Discover PerformanceHP Software's community for IT leaders // April 2014
How Big Data analytics drives success and cuts costs of experimentation
The marketplace has sped up, and as the enterprise responds, data-led experimentation is key to profiting on success, and cutting costly losses, faster.
In the age of the Internet, business ideas succeed and fail at a spectacular pace. Winning products and ideas can emerge overnight, and meticulously planned programs and products can arrive too late, or flop entirely. Agile software development and DevOps are the techniques letting enterprise IT do things faster, but without fast and robust analytics, it may just be faster guesswork—a wasteful series of stabs in the dark.
To make the right decisions at the speed the market demands requires a new level of analytics dexterity. Embracing a culture of data-centric experimentation lets you rapidly iterate through new ideas, finding (and investing in) the ones that succeed—and cutting losses on a dead end earlier than ever.
"An 'everything is an experiment' mentality benefits the business in two ways," says Mike Shaw of HP Software Strategic Marketing, "by shrinking the elapsed time between release of a product or service and starting on the next iteration, and by vastly improving the quality of feedback that drives improvement."
Five stars or die
Whether it’s building software, selling advertising, creating consumer goods, or something else, every business is under pressure to understand what the marketplace wants today—and figure out how those desires will change over time. Traditional ways of measuring consumer whims—focus groups and crude, isolated metrics—aren’t nearly agile enough.
"The Internet creates near-perfect competition," Shaw says. "The best ideas rise to the top—quickly. Conversely, inferior products 'get dissed' in public. The mobile app world calls it 'five stars or die': if you don’t have four or five stars in the app stores, you’re dead."
Ready to experiment
Today’s enterprise has more data about a consumer’s behavior and background, as well as typical behaviors of similar users. Transaction records, browsing history, social media, and calls to the help desk provide a lot more information to help a business put the exact right offer in front of a consumer at just the right time. That goes for basic A/B testing of a website experience to identifying trends to justify whole new offerings.
But few companies have made the process and cultural adjustments necessary to compete in this hyper-data-centric world. Many companies still select investment targets based on months-old data and fail to use dynamic sources of information to update developing ideas along the road to market.
To create the rapid feedback loop that lets you launch better ideas, faster, you’ll need a Big Data analytics effort capable of rapidly drawing insight from high volumes of data. Two critical capabilities are:
- Measure the changing pulse of micro-transactional data. The data points we can capture keep getting more precise. When you use an application, for example, you create “touch streams” of data about how you interact with that app. The result is a rich tableau of information about your best—and worst—product features.
- Derive meaning from human interactions. The ability to use software to derive sentiment from non-relational data sources, such as voicemails, emails, pictures, and tweets, means we can know much more about customer preferences and behaviors at a very low cost.
From these capabilities, organizations can create a cycle of curiosity that begins with a hypothesis, challenges it with A/B testing, implements findings, and measures results. But even the most capable analytics systems will be of limited value if the data science team doesn’t have the proper mindset.
Need to know
"There’s a level of intellectual curiosity that just needs to be there—a willingness to be wrong and to realize there are no guarantees," says HP Vertica Vice President Chris Selland. "The best data scientists are sometimes the people who have the least certainty about what's going to work, but the most curiosity and willingness to just try it out and see what happens."
Once curiosity has taken hold in the organization, everything can be seen as an experiment. Software, marketing, manufacturing, media, and even IT and business processes all have complex components that can be optimized quickly using an iterative process of Big Data analysis.
Welcome to a new reality of split-second decisions and marketing by the numbers.
Looking toward the era when everyone — and everything — is connected.
Introduction to Enterprise 20/20
What will a successful enterprise look like in the future?
Challenges and opportunities for the CIO of the future.
Dev Center 20/20
How will we organize development centers for the apps that will power our enterprises?
IT Operations 20/20
How can you achieve the data center of the future?
What the workforce of 2020 can expect from IT, and what IT can expect from the workforce.
Preparing today for tomorrow’s threats.
Data Center 20/20
The innovation and revenue engine of the enterprise.