Planhat: Converting data to action with a 360-degree view of customer interactions
About Planhat
Founded in Stockholm, Sweden, in 2015, Planhat now has customers all over the world. Its next-generation customer relationship management (CRM) platform combines static CRM and time-series data. Designed to power customer retention and revenue growth, Planhat's service provides its customers with a 360-degree view of every customer interaction, powering workflows and insights from signals over time.
Tell us your challenge. We're here to help.
Contact usWith hundreds of new customers, CRM platform Planhat found itself supporting more than innovating. It turned to Google Cloud to enable time-series data analysis at scale and speed.
Google Cloud results
- Customers' most data intensive dashboards now render data up to 40 times faster
- Planhat seamlessly processes 80 terabytes of time-series data every day
- The team has cost-effectively optimized its storage by maintaining large datasets in BigQuery
- No longer weighed down by support requests, Planhat is free to focus on innovation
Google Cloud is helping Planhat to render data up to 40 times faster
How well does the modern business know its customers? While companies have access to more data and tools than ever before, customer understanding is often fragmented across departments—with each relying on its own strategy, software, and systems.
Swedish start-up Planhat knew that if businesses were to truly understand their customers and stay competitive in a rapidly-changing economy, where net revenue retention (NRR) is increasingly a defining metric, they would need a modern customer relationship management platform (CRM). So in 2015, Planhat built the first next-generation CRM, powered by time-series data, which businesses could use to break through their department silos and tap into a continuous view of their customers.
By bringing time-series data to every department, Planhat CEO Kaveh Rostampor and CTO Niklas Skog wanted to empower all employees to understand each unique moment in the customer journey. From drops in usage that flag potential future customer churn to pinpointing exactly the right moment for a cross-sell, Planhat powers decisions that drive customer retention and revenue growth.
More customers, more data
As the first CRM tool to combine the insights of time-series and static CRM data, Planhat quickly grew from a one-to-watch start-up to a rapidly growing scale-up. With that growth came more traffic, higher data storage costs, and slower data processing. For the biggest customers with the highest volumes of customer data and most intense data operations, the team was beginning to spend more and more time on tech support.
"For the majority of our portfolio, our current architecture still worked pretty well. But we wanted to push our customers to push even more data. We wanted to transfer even more power to our users."
—Yaroslav Sivachenko, VP of Engineering and Analytics, PlanhatWhat it needed was a scalable, efficient architecture that could power a rapidly-growing customer base with almost unlimited volumes of time-series data at speed.
Yaroslav Sivachenko, VP of Engineering and Analytics at Planhat, notes, "For the majority of our portfolio, our current architecture still worked pretty well. But we wanted to enable our customers to push even more data. We wanted to transfer even more power to our users."
Tapping into scalability and speed
While the Planhat team compared three of the top cloud providers, when it came to speed and scalability, they found that Google Cloud Pub/Sub and BigQuery were unmatched. In particular, BigQuery promised the scalability the team was looking for. Planhat was also impressed by its UX, tools, and strong community.
Planhat had been experimenting with features of Google Cloud since 2017, using Pub/Sub to organize its data flow more effectively. But in 2020, it decided to undertake a full migration to Google Cloud.
During their initial talks, Planhat and the Google Cloud team worked closely together to create a high-level roadmap of infrastructure solutions to support Planhat's growth. The company began widely using Google Compute Engine, which now sits at the core of Planhat's new infrastructure, enabling the team to run their workloads on Google Cloud hardware.
Removing limits with BigQuery
BigQuery also quickly became a core part of Planhat's service. With BigQuery, Sivachenko says, "You get that great feeling that you're not limited in any dimension."
The team offloaded its main application stack and shifted the bulk of its computation over to BigQuery, optimizing storage and improving cost efficiencies. It began using BigQuery to store and operate on large time-series datasets and drive custom processing. It also used BigQuery to power reports in the Planhat CRM—allowing for granular filtering based on a host of user-defined dynamic permissions and object filters, such as individual user access and customer segments.
"You get that great feeling that you're not limited in any dimension."
—Yaroslav Sivachenko, VP of Engineering and Analytics, PlanhatPlanhat used Google Cloud Logging and Google Cloud Monitoring to build alerts for their key metrics—something fundamental for helping it to analyze the large volumes of data it was processing as it continued to grow.
New architecture, 100% growth
With a new powerful, scalable architecture behind Planhat, the pressure eased for its tech and developer support resources. This enabled the team to get back to innovating and go all-in on deepening the time-series analysis its customers needed to grow and manage their customer relationships in the NRR economy.
Planhat's Product Manager Alberto Lambert says, "Because we now have scalable architecture that can handle deeper functionality, we no longer have to support edge cases and focus on fixing things. Now, we can focus on deepening our functionality and innovating. We can do more, and we can do it on a bigger scale."
Today, Planhat seamlessly processes a high volume of data, day in, day out. With Google Cloud, it's storing 167 billion aggregated data points for its customers and processing 80 terabytes of time-series data every day.
"Because we now have scalable architecture that can handle deeper functionality, we no longer have to support edge cases and focus on fixing things. Now, we can focus on deepening our functionality and just innovate. We can do more and we can do it on a bigger scale."
—Alberto Lambert, Product Manager, PlanhatNot only is the Planhat team delivering more customer insights thanks to how much data it can store, but it's also delivering these insights far faster than ever before. Its speed of metric calculation has jumped by 15x and its largest customers' dashboards are rendering data up to 40x faster, in seconds.
Supporting continued growth with historic time-series data
In the future, Planhat's team is excited to dive deeper into the insights that historic time-series data can provide—both for themselves and their customers. They are focusing on a new metrics and analytics setup that will serve as the foundation for a host of exciting future builds. The team says that knowing they can call on the Google Cloud team for insights at the right time will go a long way in maintaining their high development and delivery pace.
When it comes to its core architecture, Planhat intends to further modernize its infrastructure by shifting from the self-managed Compute Engine to Google Kubernetes Engine. It will use Kubernetes to support new data flow analytics and the deployment of even more powerful metrics at scale—supporting all customers in finding new transformative insights from time-series data.
Tell us your challenge. We're here to help.
Contact us
About Planhat
Founded in Stockholm, Sweden, in 2015, Planhat now has customers all over the world. Its next-generation customer relationship management (CRM) platform combines static CRM and time-series data. Designed to power customer retention and revenue growth, Planhat's service provides its customers with a 360-degree view of every customer interaction, powering workflows and insights from signals over time.