retail Analytics, consisting of 10 QuickSight dashboards embedded directly into our POS platform
  • November 2022 – Unveiled Retail Analytics at MJBizCon in Las Vegas
  • January 2023 – Used Retail Analytics as the catalyst to migrate our customers to our new multi-tenant POS system

Moving away from a system that risked harming our business’s performance to one that delighted our customers in such a short space of time was extremely satisfying and shows the hands-on nature the QuickSight team had with building a system for our needs.

We used the SPICE system in QuickSight to create fast, effective, and reliable connectors. These synced with both our Redshift data warehouse and microservices datasets on Amazon Relational Database Service (Amazon RDS). The nature of SPICE and its automation means we no longer have to dedicate time to building and tuning queries. With data in the hundreds of gigabytes, we get powerful, incremental refresh capabilities, ready to be analyzed and queried, every 15 minutes. This keeps both our and our customers’ fingers on the pulse of their business performance.

Our success hinges on our customers’ success, and so QuickSight is offered as part of our existing software as a service (SaaS) package. Customers have access to the powerful BI dashboard system of QuickSight through their POS software, so it’s easily available for them to get insight on fresh data whenever they need.

For our internal QuickSight use, we again embedded the dashboards, this time in our Customer Value Stream portal. This means our cannabis retail experts have clear, up-to-date intelligence on how our customer’s businesses are performing, and allows them to help steer customers to greater heights from a position of real knowledge.

The following screenshot showcases our Budtender dashboard, which allows retailers’ sales staff to understand their performance trends, such as changes in average order value, average basket size, and average time spent with customers. They can even see their power ranking in several categories across their competitors.

It also helps them compare their category and brand sales to current trends within their retail location. This helps them identify opportunities for education on products or categories that represent a real opportunity.

Example of our Sales dashboard

The following screenshot shows a page from our Sales dashboard. Here, retailers can see which brands and product categories are popular. They can also understand attribution of sales to different order channels and the sources of orders from those different channels.

For example, customers can see what percentage of their online sales come directly from their website against those from a third-party ecommerce marketplace. This then informs where their marketing budget should be allocated and what return on investment it is delivering.

Example of our Sales dashboard

Sky-high impact: 190% dashboard views increase and 36% budgetary savings

We’ve seen substantial, tangible results since we moved from our legacy system to QuickSight.

Internally, it’s certainly been a major factor in us hitting our H1 2023 sales goals. Part of that is the hundreds of employee hours saved from manually firefighting underperforming systems, alongside a greater level of insight into our customers’ businesses delivered to our fingertips.

Retailers that use Treez have also responded positively to the change. We are seeing a 190% increase in dashboard views from our top-tier accounts, as they engage with a system that is intuitive, smart, and easily accessible.

Our bottom line has also seen strong improvements. Annually, we are spending 36% less than we previously were on business intelligence, despite being able to report and embed dashboards for customers.

“At Treez, we are drawing insights from data generated by users running their businesses with our software,” says Elling Hofland, Director of Product Management at Treez. “QuickSight allows us to feed those insights back into our software to allow those users to make better decisions in real time.”

Looking towards the future of Treez

It’s been a fast, exciting journey with the QuickSight team, and we’ve already got one eye on how we can continue to improve our solutions for ourselves and for customers.

Next on the agenda is fine-tuning our impressive Redshift data warehouse, moving our data timeframes from every 15 minutes to being completely real time.

The flexibility of the system also opens up opportunities to deeply tailor BI insights for individual retailers. We’re launching services for building and maintaining custom dashboards. Additionally, we’re launching a data connector product that allows our customers’ data systems to stay in sync with their POS for their own ad-hoc BI needs.

In an industry that is still growing and learning, and faced with a lot of regulatory requirements to fulfill, making the move to QuickSight has set Treez and our customers up for continued, informed, data-driven success.

To learn more about how QuickSight can help your business with dashboards, reporting, and more, visit Amazon QuickSight.


About the author

Elling Hofland is Director of Product Management at Treez and an expert in product metrics and analytics. Elling leads the development of data and analytics solutions for cannabis retail dispensary operators. Elling has a background in analytics and product management in sports data, having mined insights and created player pricing algorithms as an early employee of DraftKings. While at Sportradar, he led product management for the analytics software that powers broadcast insights and graphics for FOX Sports, NFL Network, and more.