There's a quiet war brewing among technology companies – one that , if successful, will shift billions of dollars of media spending while upending the fundamentals that underpin how marketers reach consumers today. It's the war on search, a war whose most interesting battle will be the one Google is waging against itself.
Imagining the world after search brings to mind William Gibson's lesson from his book Distrust That Particular Flavor that "imaginary futures are always… about the day in which they're written." When I predict that search will be dead in 2020, it's the embodiment of a vision that several technology companies are trying to bring to fruition today.
There are several intersecting trends that may contribute to a future where the need to search is minimized and preempted. All of the trends center on the availability of data, which will in turn be processed, analyzed, and applied in ways to anticipate consumers' needs before those needs are expressed. The biggest contributor of data is consumers' mobility, and the precise location beacons will contribute to the precision of the predictive analytics.
The idea behind this has been around for a while – the proverbial "walk by a Starbucks and get a coupon texted to you" scenario. But it's far bigger than discounts, and the data's far more powerful than that which indicates when one happens to walk by somewhere. The future hinges on mobility-fueled intelligence that knows that on Tuesday at 4, it should prepare a shopping list for you because you do your grocery shopping after work. It knows you dropped clothes off at a different dry cleaner so there's an opportunity to get you to switch. It knows that when you're traveling, it can guide you around the airport to your favorite kind of car at one of your preferred rental car companies without you ever needing to book it. All of these scenarios minimize or obviate the need for searching at all.
In late September, Google released what may prove to be an early prototype of such a service with its Field Trip app, now available for Android, with an iOS release coming. The app proactively serves tips about local restaurants and bars, architecture, historic places and offers. Google has an outsized advantage when it comes to fine-tuning recommendations, given its experience and leadership in search, display and mapping coupled with running the world's most popular smartphone operating system. That also means Google is exploring other business models that may diminish its cash cow of search engine marketing. The bet that Google is making is that consumer behavior will shift so Google needs to stay relevant and profitable however those shifts happen, even if it means disrupting itself.
Such shifts present a number of questions that marketers will need to prepare to answer in the coming years:
How likely is such a scenario to happen and how soon will it happen? This can be tested empirically by monitoring the growth and volume of various search terms over time. Over the next couple years, changes may amount to rounding errors. It's quite possible though that there could be sudden spikes. Consider what happened with mobile search. During the previous decade, most online marketers and publishers saw mobile search volume leading to their site mired in the low single digits. Then, starting roughly around 2010, that began to jump from 5% to 10% to 20% to the point now where for just about any brand and category, the mobile volume is clearly significant – and for some it rivals desktop volume. It was the embodiment of the sleeping giant phenomenon. Could predictive recommendations be the next one?
Which categories of search terms will be most affected? While many categories may be impacted, travel seems most ripe for disruption by predictive recommendations given how well it's fueled by a large volume of mobile-driven data. It's relatively easy to detect patterns there as well, distinguishing between someone who takes one trip a year or 20, or whether the person taking 10 trips a year visits the same city during the week (probably visiting another office) or 10 different trips over weekends (probably a leisure traveler). It's not just mobility that matters. What if Amazon connected with trip-planning app TripIt? Amazon might then know that a certain customer going on business trips tends to buy books about management, but the same person traveling internationally tends to buy at least five novels, plus cold medication and travel-sized toiletries. It could populate a shopping cart two weeks before the trip that one could fine-tune and before making a purchase within seconds.
Will head or tail terms be impacted more? Predictive recommendations could potentially preempt the need for a lot of the more vague and broad "head terms" in search such as "new car," "mortgage," "flowers," or "hotels" if technology can predict when people will be in the market for such products. A lot of the changes should also happen with the infrequent and specific "tail" terms. Instead of typing "one dozen roses free shipping," the data engine would already know that , when this shopper sends flowers, it's usually a dozen roses from a company that doesn't charge shipping fees – so for that customer it's just a matter of recommending flowers at the right time. People will train technology with their behavior, and the technology will train people as the relevance of its recommendations increases.
Search is too ingrained as part of user behavior to disappear overnight. But recall that in 1990, the World Wide Web didn't exist and barely anyone was familiar with the internet. In 2000, browsing directories was still a common way to navigate the internet, and Google had just barely transitioned from a research project to a company. By 2010, search became enmeshed in how people consume media and make purchases. There's no reason to assume that 2020 will be just like 2020, and the safer bet is that it will be quite different. Google seems to be preparing for such a scenario. Marketers looking to make long-term plans should do the same.