With the proliferation of alerts, targeted push notifications and mobile-only deals, it seems that smartphones and tablets are ending consumer predictability.

Human behavior is as unpredictable as it has always been. The only difference now is that there is an overwhelming amount of data about this behavior, and because it is siloed in different places, it is more difficult to build a customer lifecycle or extract insights around attribution.

?I don't think mobile will end predicable behavior per se, but it certainly has a destabilizing effect on brand marketers,? said Cezary Pietrzak, director of marketing at Appboy, New York.

?First, it allows for brand interactions to happen at any time and any place, so understanding the context of the interaction is more important than ever before,? he said. ?Second, a breed of devices complicate marketing efforts because they introduce more channels to the marketing mix and require a specific type of expertise.

"Lastly, as smartphones and tablets become more mainstream and attract a less tech-savvy demographic, we can expect our assumptions about their use cases to change. So, rather than make definitive statements about what customer behavior is and what it may be, it helps to adopt a more liquid approach ? in Coca-Cola's words ? as to what is happening at any moment."

Consumer behavior
Nowadays, consumers have very short attention spans. And, with mobile and other emerging technology, they are easily distracted, which is a big contribution to the ongoing problem.

To help deal with this, it is important that marketers segment their customers and understand that they are different and to group them into behavioral segments.

This allows for more tailored communications that will elicit a higher response.

Furthermore, marketers should use multiple channels.

The more channels companies have in their arsenal, the higher the likelihood that they will connect with consumers.

?Don't hang your hat on push notifications, because if you abuse your permission to send unsolicited notifications to people's mobile devices, they will shut you off. Look to other channels like in-app messages or email,? Mr. Pietrzak said.

?Understand attribution,? he said. ?Develop a plan to estimate attribution to see where your customers are coming from.

?But even if a channel doesn't directly link to a purchase, it may be important for other parts of the funnel. This is often the case with social.?

Many analytics and marketing tools in mobile are currently disaggregated or feature-specific, which frustrates marketers because it forces them to bring these different pieces together in-house with apples and oranges comparisons across data.

These tools will continue to converge into powerful platforms that provide them with a big-picture view of mobile behaviors and also provide a stronger link to the mobile Web and desktop Web.

Additionally, it is all about analytics and data. With these, marketers are able to look at different behavior and help in the end of predictability.

?Mobile analytics platforms will start closing the loop on attribution across the customer lifecycle, which will give marketers a much better understanding of ROI and LTV in the mobile space,? Mr. Pietrzak said.

?As marketers get more access to personal data on mobile, people are pushing back and demanding more transparency as to how this data is used,? he said. ?Apple's shift away from UDID and states' enforcement of COPA are examples of this change in mentality.?

Infinite data
Mobile provides a nearly infinite amount of data to mine to measure predictability behavior in several entirely different ways, per Andrew Walker, Atlantic Territory practice lead at Acquity Group. 

Consumers use mobile devices for directions, to check-in, to find a recipe, shop online and even hail a cab. 

The massive amount of information provided by just day-to-day tasks allows for algorithms to predict behaviors.

?Brands can use data to predict customer?s purchase trends and better customize options,? Mr. Walker said. ?This data can be used to create algorithms to predict behaviors and personalize the process for customers.  

?Some retailers are already putting these practices in play to create predictive wardrobes based on customers? purchase history,? he said. ?By using mobile to store customer histories, brands can to successfully help customers create outfit options for special events based on their own personal styles, as well as help them predict other styles and brands they may be interested in. 

?This improves the ability to cross-sell on the brand side, while improving the process for consumers. Mobile also allows for ?line-busting techniques where customers can check out on-the-spot with a mobile device, instead of waiting in line at the register. For example, if you?re in the shoe department, there is a customer representative present to check you out on their mobile phone or tablet so you don?t have to go up front and wait in line.?

Mobile a cause?
According to Vanessa Horwell is chief visibility officer of ThinkInk, Miami Beach, FL, mobile is enhancing predictability.

?That?s true for marketers as much as it?s true for mobile device users,? Ms. Horwell said. ?Retailers increasingly predict our spending habits, purchasing locations, how often we buy and what we buy.

?Hoteliers know the last time you or I booked a trip and where we stayed,? she said. ?The same goes for airlines. And the ads that populate our Google search pages are tailored to our recent searches.

?Similarly, thanks to location-based social media, sites like foursquare let friends known our location in real-time. And there are a host of apps ? free and paid ? that let users track other friends? phones. What this amounts to, in the aggregate, is the harnessing of data being generated across mobile devices to predict consumer behaviors with a lot more clarity and precision.?

Ms. Horwell does not believe that mobile is destroying predictability modeling, but rather that mobile devices collect a multitude of metrics in real-time and disseminate that data instantly to multiple interested parties.

While mobile devices may be aiding predictability, it is important not to rely exclusively on their prognostication abilities.

Human intuition still counts for a lot.

?A recent New York Times article spoke to the limits of computer intelligence, explaining that for the all the data processors out there and all the computer code being written, a growing number of humans are required to interpret that incoming data; to help search engines better tailor their results to what people are actually searching for,? Ms. Horwell said.

?For marketers, the Holy Grail will be not just capitalizing on emotions through timely and relevant offers, but actually predicting how mobile users feel in the moment,? she said. ?Think of this as reverse engineering current engagement tactics, which is already beginning to happen to some extent.?

Final Take
Rimma Kats is associate editor on Mobile Marketer, New York