How Pace Revenue halved compute costs and bolstered its algorithm with Google Cloud
Meet Pace Revenue
Hotels are designed to cater to our desires. From snug beds that seem to make themselves, to en-suite facilities, comfort is king. Consumers spend hours comparing rates and dates via search engines, hunting for the perfect room for their needs. Behind the scenes, hotel operators put just as much effort into getting their room rates right.
Deciding what to charge for a hotel room is a constant balancing act. To maximize revenues, hotel operators need to set the rate of each room at the optimal price for each day, based on peaks in demand and other factors that are constantly in flux. In fact, the hospitality market is so fast-moving that itโs impossible for a human to keep their finger on the pulse of these factors. Thatโs whereย Pace Revenueย comes in, with a software that gets the job done almost imperceptibly, like a hotel bed made up every morning.
The challenge
โWhat our clients like about our pricing engine is the simplicity,โ explains Matt Yule-Bennett, Chief Technology Officer at Pace Revenue. The software uses machine learning to automate pricing decisions entirely or provide pricing suggestions, increasing revenues by 10% on average. โThey just let our algorithm do its thing, and theyโre often surprised at how much they can charge for certain rooms.โ
Helping hotels save big with a lean IT crewย
For the engine to โdo its thingโ, Pace Revenue needs to continuously run property management system data through complex machine learning algorithms. To scale with these data volumes, Pace Revenue leverages a lean infrastructure built entirely onย Google Cloud. โAll of our workloads are containerized, and almost everything runs onย Google Kubernetes Engine,โ says Matt. โThis resource-efficient Google Cloud stack enables us to scale on demand to handle large data volumes, even with a small IT team.โ
With less time needed to maintain infrastructure, the team is free to focus on new solutions and optimize the quality of the Pace Revenue pricing engine. This agility drives innovation at Pace Revenue, which has grown steadily since the first client signed on in 2017. To keep up with this success, the IT team often needs to master new cloud technologies and solutions.
Checking the stack, curbing the spend
โWe didnโt always have time to establish best practices before we started using cloud solutions, and we wondered if there was untapped savings potential in our stack,โ says Matt. โAnd of course, reducing our cloud spend became even more important when COVID-19 hit. Thatโs why we approachedย DoiT.โ
In selecting a partner, Matt had two main criteria: support to build sustainable expertise in the company, and affordability. DoiT surpassed Mattโs expectations on both counts. โWhat immediately stood out was DoiTโs business model,โ he says. โIf we buy our cloud services through DoiT, its consulting and technical support is included at zero cost. And from the get-go, we felt like the team went above and beyond to respond to our questions.โ
The solution
To begin working together, Pace Revenue first provided DoiT with read-only access to its Google Cloud projects configuration. One week later, Matt and his team received a comprehensive report highlighting all of the potential savings in the Pace Revenue IT stack, with detailed instructions on how to implement the changes. โDoiT gave us an incredibly thorough review of everything weโre doing on Google Cloud,โ says Matt. โThis included one major savings potential in our use of Google Kubernetes Engine that promised a massive impact on our bottom line.โ
Halving compute costs with a few tweaks
To power its pricing engine, Pace Revenue runs workloads on on-demand compute nodes with Google Kubernetes Engine. DoiT saw an opportunity to reduce IT overhead with little effort by running these workloads on preemptible VMs instead.
Preemptible VMs offer the same machine types and options as regular compute instances and last for up to 24 hours, while operating at a much lower cost than regular instances. DoiT showed the Pace Revenue engineers how to reorganize the architecture to make it more resilient and allow for the use of these more short-lived VMs. In less than a week, the team had implemented all the suggested infrastructure changes, cutting its compute bill in half almost immediately. โDoiTโs recommendations enabled us to swap 80% of our on-demand nodes for preemptible VMs on Google Kubernetes Engine, which come at only 20% of the price,โ says Matt. โThat amounts to a reduction of more than 50% in our overall compute cost, with very little effort.โ
Making time for hospitality
Instead of spending time and effort researching cloud optimization techniques, the Pace Revenue team can now focus its energy on tasks that directly improve the customer experience. โWe no longer have to spend time understanding how Google Cloud works because DoiT does it for us,โ explains Matt. โCombined with the compute savings, we can use these resources to make our algorithm more capable and run it more frequently. And that adds directly to our customersโ bottom line.โ
Working with its previous support services, Matt and his team often found it hard to get the point across with their questions, and misunderstandings were common. Today, this is no longer an issue, as Pace Revenue has a direct line to cloud engineering experts at DoiT for any cloud challenges. โWith DoiT, support tickets immediately go to reliable subject-matter experts. The team always understands where weโre coming from, and helps us get solving immediately. This creates a level of trust that we havenโt experienced elsewhere.โ
An all-inclusive cloud experience
Trust is a key ingredient in the relationship between Pace Revenue and its hospitality clients, and with DoiT, the team feels more empowered to make decisions that its clients can rely on. โOur customers are early adopters, and they trust our technology to take business-critical pricing decisions off their hands,โ explains Matt. โWeโre able to meet that trust by making smart decisions for our IT stack. And partnering with DoiT has empowered us to make smarter decisions that maximize profits for our clients.โ
The result
Besides more profits, clients can also expect more features thanks to Pace Revenueโs new analytics platform, built with the help of DoiT. Powered byย Google BigQuery, it helps hotel operators visualize data to drive decisions and strategies. โStanding on the shoulders of Google Cloud, and knowing that DoiT has our back, we can build better solutions faster, even with our lean IT team,โ says Matt. โIt enables us to leverage technology effectively and offer top-notch services that our hotel clients can depend on.โ
What's next?
Now that it has quick access to proven cloud expertise whenever needed, Pace Revenueโs IT team is ready to meet new challenges head on. โOur clients are a forward thinking-crowd, and weโre always trying to stay at the forefront of technology,โ says Matt. โOur ongoing partnership with DoiT ensures that weโre aware of all the amazing things happening in the cloud, and that weโve got the expertise at hand to make that technology work for us.โ