How Big Data Yields Better Business Results

Every week, it seems, there is a new article touted by The New York Times or The Wall Street Journal acclaiming the use of big data within a well-known small business or startup that has consequently jettisoned its way to becoming a household name. 

Uber used big data on mobile to cater to those who needed rides. Netflix used big data to create an insanely successful TV series. AirBnB used big data to put a dent in the hotel industry. YPlan used big data to find the right concerts for the right customers at the right time.

Everywhere you look, startups are using big data to become something more than a six person team in a rented out space. And yet for all the success, all the articles, all the accolades, few big businesses are picking up on big data’s big benefits. And that slow-to-change mindset has got to pick up pace. 

Big Data Best Practices, or Lack Thereof

Sure, the bigger the business the bigger the opportunities, or visibility, for failure. After all, Google Flu Trends was mocked, torn apart and thrown to the side as an infamous example of big data gone haywire. Amazon recommendations often get the same treatment from customers who don’t find much variety within what the Amazon bots have to offer. Facebook also gets lashed with wrath as critics accuse the social media site of selling out their user base as mere data points with which to increase profit.

Just because data is becoming the driving force behind business strategies and policies doesn’t mean that we’re losing our human touch.

Certainly, when a business behemoth makes a big data move, the world watches and waits for the wreck. But the failures found within a heavy data-reliance are no reason to back away from the industry – especially when the little guys are figuring out how to make big data work for them in a very real, very big way. 

There must be a big data secret then, no? There must be something the little guys have understood that is missed entirely by the names you’ve known for so long. And the more you look into the big data uses among the smaller companies, the more you realize that a data-driven business is most often successful when following a few key best practices.

These best practices were or continue to be missed by the Googles and Facebooks of the world. These best practices are often overlooked by the overarching business strategies produced at large scale for large scale implementation. And here are these best practices, for you to read over, chew on and fully digest to help your company achieve data-driven success.

Big Data Must Be Small

Specialization matters when it comes to big data – because without it, you’re just dealing with a massive amount of 1s and 0s that relate to each other in ways even batch computing would have a difficult time understanding. 

Uber focused on public transportation needs – and then further specialized by zeroing in on the metros that would most benefit: San Francisco, New York and London. YPlan followed a similar route, focusing exclusively on concerts and launching in cities where last-minute plans can easily be attended (AKA, you’re often on foot): New York and London. 

Whereas Google Flu Trends was specific to the flu, the data it was analyzing was amorphous, relying a bit too much on keywords, rather than allowing a user to simply say, “Yes, I have the flu” or “Yes, I have these symptoms.” By following a set of keywords rather than specific user-input, as Uber and YPlan’s apps allow a user to do, Google Flu Trends misinterpreted data, overestimating flu-infected areas and losing public trust in data-driven predictive analyses. 

Democratize the Data – For the People

Data is merely human-generated information, and the best, most accurate human-generated information is that which comes directly from the horse’s mouth. Where Uber, AirBnB, YPlan and Netflix all do well is in asking their user-base to input information. Want to rate the movie or TV show you just watched? Give us feedback on the vacation home you rented. 

What type of music do you like? In which type of vehicle do you prefer to travel? 

These are data points that could be guessed, could perhaps be part of a predictive audience analysis about a particular demographic – but startups don’t use them in that way. Amazon and Facebook often do. This is why if you or a group of your friends like a particular Facebook page, you’ll see ads about similar products, because predictive analysis has you in a conversion look-alike funnel. It’s why on Amazon, when you put a particular book in your cart, the options for what you’d also like are near identical to what you just indicated you are considering buying. 

Data is people-based, and our uses of it must be as well.

It’s about a conversion funnel. 

For startups, though, its about building immense loyalty and thus a word-of-mouth fanbase. To do that, you treat people as the owners of their data, allowing them to share on their terms, rather than on the terms of their digital behavior. 

Embrace Skepticism 

When you’re fighting for the betterment of your users, embrace skepticism when and where you can – it is what will propel you forward. Uber is consistently questioned and even banned from cities where cabs and taxis have monopolies. But that doesn’t stop them from going where the data points, and then petitioning the local laws. 

Netflix’s implementation of weather data to help determine which TV shows or movies to recommend was questioned – and turned out to be one of the better data use experiences for the site’s customers. “Viewing behavior is the most important data we have,” said Carlos Gomez-Uribe, VP of product innovation and personalization algorithms at Netflix. 

AirBnB turned what was once a creepy idea (you know, staying at a stranger’s place), into something that is cheap, fun and profitable.

Embrace skepticism, but follow the data – as long as a better user experience is the end goal.

Embrace skepticism, but follow the data – as long as a better user experience is the end goal. Even Facebook is catching on to that fact with their anonymous login and the rolling out of policies that won’t allow brands to take identity data points from Facebook user’s without proper usage. 

In the end, data-driven success is about smart data usage and treating your users like people. After all, just because data is becoming the driving force behind many of our business strategies and policies doesn’t mean that we’re losing our human touch. Data is people-based, and our uses of it must be as well.