Enter your keyword

Big Data is Big Hype, but Data Analytics Delivers

Big Data is Big Hype, but Data Analytics Delivers

Big Data is Big Hype, but Data Analytics Delivers

At Ascent, we have been focused on the growth of data analytics – or as we called it back in 2008, the ”Explosion of Data” – for close to a decade. In the early days we backed companies such as ZoomInfo, Cymfony (acq. by TNS) and Bizo that comb the web, collecting data to build professional profiles or understand customer sentiment on products. More recently we have seen a revolution on the infrastructure side bringing new technologies and databases to market that greatly accelerate the speed at which analytics can take place, including portfolio companies Interactive Supercomputing (acq. by Microsoft), Terascala and ScaleBase. 

Today we are focused on new opportunities to harness data in previously impossible ways that combine the volumes of new data sources with new underlying infrastructure to drive unconventional and deep business insights. Companies such as ClickFox and CargoMetrics are perfect examples. We are well aware of the challenges that exist in extracting real value from the swelling torrent of data, but we also see great potential and have witnessed the power when data, algorithms and technology are combined and harnessed properly. 

While the media has often focused on the concepts and created a lot of hype around “big data,” we felt it was time for a B2B Forum focused on what emerging leaders in the application of data analytics are actually doing today.

The ways most enterprises are applying data solutions are “outdated and appalling,” claimed Simon Moss, CEO of Pneuron, at the opening of Ascent’s recent B2B IT Forum in Cambridge.

And he’s right.

That’s the main reason why 55 percent of “big data” projects fail today. That’s not all. Perceived high costs, security concerns, and a skilled worker shortage are all factors. But the biggest reason data analytics projects fail, in my view? The political, cultural and logistical issues of gaining access to what is usually siloed or isolated enterprise data sources.

So what is working? I asked our forum panelists to each describe an innovative use of data analytics from one of their customers.

Rajeev Sharma,CEO of Video Mining, a company that crunches data on in-store consumer behavior to deliver new insights to package goods manufacturers (think P&G), is bringing the granularity of online analytics to physical stores. As an example, Rajeev said that by collecting data about every shopper who came through a supermarket, Video Mining can help brands like Coke and Pepsi optimize their in-store marketing by showing where most consumer purchasing decisions are made. And interestingly, Rajeev said he has determined the highest conversion rate of a common food item: bread. He said that by quantifying varying methods of bread placement over millions of in-store queries (or purchases) – by company or by bread category (white, wheat, gluten-free, etc.) – they had found that more bread sells when it is categorized by type. This may seem like a trivial insight, but it adds up to tens of millions of dollars annually for brands and retailers fairly quickly.

In another interesting case, Rama Ramakrishnan, CEO of email marketing technology provider CQuotient, said their customers can achieve a 500% increase in marketing effectiveness. For example, he’d found that using the words “angel,” “angelic,” “cherub” or similar descriptors of children’s clothing in subject lines resulted in a 5x increase in email open rates. CQuotient made this breakthrough by studying the language used in product reviews on sites like Amazon. Those “angelic” words bubbled to the top in CQuotient’s data collection, and delivered a very high ROI for the new marketing campaign.

Bill Hawley, COO of Ascent portfolio company ClickFox, said that by looking at data from all customer interaction points, his company can uncover very powerful new insights about the “total customer journey.”  As one example, analyzing cross-channel behavior trends and issues across

all service touchpoints can show if a retail store or agent is in need of training. If they need help on multiple occasions, ClickFox pinpoints the store or employee’s lack of training as potentially negatively impacting store sales. By identifying these kinds of problems early, ClickFox is improving the customer experience and saving major brands more than $100 million annually, Bill said.

Paul Markowitz, Global Insights Consultant at Bain & Company, shared that he has been working on data analytics projects in areas previously unexplored by customers – virtually shining a light into rooms full of data that were previous pitch black. For example, Comcast might send hundreds of mailings a month, but are they working? Previously the company wouldn’t understand which messages resonated with customers, but increasingly a company like Comcast is using data analytics to pinpoint the direct mailings and email messages that a customer wants, and eliminate the spending on messages they don’t.

But data insights can go way beyond targeting sales. Simon has seen it used to combat money laundering within a large financial services customer, where data systems belch huge amounts of false positives that all need to be investigated. Contrasted to the 30 minutes needed by a human investigator, Pneuron was able to crunch the information in 10 seconds, saving the firm a great deal of money and increasing the productivity of its fraud detection group by 95 percent.

These examples provide compelling evidence that strides are being made, but the data analytics sector overall is still immature. While the amount of data being collected is skyrocketing, projected to grow 50x this decade, a CIO Insight survey found that 41% of global corporations are either somewhat or completely incapable of collecting and analyzing data. Still, 63% believe that allowing workers to harness data will help them make better and faster decisions. So how do companies use data better?

Bill suggested being very thoughtful about identifying the specific business problem first, then exploring what data can be used to solve it. Including the data that really matters is more important than using data that is easiest to access. The next step is to have “an operational lens” to ensure a practical and robust process, and involving teams with extensive experience in data mining and analytics is critical to ensure you do not repeat common mistakes.

It is also important to be sensitive to privacy.  The depth of big data analytics can take things to creepy levels. As an example, instead of “We picked these recommendations just for you,” Rama suggested a more masked approach, “Customers like you like this…”

One of the overriding themes of the evening’s discussion was that data analytics is not new, but it is seeing a major resurgence as new technologies come to market that make previously inconceivable analysis relatively simple. Perhaps we should look at “big data” as a “revolution in measurement,” as Rama suggested, rather than a ground-breaking technology. Either way, it was clear from our Forum that while we have a ways to go with applying data analytics broadly across many industries, we are making headway and developing new data-driven solutions to long-standing business problems, and there remains tremendous potential yet to be tapped.

No Comments

Post a Comment

Your email address will not be published.