You can’t miss the buzz about Big Data—massive quantities of valuable  information that can be hard to analyze and act on—and how ecommerce companies can use it to deliver more relevant customer experiences in real time.

Most companies understand the potential Big Data holds, and the opportunity it gives them to spot trends in customer behavior and respond quickly.

But many are still unsure how to use, share, and act on the Big Data they have. According to one of research firm eMarketer’s latest reports, “How Retailers are Leveraging Big Data to Personalize Ecommerce,” 45% of companies reported they weren’t using data effectively to personalize their marketing efforts.

So how can you turn your big datasets into an asset—instead of an untapped resource? A few keys:

1. Develop a data strategy. Big Data is more than just reporting. It’s a different set of data that offers another way to view your customer. That’s why it’s critical to figure out exactly what you want to know before you do deep dives into the data. Do you want to use Big Data to detect fraud? Stay compliant? Serve more accurate product recommendations? Big Data can do a lot and it’s critical to have a clear goal in mind.

2. Know how to use your data. You have to know how to view the data, how to slice and use that info to develop features that matter to your customers (e.g., product recommendation or personalization engines). So it’s likely you’ll need to partner with a vendor, or hire a data scientist who can help you build and maintain a data system. Bottom line: Working with Big Data is an investment, and you’ll need experts who can help you.

3. Test your ideas. Once you’ve sliced your data down to focus on a specific issue, and discovered trends in the numbers, it’s time to think outside of the box. Ask for ideas about tests you should run to see if you can respond to the trends. Example: You can use Big Data to find underserved segments of your website visitors, like international browsers who abandon before checkout. But then you have to figure out how to serve them. That could be by testing geotargeted banners that let those visitors know you ship to their country, or setting up your site to automatically translate to the language of the country they are browsing from. The point is that once you understand the data, you need ideas and tests that will use that information to optimize the website experience for your visitors.

4. Keep your eye on technology. Right now, Big Data is a complement, not a substitute, for understanding your customers. So it’s important to stay on the lookout for developing technology. As you get more comfortable, you’ll be able to pose and answer really innovative questions about personalizing the online experience.

The truth is that we’re just scratching the surface of what big data can do. But it’s critical to start developing a plan and resources for where to dive in.

Tom Janofsky is Monetate’s Vice President, Engineering, and a software engineer with 15 years of experience in successfully delivering enterprise and internet software. Prior to joining Monetate, he started and ran a software development consulting firm, where he designed and developed software as well as led teams that built large-scale, web-facing transactional software for clients that included EMC, FEMA, JBoss, and Pearson.