Analyze-Automate-Accelerate: The New Model For Customer Focus

BrandPost By Marty Brodbeck
Oct 28, 2020

Here is Priceline’s three-step framework for doing more with less, and responding more quickly to evolving customer needs.rn

istock 1138834713
Credit: iStock

“Do more with less.” Sound familiar? For years, that’s been the dominant marching order for anyone running a company’s information technology department. What we don’t have is much of an industry standard method. As CTO at Priceline, I’m responsible for Product Engineering, Infrastructure, and Technical Operations, as well as thinking about our overall information architecture and our long-term technology strategy. I also work with our teams to build a corporate culture where we commonly think about designing new solutions and trying new approaches.

That’s not as challenging as it sounds, because we focus on one guiding principle: Whatever we do, it must be done with the mindset of, “how does this matter to the customer?” We measure customer demand, customer satisfaction, customer responsiveness—basically, if you can put the word “customer” in front of it, we’re interested.

We are working on a new model that helps our people do more with less, and respond more quickly to evolving customer needs. It began with Priceline’s three goals for IT in 2020. They are to 1) analyze data more effectively, 2) automate processes in the cloud, and 3) accelerate developer productivity. All are strong as separate goals, but they unify into something greater. Here’s the three-step framework we’re using to make these goals a reality.

1. Analyze data to use it more effectively

Enable data-driven decisions wherever possible. Measurement is the only way to achieve meaningful change in a broad way, and what you measure is a good way of staying close to your core goals.

In our case, these are centered on the customer experience. Our data analysis focuses on real-time information about how online shoppers are viewing and acting on our site and our products, and how they interact with us in the future. Much of the analysis is done in BigQuery, which supports streaming analytics, and is at the center of our cloud-based data infrastructure. We are now in the process of building new personalization and recommendation algorithms with machine learning, that we can scale up automatically to meet new demand. 

2. Automate processes in the cloud

Move more assets to cloud frameworks to minimize rote, individual work, and develop a single digital view of your online enterprise.         

Moving to the cloud to take advantage of managed services that reduce operational overhead has been key to our strategy. Our customer-facing front end, for example, is moving to Google Kubernetes Engine, which will enable us to design, launch, and respond to changing behaviors much faster. Google’s rich background in both data analytics (useful in leveraging more fixed data, like revenue forecasting) and ML (particularly useful with the more fast-changing customer and market data), along with its ability to automate and scale quickly and easily, were behind our choice of this cloud provider. 

Leveraging Google Cloud Platform to create a real-time data infrastructure, we can make data-driven decisions much faster. For example, as COVID-19 impacted the travel business, we quickly observed a drop in flights, along with a relative increase in hotel and rental car bookings. As a result, we tailored our offerings on to adjust for the change in demand. We were fast in spotting this opportunity because of the additional analysis and developer resources freed up by not having to do custom work. 

3. Accelerate developer productivity

Increase developer productivity and satisfaction by speeding their innovations to customers. Scaling benefits (and ending unsuccessful efforts early in the process) raises both developer and customer satisfaction. 

Already, we are seeing our software build times, and our time-to-market of new features, shrink from days to minutes or hours. This is because we have reliable, automated processes underlying the more value-adding human work of human innovation, trying things, learning, failing fast, moving forward. Naturally, access to a public cloud also means we don’t spend a lot of time and money on installing hardware and software to support innovation. We simply dial up and down our consumption, as needed.

Stronger together

Each element of this could be successful as single processes. Data analysis affords insight. Going to the cloud can save money and build efficiency. Developer productivity is increasingly important in a digital world. There’s even greater power, though, in viewing them as a single, linked effort. 

By finding and encouraging the linkages between the three missions, we create that process. For example, our online business has long experimented with innovations through A/B testing, choosing one of two approaches that resonate better with customers. Now, we’re leveraging a developer initiative with data, and then automatically scaling it with cloud services.

I must underline that this is both a process and behavioral change. Just lifting and shifting existing assets—a frequent recourse when people are under pressure to do more with less—can’t get you all the way there. You must bring to bear the most forward-looking and change-oriented parts of your company’s culture. Now is a good time for that, because few will question why, in these days of COVID-19, you’re trying something new.

The analyze-automate-accelerate framework of doing more with less is compelling precisely because it is responding to the customer, whose tastes and habits are always changing. Our real innovation is based on how well we think about that, respond to that, and build up for that. In a world where you either become a dinosaur or you evolve, it’s good to have a goal that’s always moving forward.

Keep learning: Discover how to drive large-scale digital transformations with a serverless smart analytics program built on BigQuery. Get the whitepaper.

About the author

Martin (Marty) Brodbeck is a Chief Technology Officer with over 20 years of experience across the financial, pharmaceutical, biotechnology, consumer products, media, and digital commerce industries. Martin specializes in leading technology transformation programs where technology is a key enabler for business change and growth. Over his 20-year career, he has led companies through mobile, cloud, big data, cyber security, infrastructure, content, and product engineering changes that have driven new revenue, costs savings, and productivity gains across enterprise companies. Martin holds a BA from the University of Richmond and an MS from the Stevens Institute of Technology. He lives in Armonk NY, with his wife Blakely and kids Connor and Charlotte.