Digital is such an ingrained part of our world that it can be difficult to remember how young—and therefore still evolving—it is. Ten years ago, you could pull easily available data directly from Google, find some relevant keywords at a decent price with satisfactory traffic, and sprinkle those through your website or other content.

In 2011, Google cut off access to public search data. Two years later, with the release of Hummingbird and introduction of semantic search, traditional keyword search became irrelevant. Semantic search gives Google the ability to decipher query intent to produce better search results, which changed the customer acquisition game.

Now, with the advent of artificial intelligence, natural language processing and virtual assistants such as Amazon’s Alexa and Microsoft’s Cortana, we’re starting to see the next evolution of online customer acquisition, according to a panel of experts at SXSW 2017.

For example, if you wanted to stop and get a bottle of wine on your way to your friend David’s dinner party ten years ago, you’d figure out the general area and do a Google search for wine shop, Hyde Park. Then you’d spend half an hour combing through the results to see which store was on the route you wanted to take and would be open when you needed it. If a wine shop in the Hyde Park area was looking to nab those customers, their best bet was simple: just grab the keywords.

Today, you might pull up Google Maps, get directions between your house and David’s, then do a search for wine shop right there in the map interface. The information you need—such as store hours, address, phone number and even customer rating—is all right there on the left side of your screen. In just ten years, the entire nature of search has changed. Semantic search has enabled a much more intuitive user experience, and customer acquisition strategies evolved from strictly employing keywords to deciphering a customer’s intention.

But in the not-too-distant future, that same search will be more like this: Alexa, where can I buy a nice bottle of wine between home and David’s house? That’s a whole new spoken world of search and discovery that poses both opportunities and problems for marketers and business owners.

The future of customer acquisition

Search today is a “lean forward” activity. But it’s turning toward a “lean back” approach. In other words, today users sit down at their computers and actively engage in searching for what they want. They try different tactics if their first search results aren’t what they’re looking for, and they’re willing to scroll through a couple of pages of results or click on a few links until they find what they want.

But it won’t be long before this type of intensive search is a thing of the past. With AI and machine learning, users will be presented with results before they even ask the question. Search will become a two-way conversation.

Alexa, where can I buy a nice bottle of red wine between home and David’s house?

How much do you want to spend, Kris?

Under $25.

Joe’s House of Spirits has a Beuhler Zinfandel Napa Valley 2012 on special for $20. Would you like to pre-order for pickup?

What all will Joe’s House of Spirits have to think about and anticipate before nabbing the top recommendation, and what will such a platform even look like on the backend? Even the experts don’t yet know.

Marketers and business owners need to begin focusing on where they can add value for the user instead of on keywords. They’ll need to anticipate the situations that will create opportunities for their brand and find ways to leverage them. It’s similar to the channel’s growing focus on solving for business outcomes.

For example, maybe a small business is growing out of its current space and looking to move. They may do a mobile search for office space for rent in Oak Lawn or tell their virtual assistant Find me a commercial realtor in Dallas. Odds are they’re going to need someone to move their IT infrastructure, too, and perhaps configure it to scale with their growing business. Partners that can figure out how to get in front of new technology will have a big head start.

Managing the data chaos

Over the last few years, we’ve become a data-obsessed world. The more data we have, the better we can target potential customers and identify all those wonderful “actionable insights.” But Amanda Richardson from HotelTonight told the roomful of attendees she’s not even sure how much data she actually wants to own.

In the world of last-minute lodging search and discovery, there is a massive amount of data points that ultimately play into which hotel to recommend to users. What’s the weather like today? Does the user want to go out or stay in? Is this a romantic getaway or a business conference? Does the user want a spa? A golf course? A four-star restaurant? Vegan room service options? Eco-friendly housekeeping?

To build a system that can gather, manage, and mine that data would be an enormous undertaking, says Richardson. But through strategic partnerships, HotelTonight can gain access to that data without having to actually own it. The National Weather Service can provide one set of data points, for example, a directory of vegan-friendly eateries another. Find a business with a complementary (not competitive) offering and see if you can leverage the data they already have to create your AI layer.

These issues aren’t going to become a point of concern overnight. We’re talking about the future here, after all. Alexa is still in its infancy, as any user who’s spent frustrating and fruitless minutes trying to find specific information can attest. But like all technology in the digital age, it will be here before you know it. Businesses should start thinking about where they’ll fall in this new ecosystem and then see how it evolves over the next couple of years. That way they’ll be poised to pounce when the time comes for Alexa to start making natural recommendations for IT service providers.