“Faster-than-real-time” was the buzzword used to promote the first edition of LeWeb’s summer edition in London, an event that brought together some of the most influential actors in the international internet ecosystem. In this case, the meme had well-defined boundaries: trying to understand how companies stayed ahead of the curve, navigating through the rapidly changing technological environment. But as a more generalized term, “faster-than-real-time” is also the Holy Grail of innovation. Being able to predict what people will want tomorrow is an incredibly powerful asset that can “pull” rather than “push” research and development forward.
And companies are better today at predicting consumer behavior than at any moment in time. How do they do it? Big data.
Nothing here is really new. Recommendation engines are already a boon for companies like Amazon and Netflix (75% of rentals) and a source of angst for writers like Eli Pariser, who in his 2011 book The Filter Bubble highlights the dangers in letting algorithms run our lives. An eerie passage even looks at prediction capabilities of networks we use every day, like Linkedin, that supply enough data to let employers not only know how good you are for the job, but how long you will stay at the company. Your profile could therefore be judged not on past employment history, but on future employment history of people with similar profiles. In a way, Philip K. Dick wrote a great novel about this.
Such is the power of big data that private companies like UBS Investment Research use Cold War-style satellite surveillance to spy on Walmart parking lots to predict the company’s quarterly results. And advise hedge funds. Of course, these satellite images are set in time, but real-time data, using a host of other sources are already available and used for prediction models. How many will go shopping tomorrow? And more importantly, what will they want to buy? It took time for marketing departments to “discover’” the weather as one of the most important parameters for influencing consumer habits. But what is the “weather” of innovation?
The acceptance that some things are better left to others is at the heart of any company with a sustainable growth rate.
At a conference last year, the R&D director of one of the largest European car manufacturers explained that one of the major obstacles for his industry was closing the time-to-market gap. For him, the record in the car industry belongs to Toyota, which is able to move a car from blueprint to the road in about four years. His was double, “how can we predict what type of car consumers will want to buy in 2020?” That could be a marketing question, and having consumers draw up the plans for the products they buy is already a reality for many brands, from Starbucks to, say, Boeing. But stepping up the innovation game would translate in accelerating both the ideation and manufacturing processes to close the time-to-market gap à la Zara (2 weeks from drawing to store). Of course, a car is more complex than a suit.
Companies like Apple—and its relentless product launch schedule—can teach us a thing or two about this. Though it has been argued that Apple is the antithesis of Open Innovation, how can a company with such a limited patent portfolio (in numbers, not in reach) produce complex electronic equipment without calling onto others, via co-development, acquisitions, and collaborations? The acceptance that some things are better left to others is at the heart of any company with a sustainable growth rate. Globalization, a narrowing market window, and the quickening pace of technology have pushed companies to devise better and faster ways to innovate. Problem-solving platforms and big data analysis are the drivers of this acceleration.
Coupling this technology with classic problem-solving platforms can turn any relatively small and “siloed” R&D department into an international research powerhouse.
The possibilities of coupling big data analysis with research are endless, and will profoundly affect how discoveries are made. Companies like brainSCANr (Brain Systems, Connections, Associations and Network Relationships) analyze 3.5 million scientific articles on neuroscience to draw unexpected correlations between various parts of the brain. Real-time analysis of data also predicts trends, like Project Heathmap, which after the 2010 Haiti earthquake anticipated the spread of cholera by correlating online conversations with official statements. Yet scientists haven’t waited for big data to become a buzzword. Computer models of everything from species extinctions to the climate and ocean acidification are their most powerful allies. What’s new is the amount of real-time data that can be processed in a relatively short time, and for little cost. Scientists are working hard to develop new and better algorithms to extract not only statistical correlations, but parallel network of concepts, establishing strong links and, more importantly, weak links between areas of research. Coupling this technology with classic problem-solving platforms can turn any relatively small and “siloed” R&D department into an international research powerhouse.
Problem-solving platforms play a key role in accelerating innovation, not only by finding relevant experts outside the field of the problem, but by letting companies post several problems at once to multiply the number of possible avenues for their research, establishing different scenarios, or alternative timelines and even alternative ways a product could be manufactured—in real time. As soon as a problem is solved on one of these timelines, the others are sometimes abandoned and another problem map is established stemming from the new “node.” Looking at a maze from above, it would be like filling all the paths at the same time, meaning less costly dead-ends and zero turn-arounds to find one, or many exits.
Ask any weatherman and they will tell you that predicting the future is far from easy. If innovation has a speed limit, new methods are quickly catching up to it. And sometimes along the process, you’re able to break it. New algorithms, for example, are able to identify experts who do not yet know they can solve a given problem. The Web’s vast expanse of information, coupled with increasingly powerful collaboration tools, spell a new dawn for research that companies will have to adopt and adapt to their existing process. Because faster-than-real-time, let’s face it, is the future.
By Saman Musacchio
Saman Musacchio is VP of communications for Hypios, an open-problem solving platform based in Paris. He holds an MA from the University of Alberta (Canada) and contributes to various publications including CNRS International Magazine, Business Insider or Bloomberg. Twitter: @musacchios.