Data done right
learning to target the right type of data can uncover insights that drive project success
BY STEVE HENDERSHOT
Big data is a big deal.
More than 75 percent of organizations are planning big-data investments in the next two years, according to a 2015 Gartner report. Yet, only 37 percent of organizations have deployed big-data projects in the past year, suggesting that filling the gap between data and insight remains a substantial hurdle.
Nearly 50 percent of organizations that use big data say it has been more expensive than they expected, according to a 2015 survey by Vanson Bourne. Likewise, budget limitation is the most common reason that organizations don't use big data, the survey found. But a “go-big-or-go-home” philosophy on data might not be in an organization's financial interests, says Hannu Verkasalo, PhD, CEO and founder, Verto Analytics, an Espoo, Finland-based firm that measures consumer engagement with brands online.
“Companies large and small struggle to use big data for business benefit,” he says. “Even those organizations with vast amounts of data available are not finding it easy to aggregate, analyze, quantify, populate or derive metrics that provide actionable insights for decision-making.”
Rather than casting a wide net to blindly mine data, some organizations are taking a more focused approach to capture a greater ROI. By using targeted data to streamline project portfolios, organizations are finding that so-called small data can make a big impact, Mr. Verkasalo says. For instance, organizations often miss the small, targeted data that helps determine how customers actually use their products, he says.
“Companies large and small struggle to use big data for business benefit. Even those organizations with vast amounts of data available are not finding it easy to aggregate, analyze, quantify, populate or derive metrics that provide actionable insights for decision-making.”
—Hannu Verkasalo, PhD, Verto Analytics, Espoo, Finland
Nearly 50% of organizations that use big data say it has been more expensive than they expected.
Source: Vanson Bourne
“How many times did he or she log in to the service in one day? What device were they using? For how long? All these bits of small data can add up to a much larger picture that, in turn, can inform better product and business decisions,” he says.
By relying on this targeted data, organizations that have limited resources still can realize the benefits of analytics without committing to a major investment. This approach lets them sidestep big data's potential to overwhelm, distract and drain budgets and resources, Mr. Verkasalo says.
For instance, IssueLab is a New York, New York, USA-based organization that collects case studies, white papers and other research generated by foundations around the world. In 2015, the organization ran a targeted one-year data project that cost roughly US$200,000 to make all the materials in its 18,000-document catalog more searchable. The project leveraged a software program that relies on titles, topics keywords and secondary themes to better organize all materials. The result was a search engine that is both better at locating relevant content and at filtering out off-topic research.
“We had a mess—an enormous and overwhelming hairball of information,” says Gabi Fitz, director of knowledge management initiatives for IssueLab's New York-based parent organization the Foundation Center. “We recognized that if users are overwhelmed by the amount of information that's presented, then it's a mistake to say they have genuine access to that information.”
While big data requires organizations to sift through large buckets of information, capturing and using targeted data means organizations can focus on only the amount and type of data that's relevant to a specific project. For instance, organizations with projects impacted by weather could narrow the climate data they gather to a particular postal code rather than a city or region to best define risk, says Hyoun Park, a former project manager and portfolio manager who now is a big data analyst and chief research officer at Blue Hill Research, Boston, Massachusetts, USA.
Motivation for Metrics Goals that drive data projects for organizations:
Source: Gartner, 2015
“[Targeted] data is not a naive approach; it can be very sophisticated when aligned with key drivers and risks. In theory, this will allow us to conduct projects faster, easier and cheaper.”
—Hyoun Park, Blue Hill Research, Boston, Massachusetts, USA
“[Targeted] data is not a naive approach; it can be very sophisticated when aligned with key drivers and risks,” Mr. Park says. “In theory, this will allow us to conduct projects faster, easier and cheaper.”
While small data can offer big business benefits, gathering targeted data requires careful risk management. For instance, a project to collect data from members’ activity trackers run by CSS, a health insurance company based in Lucerne, Switzerland, had to get approval from an ethics commission and meet the highest data-protection standards before it could proceed.
The company's goal was to offer discounts on premiums to customers who wore activity trackers—if the targeted data mined from those devices (such as average number of steps taken per month) indicated that the participants led healthful lifestyles. The six-month MyStep pilot project cost CHF60,000 and was a joint venture with Zurich, Switzerland-based technology researchers Health-IS Lab. But before the project launched in July 2015, the team had to manage security and access risks, says Niklas Elser, project sponsor, CSS, Lucerne, Switzerland.
For starters, the companies that make the fitness trackers also harvest and store their users’ data, so CSS had to make sure their access to that data wasn't interrupted. Although manufacturers had the right to change interface programs that could block data transmission to CSS—and, therefore block access to data that CSS needed for MyStep—the organization engaged with legal departments for each device manufacturer. Through those meetings, CSS received assurances that the companies wouldn't alter CSS access—although no companies would guarantee no changes. “We considered the risk to be small and bore it,” Mr. Elser says.
The pilot project was so successful that CSS will offer the incentive to all customers starting in the second quarter of 2016, he says. “We want to take these experiences to the next level, designing personalized digital programs to address chronic diseases that really work, and have that lead to a win-win for both the patient and health insurance.”
To deliver the greatest ROI, data projects must support a specific business objective. For the U.S. pizza chain Papa Murphy's, the goal was to streamline the construction of its restaurants.
In April 2015, the company started using an application that collects a wide range of task-completion data that can be custom-sorted to show only the information that each project team member or stakeholder needs. For instance, the point-of-sale installer sees only the progress of other subcontractors working on the store that's relevant—such as when the electrical system and countertops are in place—so that team knows when its service is needed. But Papa Murphy's executives get a big data view of on-site progress for all locations under construction.
“In order to be useful, data needs to be contextualized and cleansed and brought into the context where each stakeholder lives.”
Contractors have full access to data but can filter their view to “get a narrow view that doesn't overwhelm them with more than they need to see,” says Jennifer Doyle, director of information and technical services at Papa Murphy's, Vancouver, Washington, USA. Project leaders and executives are “able to pull out data and do higher-lever analytics to see things like whether one architectural firm is working faster than others,” she says.
This approach makes the store-opening process more efficient, and she predicts restaurants will open a few weeks sooner because of the little data project approach. In 2016, Papa Murphy's also will start to analyze how cost estimates and invoice tracking via the app can help the organizations identify opportunities to reduce costs, she says.
Executives keen on data-driven decision-making likely will continue to invest in big data projects, but project and portfolio leaders must learn to capture and prioritize insights to deliver real results, Mr. Park says.
“There's a temptation to believe more data is better, but the truth is, there's a limit to it. In order to be useful, data needs to be contextualized and cleansed and brought into the context where each stakeholder lives.” PM
PM NETWORK MARCH 2016 WWW.PMI.ORG
MARCH 2016 PM NETWORK