You’ve heard the hype: Companies that use big data analytics are 5 times more likely to make decisions “much faster” than their competition. Great. And you’ve built yourself a data science team and started collecting data. Even better. But, now you want to start innovating, pushing the envelope, become the next Google.
Well, you’re probably not going to be the next Google – not unless you have an extra $8 billion lying around. But, just because you don’t have their budget, doesn’t mean you can’t learn from the what the great data innovators are doing.
Facebook basically has free reign to stick whatever it thinks you’ll love into the News Feed, and recently it’s been focusing on video. Every video Facebook shows you is essentially a recommendation from Facebook. Using the data they collect, Facebook can crunch insanely specific information on viewership behavior to learn what you want to see.
They measure what do you watch on silent, what you click on for sound, what you watch in HD not to mention what you like, comment on and share. Based on this data Facebook knows that a video popular with some people will probably be interesting to people similar to them based on all the biographical and other behavior data it has. Meaning they can give you a data-driven personalized news feed of the exact videos you want to see. Even if you didn’t know you wanted to see them! Which makes a site like YouTube seem a lot less useful.
The lesson Facebook teaches us is that things like mass marketing are dead. Every person is unique – yes your mother was right – and therefore every person should be treated as an individual. With enough data it is possible to do that at scale using algorithms to create material or suggestions that are hyper-targeted to each individual consumer.
Amazon has taken personalization beyond marketing to people and started using it to inform interactions with people. For example, if you call up Amazon’s customer support, you will be met with a person who has all your information in front of them. Which they can use to help you sort out your issue. No more explaining yourself 500 times to 10 different people because Amazon’s customer team already knows.
In general, customer service is still treated as an expense to be minimized, not an opportunity to be developed. But – as Amazon teaches us – with good data you can treat each customer as an individual and turn what is usually a highly unpleasant experience into an opportunity to promote your brand and gain customer loyalty. The best part is that good data support doesn’t require a vastly expanded workforce, or even a new type of employee—these are conversations that people already know how to have.
Next Big Sound
The music industry generates an extraordinary amount of data, especially now that the consumption of music has moved away from simply purchasing an album, to consuming that music through streaming sites like Spotify and Pandora. Next Big Sound is collecting and analysing data from Spotify streams, iTunes sales, SoundCloud plays, Facebook likes, Wikipedia page views, YouTube hits, Twitter mentions – basically everything – to help predict what the next hit band or artist will be.
Their analytics can explain which bands are about to break, which late-night shows really impact an artist’s trajectory, and more. In addition to helping artists and music execs understand what will help make them successful, they’ve also partnered with companies like Pepsi and American Express to help steer the $1 billion plus that is being spent each year by brands on music-related marketing and sponsorships.
So, even if you aren’t generating much of your own data, you can still benefit from collecting and analysing data from elsewhere. And when you start analysing data from across a multitude of sources, you can start to actually predict the future.
General Electrics’ many machines — they make everything from power plants to locomotives to hospital equipment — now pump out sensory data about how they’re operating. Which GE’s analytics team crunches and uses to make them more efficient.
For example, they’ve been working with the Italian airline, Alitalia, to monitor wing-flap positions and relay adjustments to minimize fuel usage. Through this method, GE managed to save Alitalia an estimated $46 million in just 2 years. And, they estimate that self-reporting machines, such as the engines on enormous airliners, could save the airline industry more than $30 billion over 15 years.
In fact, GE has gone on to speculate that data can boost productivity in the U.S. by 1.5%, which over a 20-year period could save enough cash to raise average national incomes by as much as 30%. The important lesson here is not to pigeonhole your data collection to data produced by people because machine generated data can be just as useful.
It might surprise you to see an old brick and mortar department store on a list of data innovators, but for one of the oldest retailers in the US, Macy’s is remarkably ahead of its time. They’re using the latest Big Data Analytics technology, to change the in-person shopping experience of their customers.
They analyse a large amount of different data points, such as out-of-stock rates, price promotions, sell-through rates etc. and combine these with data from product sales at a certain locations and times as well as customer data in order to optimize their prices in those locations. This takes the practice of differential pricing, which has been around for a long time, to a whole new level.
Using this method they’ve been able to reduce the time it takes to optimize the price of an item from 27 hours to just 1. Which in turn has increased sales 10% in the last few years. Pretty impressive when you consider that just 4 years ago they were relying on Excel for all of their analysis. Proof, I guess, that you can indeed teach an old dog new tricks.
The important lesson here, is that while it may require a heavy investment now, in the long run investing in good data practices can save you a whole lot of money.
There are over 180,000 felons on the loose in the United States that police can’t track, but one innovative startup is using data to help change that. Mark43 is developing software to help the police to collect and process data better. Using their cloud-based records management and analytics system, their goal is to make policing smarter, more efficient, and more accountable.
To do this, they’ve created a web app that makes it easier and faster for cops to enter arrest and incident reports. Thereby reducing a process that can take several hours on decades-old legacy systems to simple 30 minute task. They then then add a layer of algorithmic intelligence, connecting data such as social media activity and phone records (obtained via warrant) to those police reports to give patrol cops a more complete picture of a criminal or a syndicate. Mark43 is already in use by Los Angeles County Gang Taskforce and soon will be officially deployed across the entire major metropolitan department.
The way you collect data is just as important as the way you analyse it. If collecting relevant data takes too long, it starts to lose its usefulness. It is essential that your data be timely so that your insights are about what is happening now, instead of what happened last week. You should be building in data collection systems that are fast, easy to use and sorted immediately into the relevant sectors. You can save a lot of time in the processing phase of data analysis by making sure the data is collected right in the first place. In addition, Mark43 demonstrates the incredible added value of data when it’s combined with other data (like social media) from around the web.
Netflix has proven that data-driven development is the future. One popular example of this is their first hit show House of Cards. After outbidding networks including HBO and ABC for the rights to the show, they were so confident that it fit their predictive model for the “perfect TV show” that they threw away the convention of producing a pilot, and immediately commissioned two seasons comprising of 26 episodes.
Literally every aspect of the production under the control of Netflix was informed by data. They even analyzed the color pallets of successful show covers to come up with the now iconic Kevin Spacey cover.
Using data from day one is key. You should be using data when creating your ideas and products, not just when you’re marketing them.
Uber is the ultimate data-driven company. They use data to inform every part of their business strategy. Their algorithms monitor traffic conditions and journey times in real-time, meaning prices can be adjusted as demand for rides changes, or when traffic conditions mean journeys are likely to take longer which they use to build a predictive model to estimate demand in real time.
In fact they have so much data on you, their latest offering UberPool suggests you share rides with other Uber users who make similar journeys at similar times – reducing traffic in congested areas. In addition, their rating algorithm ensures that bad drivers (and passengers) are pushed out of the system because they will find it hard to get chosen.
But what makes Uber a truly great data company, is that they recognize how valuable their data is internally, but also how much it is worth to others. Recently, Uber launched a service that lets its customers connect their Uber account to their Starwood Prefered Guest account so you can earn hotel points for taking Ubers. In signing up for this service you agree to give Starwood all of your Uber location data – which they can use to retarget you if they see Uber taking you to a non-Starwods property
Uber can run the same program with airlines, restaurants, nightclubs, bars and pretty much every place you could ever choose to go. Every time you go from point A to point B in an Uber, “A”, “B” or both represent a new potential consumer of your data. Uber knows the hottest nightclubs, the best restaurants and now has as much data about traffic patterns as Waze – and all that data is worth a LOT of money.
So the lessons we can learn from Uber are 2-fold. First, data should inform every element of your business. And second, never underestimate the value the data you collect has to people outside your organization.
Updating your data strategy
Is it a coincidence that companies who are strong at innovation are three times more likely to rely on big data analytics and data mining than their counterparts? Or is it the data analytics that is making them more innovative in the first place?
We may never know, but with the average business expected to spend $8 million this year on big data and related projects, it’s clear that companies are racing to take advantage of the insights and savings that data can provide. And if you want to be competitive in your market, you should too.
This post is taken from a talk given at Tech Open Air