Is the person who viewed those leather boots on a laptop the same shopper who browsed the pricey footwear from a desktop but ended up buying the product using a tablet? That tricky question is behind the rise of cross-device tracking tools, and it’s the reason why popular “walled-garden” sites like Facebook ( FB ) are gaining in the digital ad game. In the past, cookies were the universal bread crumbs of the Internet, helping advertisers find out which websites a particular user favored. Cookies are tiny text files that let websites recognize users — usually by their IP (Internet protocol) address — when they return to a website, thus learning their preferences. With this data, companies can serve targeted ads — ads that they believe are most likely to lead to a sale. With the transition of Web users to mobile devices and apps, cookies aren’t as effective, and advertisers are looking for new ways to identify users and learn their preferences. “When you understand who a user or person is, you can provide them with more relevant advertising or more personalized information if you are a publisher. That identity is potentially helpful to improve the experience,” Greg Sterling, vice president of strategy and insights at the Local Search Association trade group, told IBD. Determining which exact ad prompted a shopper to buy, a process called attribution, “is the holy grail of advertising right now. It’s what everybody needs to understand,” Anna Bager, senior vice president and general manager of mobile and video at the Interactive Advertising Bureau trade group, told IBD. Learning more about cross-device attribution is high on the must-do list for many ad pros this year, marketing firm Rocket Fuel said in a research report last month. A Rocket Fuel survey found that ad professionals also intend to learn more about data management platforms in 2016. DMPs help businesses get a panoramic view of their customers by collecting marketing information from a variety of channels, such as websites, emails and mobile ads. But first, advertisers want to be certain of just who those shoppers are. “The technology and the different methods to identify people across devices in the aggregate have become more sophisticated,” said Sterling, who calls Facebook the leader in multichannel identification. “You have to sign in to use Facebook, so they will know if you’re on a desktop computer, a laptop or a mobile device,” Sterling said. “They can provide whatever experience they want to provide because they know who you are, and there’s persistent identity there.” But Sterling says that the shift from using just one device to seeking content from a hodgepodge of desktops, laptops, tablets and smartphones has created “blind spots (where) people are not revealed to a Website publisher or a marketer. Somebody may sign in on one site and then go on to use the site on a mobile device and not sign in.” This trend is putting cross-device technology into the spotlight, as digital advertising, including mobile advertising, continues to boom. One trend is programmatic ad buying, where software determines ad buying and placement. It works best with ads targeted to user preferences, no matter the device used. Ad Firm Criteo Predicts User Behavior Market research firm eMarketer estimates that programmatic digital display ad spending in the U.S., which reached $15.43 billion last year, will rise to $21.55 billion this year and $26.78 billion in 2017. Paris-based ad tech firm Criteo ( CRTO ) says its cross-device advertising tools work anywhere online, including on the mobile Web and inside apps. While cookies are a common standard used to identify and follow users in a browsing environment, that information can be “captured through other mechanisms when you are in applications,” Criteo CFO Benoit Fouilland told IBD. “Our technology focuses on predicting the behavior of the user based on shopping intent information. That’s an area where we are developing pretty unique capabilities. “There are not many players in the industry able to develop cross-device capabilities. We are one of those few companies and we are investing significantly in this area of what we call ‘universal matching.’ ” Top social networks Facebook and Twitter ( TWTR ) are expected to emerge as big winners in 2016, as digital display-ad spending, which includes mobile ads, overtakes search-ad spending in the U.S. for the first time, according to separate reports this year from Cowen and eMarketer. Google owner Alphabet ( GOOGL ) is seen as another winner, with its browser and slew of email, search and other services that have users logging in with their Google passwords. Facebook remains a dominant pick of the ad buyers surveyed, Cowen’s report said, followed by Twitter, LinkedIn ( LNKD ), Facebook-owned Instagram and privately held social sites Pinterest and Snapchat. EMarketer predicts that 2016 will be pivotal for Facebook; it’s the first year in which more than half of the U.S. population is expected to use the social network. Facebook will capture 73% of social network ad spending in the U.S., or $9.9 billion, this year, eMarketer said last month. Twitter will be a distant No. 2, getting 14% of social network ad dollars, or $1.9 billion. The research firm also says that nearly 53% of U.S. mobile phone users will log onto Facebook at least once a month this year. Google is expected to remain No. 1 in 2016, with a 33.3% share of mobile ad revenue globally, eMarketer said in a report in March. It says that No. 2 Facebook’s share will reach 17.7% this year. User Tracker Methods: Deterministic And Probabilistic To learn user identities across devices, measurement companies rely on two methods — “deterministic” and “probabilistic.” Deterministic matching is at the core of the formidable cross-device matching success of Facebook, Google, Amazon.com ( AMZN ), eBay ( EBAY ), Twitter and other so-called Walled Garden sites, eMarketer analyst Lauren Fisher said in an online seminar in January. Walled Gardens are websites and apps that require passwords for entry and have a treasure trove of information about users, making it easier to match consumers across multiple devices. So, in a nutshell, trackers can pretty much determine who a user is. In being able to identify users, Walled Gardens are “not 100% percent, but press close,” Fisher said in the seminar. “Think about it — even now retailers are making a practice of requesting email addresses when you check out in a store. “If I have Facebook on my tablet, on my phone, on my laptop at home and on my work computer, then Facebook knows all those devices belong to me because I’m logged in and they’re using that login to tie everything together.” Fisher says eMarketer believes “advertisers are going to continue to flock to the Walled Gardens throughout 2016.” While nearly foolproof, the shielded deterministic method used to match, measure and target customers can’t be used on Web properties outside of the Walled Gardens. That’s where probabilistic identity matching comes in. Companies using the probabilistic method analyze wider, non-proprietary digital tip-offs that people leave behind online when they have not used a sign-in. Such data could include which Web browser version a person uses, their physical location and their go-to content sites through use of cookies and such technology. With these methods, tracking is a more of a probable than a determined fact. “When companies start to collect that information and analyze it over time, they can begin to say with a certain amount of probability that a particular device belongs to a specific individual,” Fisher explained. Some Web publishers rely on both the deterministic and probabilistic methods to cross-check and boost their device matching accuracy, Fisher said. “The bottom line,” said Fisher, “is there’s no right or wrong approach.”