I don’t always go looking for this stuff – well, OK, maybe I do. I guess I’m always looking for answers to the strange and the mysterious. I’m a reporter, it’s in my blood.

So I decide to check around to see if there’s anybody out there using big data analytics to solve some of the world’s unexplained phenomenon.

Like many people, I’ve always wondered about the Loch Ness Monster, known affectionately as Nessie. You know, is it live or is it Memorex?

But most people never really do much of anything to get to the truth about Nessie.

Enter Dr. Charles Paxton, an ecologist and statistician at St Andrews University in Scotland.

For the first time, Paxton has complied a comprehensive database of all the reported sightings of Nessie – about 1,000 reports in the past 80 years – and he has analyzed about 853 of these sightings.

Paxton is conducting a statistical analysis of the report data and analyzing it for clusters or patterns. His goal is to figure out if Nessie sightings can be explained by natural phenomena. For example, have people reported seeing this serpent-like beast at a site where unusual waves frequently form.

“I am looking at the last 80 years’ worth of Loch Ness sightings report data and asking the question: does it form into any natural clusters, in terms of what is seen? Other people have catalogued some of the reports, but I am trying to get everything that has ever been written down, so it is a pretty comprehensive database,” he tells the Herald Scotland News.

Paxton’s statistics also include such factors as the frequency of sightings by year, what month of the year, time of day, and estimated distances.

The statistics also include a colorful list of descriptions used to identify the shape, color, and texture of the creature: black, grey, green, pink, swan-like, horse-like, eel-like, tadpole-like, hedgehog-like, smooth-skinned, slimy, rough, elephant-skinned, shiny, polished-but-rough, etc., according to a review of a talk Paxton gave to an organization called the Edinburgh Skeptics last August.

“Most interesting than the statistics themselves, however, was Paxton’s discussion on how we should treat anecdotal evidence,” according to Skeptic Sheridyn Woodward.

“Though often wrong in its assumptions and interpretations, anecdotal evidence has historically been useful in identifying previously unknown phenomena, thus establishing itself as potential data which should be treated critically,” Woodward notes.

Although Paxton doesn’t believe there’s some exotic beast swimming around at the bottom of Loch Ness, he does believe that the anecdotal “data” is amenable to statistical analysis.

And it’s his opinion that the reported sightings, although biased and not very precise, might still reveal something of scientific interest if these sightings are treated critically.

“It’s a potent reminder to us that, in our quest to expose frauds and falsehoods, we do not allow our own acquired prejudices to dismiss the prerogative for a critical and unbiased treatment of all data being presented,” notes Woodward.

In April, Paxton, who ultimately plans to publish a scientific paper on his research, discussed his findings to date at the Edinburgh International Science Festival to mark the 80th anniversary of the first “modern” sighting of the monster.

Half of the speakers who participated in the conference were skeptics who didn’t believe the monster was a physical, living being, while the other half were believers, according to Gordon Rutter, a co-organizer of the conference.

“There is the romantic in all of us, I suspect – a monster living nowadays relatively close to a major population is just a cool idea – a bit out of the ordinary from our everyday lives,” Rutter notes. “People claim [to have seen it] and we can’t just dismiss the claims out of hand.”

Next Steps:

• Subscribe to our blog to stay up to date on the latest insights and trends in data analytics.

Linda Rosencrance
Spotfire Blogging Team

Print post

The webcast also discusses how Spotfire is used to seamlessly access Salesforce.com data and enhance traditional reporting capabilities.

One of the most common reasons given for higher than anticipated losses in any quarter is that deals in the pipeline fail to close as expected.

Here’s a sample clip from the webcast followed by a post on how data analysis can help companies better manage the sales pipeline.

Big data and analytics are commonly associated with a firm’s marketing efforts. But, data analysis can be crucial to supporting the sales process and providing visibility into a sales pipeline that often can be shrouded because limited information results in decisions fueled by gut feelings instead of being driven by data.

In fact, 71% of the best performing companies have a central repository of financial performance information to ensure that employees across the organization are working from a “single version of the truth,” according to a recent report from Aberdeen research.

In addition, this route allows companies to overcome the problem of silos of information within departments. Aberdeen finds that 50% of the top performing companies have an open exchange of data across business functions.

“Analytical insight in one department can have an impact across multiple client-facing line-of-business,” the report notes. “Dwindling sales figures could lead customer service to reach out and discover unknown issues, or marketing data might prevent salespeople from wasting resources on an already saturated market.”

Additional insights from the report include:

• Fifty-four percent of the top performing companies have a formal definition of progressive sales stages that are used to weigh sales forecasts
• Forty-four percent of companies who say they have formal deal abandonment or deal win/loss analyses in place are the top performing companies in the survey
• The top performing companies are twice as likely as all other companies to measure the effects of deal abandonment, discounting or sales incentives than other companies

But the first step for many companies is identifying the best metrics to put into place to better manage the sales process. And that means companies should focus their data analysis efforts on deal quantity, quality and velocity, according to Paul Vinogradov, vice president with the sales consulting organizations the Alexander Group.

Deal quantity involves the number of leads, number of deals and the value of the pipeline in terms of expected value for a specific period of time, he notes.

“Quality metrics are more difficult to measure but using them adds a critical lens to the effectiveness of the sales process,” Vinogradov says. “Are leads qualified effectively? Are deal sizes and types attractive? Close rate is the classic quality metric. Win/loss analysis can provide valuable insights into strengths and weaknesses of your sales effort and the advantages or shortcomings of your offerings.”

Finally, deal velocity measures the turnover of deals, or how quickly a lead moves through the sales funnel after it appears.

“How much time in both total duration and in actual rep selling hours does closing the deal require? You can gain valuable insights from assessing where deals get hung up in the process and what constraints, both internal and external, may be causing this,” he adds. “The number one shortfall for most sales forces is the reluctance to qualify out deals.”

Next Steps:

• Subscribe to our blog to stay up to date on the latest insights and trends in data analysis and sales.

Print post

Risk concerns continue to be top-of-mind for bankers. A combination of lower asset yields and loosening loan terms for mid-market and large businesses amid an uptick in commercial and industrial lending is increasing risks for banks, according to an article in American Banker.

“As profits on loans narrow and competition for borrowers intensifies, banks tend to underwrite a greater number of loans to companies with looser lending restrictions,” according to American Banker.

The good news is that banks and other financial services companies are recognizing the role that analytics can play in helping to mitigate risks. Banks are increasing their year-over-year investments in analytics more than any other industry, including the insurance and chemicals industries, according to a report by Accenture.

In fact, 73% of banks plan to increase their investments by more than 10%, according to Accenture. The areas where banks are expected to drive the highest amount of investment in risk analytics are data quality and sourcing, systems integration and modeling.

These are certainly areas where banks need to pump up their investments. Too many banks and financial services companies have silos of customer information that are spread out among different channels (mobile, online, branch, ATM) and among lines of business (checking and saving, loans and mortgages, investment accounts, etc.).

By pooling information together across these various functions and channels and applying analytics against a broader swath of information, account managers and product leaders can obtain a more complete view of customer accounts, including personal and commercial accounts, to better assess the risks associated with individual customer activities and portfolios.

There are numerous ways that banks of all types can apply analytics to better mitigate and manage risk. For instance, investment banks, asset management firms, and hedge funds can use risk analytics to adjust their operating principles to align with market and investment shifts.

Banks can also use predictive analytics for risk management.

“There is no exact science for measuring risk. But with analytics, you can build measurement parameters that can help you establish and examine likely risk scenarios,” according to Deloitte. “From there, it’s easier to understand the potential impact of a risk – and start planning around it.”

Next Steps:

• Subscribe to our blog to stay up to date on the latest insights and trends in analytics and the financial services industry.

Print post

Companies looking to reap the benefits of Mother Nature’s renewable energy sources can use analytics to evaluate multiple factors, including the average amount of cloud cover, direct sunlight, and the wind energy potential that’s available to justify investments in renewable energy systems.

For instance, even in sunlight-rich regions such as the Southwest US or the Middle East, executives would want to evaluate the estimated amount of daily cloud cover since any shading that might occur on a given day could dramatically reduce the output of a solar energy system.

Because data centers consume so much energy, a growing number of organizations are exploring the potential for placing photovoltaic cells on the roofs of their data centers to help power these facilities. Electricity used by data centers worldwide rose by 56% between 2005 and 2010, according to The Data Center Journal.

For organizations exploring the potential use of solar energy systems, big data analytics can be used to evaluate a number of requirements for adoption. This includes the available roofing space that can be used, the remaining lifetime of the roofing infrastructure and how quickly it would need to be replaced, along with the pitch of the roof and the amount of shade received by the roof’s surface.

Meanwhile, recent studies suggest that the wind offers the potential to power the entire world – and then some. In fact, wind has the potential to generate more than 20 times the amount of power that the world currently consumes, according to a study by the Washington, DC-based Carnegie Institute for Science,

The potential behind wind energy has led US government officials to push to expand the development of wind farms on the Atlantic coast. Roughly 1,900 square miles of Atlantic Ocean waters are being made available for building offshore wind farms, according to US officials.

One of the strengths of offshore wind farms is that there’s nearly always a steady flow of wind coming ashore to turn the turbines. However, such investments still need to be cost-justified by investors, many of whom are incented by federal tax credits.

In addition to construction costs, developers also have to consider the costs of conducting environmental impact studies. And the biggest challenge is determining who is going to buy the power, notes Brian O’Hara, president of the North Carolina Offshore Wind Coalition.

Power contractors can use analytics to better determine the full amount of construction costs, including estimates involving overtime labor costs and weather-related construction delays, along with the likely market rate for the electricity to be generated and the amount of power that’s expected to be consumed in a particular region.

Next Steps:

• Subscribe to our blog to stay up to date on the latest insights and trends in analytics, and the energy industry.

Print post

But King James’ less-than-stellar performance on the basketball court happened before he took a hard look at the analytics behind his play, notes Michael Schrage, a research fellow at MIT Sloan School’s Center for Digital Business, in a post in Harvard Business Review.

“Nothing makes serious competitors more open to analytics than losing,” the HBR article notes. “A basketball genius frustrated with his professional failings decided he wasn’t as good or as smart as he needed to be. James took a good hard look at the analytics … then he hired retired NBA legend [Hakeem] Olajuwon … to help remedy the analytically undeniable flaws and shortcomings of his game. [James] explicitly linked analytics to his personal [and] professional transformation.”

But in the world of sports, analytics are playing an even more critical role than helping players improve weaknesses: increasingly predictive analytics are being used to predict future performance. So, no matter how well a player performs in this year’s playoffs, decisions for next year will be made by predicting future results.

That’s according to another HBR post by Schrage, who posits the same approach can be used by Procter & Gamble or other companies to predict which marketing executive will come up with the best idea next quarter or which sales team will make its sales goals next year.

“Future potential matters (much) more than past performance,” according to Schrage. “That’s the new quantitative consensus reshaping professional sports worldwide. After looking hard at the numbers and algorithms, the smartest — and richest — general managers and franchises have made up their collective minds: They’re not paying a premium for yesterday.”

This next-generation “Moneyball” is more commonly being used to predict – taking into factors like age – which players are about to peak and which have past their prime.

“If you’re running Procter & Gamble, Unilever, Google, Exxon Mobil, or Ford, you have comparable concerns about making sure you’re getting the best possible returns from your talent and human capital investments,” according to Schrage.

“You should be concerned about the aging curves of your marketing people; you should want to know if your tech support folks will deliver better outcomes tomorrow than today; you should be predicting which sales teams will procure the most lucrative contracts with the minimum risks,” he adds. “Think of it as Six Sigma predictive analytics for talent.”

Still, it’s difficult to measure the role that personal loyalty plays in inspiring extra effort that leads to better results, the post notes.

“The classic response, of course, is to insist that, ultimately, these decisions come down to human judgments rather than computational dictatorships,” the post concludes. “But that’s exactly why acknowledging the current trend is so important: The leaderships of the richest and (ostensibly) best-managed franchises in sports have effectively declared that the costs of preserving past values are too high relative to the potential for tomorrow’s performance.”

Next Steps:

• Subscribe to our blog to stay up to date on the latest insights and trends in data analysis and predictive analytics.

Print post

Executives across 30 communications and media companies say they plan to invest in network analytic tools to improve network planning, enhance the quality of service, and improve the customer experience, according to the survey.

Carriers commonly use network analytics tools to analyze network data alarms, performance measures, trouble tickets, and customer churn due to dropped calls and other network performance and reliability issues.

Seventy-seven percent of the respondents say that managing the quality of services is the biggest challenge for telecom operators, particularly in a shared network environment. The use of analytics software can help carriers better prepare for and respond to these challenges in a number of ways.

As consumer demand for mobile data continues to surge, carriers will increasingly rely on big data analytics to determine gaps in network performance and capacity.

However, even though carriers are investing heavily in 4G LTE (long-term evolution) communication services and radio network upgrades, operators may not be planning sufficient investments in backhaul to meet consumer demand over the next five years, according to a study conducted by Tellabs.

Global mobile data traffic has increased 13 times in the past five years and it’s anticipated to grow five or six times more by 2017, according to the study, which also predicts a $9.2 billion global backhaul funding gap with a 16 petabyte shortfall in backhaul capacity by 2017.

Individual carriers and those that are partnered in a shared network environment can use analytics to evaluate the likely growth in data traffic for specific customer segments and geographic areas. Carriers can then use that information to craft strategies for building out network capacity in different locations.

Carriers can also use analytics to help create virtual network models by combining network technology configuration and topology with the behavior of network elements to more accurately predict utilization for dimensioning and to predict resource shortages, notes GSMA, an organization that represents the interests of mobile operators worldwide.

As carriers build out network capacity, they can also use analytics to identify new business opportunities and services they can offer consumers and subscribers.

For instance, in a report that focuses on opportunities for mobile operators to deliver “smart” mobile services, Accenture rattles off a number of different applications and services carriers can develop in cities around the world including:

• Mobile health remote monitoring solutions that can be used by chronically ill patients to optimize home medical visits around specific medical needs
• Smart traffic control through the use of embedded sensors in roadway equipment that can be used to proactively reroute traffic to avoid congestion and direct traffic to underutilized roads
• In-vehicle telematics applications that can be used to help stoke eVehicle adoption by alerting drivers to nearby charging stations, wait times and availability

Next Steps:

• Subscribe to our blog to stay up to date on the latest insights and trends in data analytics and telecommunications.

Print post

Whether a customer is browsing a company’s product pages to gather information about a product or she engages in a chat discussion with an agent about a service issue, companies can use this information with analytics to gain a richer understanding of each customer and to better tailor offers, messaging, services, and products for them.

Companies can make clever use of historical transaction data and recent behavioral information to provide customers with proactive communications that are personalized and relevant, as Don Keane notes in a recent blog post for Business2Community.

Amazon is famous for doing this. Let’s say an Amazon customer places an order for a Stephen King book he has previously purchased from the e-tailer. Drawing off the customer’s purchasing history, Amazon detects that fact and then sends the customer an alert, asking if he really means to purchase the same book.

By looking out for their customers’ best interests, Amazon and other companies that use analytics to provide customers with personalized, proactive communications stand to strengthen their customers’ trust and reinforce their loyalty.

There are numerous ways that companies can use customer data and data analysis to provide customers with relevant and personalized messaging and experiences.

If a customer orders a 55” big screen TV online from an electronics retailer, the retailer could analyze the customer’s previous purchasing history and send the customer an offer for a package of related components such as a TV stand, home theater seating, speakers, etc.

Maybe the customer has already purchased some or all of these items from another retailer or maybe the customer just isn’t interested in them. Still, the retailer stands a better chance of cross-selling at least some of the equipment to the customer based on the relevance and timeliness of the offer.

Contact center interactions conducted through various channels (voice, chat, email, social, mobile) also give companies a lot of information about consumer channel usage, preferences, and behavioral insights they can use to craft personalized support experiences for customers, especially high-value clients.

Relationship managers and other decision makers for a regional bank could use analytics to determine exactly when a high-value customer uses the bank’s interactive voice response system to check account balances.

The bank could use this information to offer to push account balance updates to the customer through SMS alerts to his mobile device or via email during those times.

By customizing this service for the customer and by offering the service proactively, the bank demonstrates its understanding of what the customer wants while making it easier for him to access the information he needs.

Next Steps:

• Subscribe to our blog to stay up to date on the latest insights and trends in analytics and the customer experience.

Print post

A recent post in the UK’s Guardian newspaper sums up what many marketers are facing with a quote that’s often attributed to Albert Einstein: “Not everything that counts can be counted, and not everything that can be counted counts.”

All decisions can’t be based on real-time, infallible data, according to the post. For example, companies often can’t directly attribute sales revenue to social media outreach efforts.

Thus, companies should focus their data analysis efforts on areas that have more readily available measures that can track performance against strategic objectives.

For instance, businesses should consider these questions:

• Compared to other marketing channels, does the company have good reach across its target audience on social media?
• How does the target audience engage with the company’s content compared to how the audience engages with the content of the company’s competitors?

“Align the data you need to your strategic objectives, assess the options and make the decision accordingly,” the author of the article advises. “This could be big data, small data or somewhere in-between data. These principles hold true now, as they did when customer relationship management was marketing’s next big thing, and before the quote itself was conceived.”

Companies should also think about the answers to these questions:

• Is there a clear connection between strategy and the data being analyzed?
• Will the company collect the data it truly needs to make better informed decisions, despite the size of the data?
• Does the company have a roadmap for what to do with the insight gathered after the analysis?
• What experts – such as data scientists – does the organization have or need for data analysis?

The first step is for organizations to tap the data to be measured that best aligns with corporate objectives. The next step is ensuring that the insight generated via data analysis is presented in the most compelling and effective way to the decision makers.

One of the best ways to do this is through data visualization, telling the story of the data by graphically depicting statistics. But, what is the best way to tell a story through data?

A recent Harvard Business Review post suggests that the most important first step is identifying a compelling narrative.

“Along with giving an account of the facts and establishing the connections between them, don’t be boring,” according to HBR. “You are competing for the viewer’s time and attention, so make sure the narrative has a hook, momentum, or a captivating purpose.”

It’s also important to carefully consider the audience and what it may already know about the topic; the visualization should be built around the information the audience already has, both correct and incorrect, the post notes.

“A visualization should be devoid of bias. Even if it is arguing to influence, it should be based upon what the data says–not what you want it to say, according to the post. “Viewers and decision makers will eventually sniff out inconsistencies which in turn will cause the designer to lose trust and credibility, no matter how good the story.”

Next Steps:

• Subscribe to our blog to stay up to date on the latest insights and trends in data analysis and data visualization.

Print post

In fact, IT and business leaders are planning M2M implementations in areas ranging from smart grid energy networks to manufacturing and industrial plant monitoring to patient monitoring in healthcare, according to a recent study by TechRepublic and ZDNet.

Almost half of the survey respondents (48%) say they’re using or plan to use M2M technology for energy-related activities, such as smart metering, or communications between wirelessly-connected grid assets.

Forty-six percent of respondents are using or will use M2M communications for IT and network monitoring applications such as network traffic monitoring, while 43.9% are using or plan to use M2M technology for automotive, transportation, and logistics apps such as vehicle telematics and fleet tracking

Still, there are myriad opportunities for using M2M technologies – and big data analytics – in a wide swath of industries for different purposes.

For instance, a growing number of hospitals and other healthcare practitioners are using M2M data and analytics to track patients’ drug interactions and other aspects of patient monitoring.

Caregivers and insurers intend to use analytics to evaluate clinical outcomes (64%) and performance measurement and management (also 64%), according to a recent study by IDC.

For instance, hospitals, clinics, researchers, and insurers can analyze M2M data from patient drug interactions to help determine when patients are most likely to develop reactions to a particular type of drug (e.g., within the first six hours of initial intake) to train nurses and medical staff on monitoring and procedures for addressing dangerous interaction situations quickly.

Meanwhile, manufacturers also have numerous opportunities for applying analytics to operations.

For instance, sensory data gathered from a plant floor may determine that the bearings used for a manufacturing belt are running hotter than normal or are emitting vibrations that are outside of the normal range.

The use of analytics can help manufacturing leaders determine whether these indicators are symptoms of a pending breakdown.

The manufacturer can then use this information to order and install a part and shift production to another manufacturing line without have to disrupt operations. This is a much more cost-effective approach than incurring a breakdown and having to halt production until a replacement part can be ordered and installed.

Retailers are also finding ways to benefit from the use of M2M data and analytics.

British retailer Marks & Spencer has partnered with Avery Dennison to implement RFID technologies to make its inventory replenishment faster and more efficient. As merchandise is sold in a Marks & Spencer outlet, alerts are sent to the company’s distribution centers to automate the replenishment of merchandise to ensure that popular items are available to customers in store and online.

Additionally, retailers can use analytics against RFID data to glean other actionable insights.

For instance, RFID data that’s gathered and analyzed could inform retail executives whether an apparel combination that’s on sale (e.g., a specific brand of women’s blouses and skirts) is selling well or if certain color combinations are faring better than others.

Executives can then use this information to ensure the right mix of inventory is available for the stores where certain combinations are performing well. Executives can also use these insights to help determine the right mix of items and prices for future offers.

Next Steps:

• Subscribe to our blog to stay up to date on the latest insights and trends in data analytics.

Print post

The fact is, users of in-memory analytics are able to process more than three times the volume of data at speeds more than 100 times faster than their competitors, according to a study conducted by Aberdeen Group.

That’s largely because users of in-memory analytics tools are able to access and act on data so much faster than users of traditional analytics systems since in-memory technologies can avoid latency issues.

Ultimately, this provides companies that use in-memory technologies a leg up on their competitors in terms of business and productivity opportunities.

Consider some of the productivity benefits that leader organizations (top 20% of performers) have obtained compared to laggards (bottom 20%). For example, on average, it takes best-in-class companies nine days to integrate data sources compared to 137 days for “laggards,” or 15 times longer as measured by Aberdeen Group.

Additionally, 38% of leader organizations have adopted in-memory technologies while none of the laggard organizations have, according to Aberdeen.

Many companies struggle to integrate stovepipes of data from different channels, business units, and organizational functions, according to McKinsey & Company. In-memory technologies can help companies unify this information and provide decision makers with a more holistic view of customer data and other data sets that can strengthen decision making.

Companies that embrace analytics deep into their operations can deliver profit and productivity gains that are 5% to 6% higher than their competitors, according to McKinsey.

By making data readily available, in-memory analytics can enable companies to exploit information more quickly and make critical business decisions faster.

According to Aberdeen, top performers improved the accessibility of their data by 35% last year while organizations in the bottom 20% witnessed a 10% reduction in the same metric. Meanwhile, leader companies reported that a whopping 93% of their data was reliable while stragglers claimed that just 57% of their data was accurate.

Taken together, the enhanced access and response capabilities offered by in-memory technologies help organizations deliver the right information to the right decision makers at the right time. These capabilities have shown to deliver higher-than-average business benefits.

According to the Aberdeen report, best-in-class companies reported a 14% year-over-year improvement in customer retention compared to just 3% for laggards. And it’s at least six to seven times more costly to acquire a new customer than to keep an existing one, according to this article in Forbes.

To help put this in some kind of dollar perspective, consider this: It can cost a financial services company $175 to acquire just a single customer, according to author Lon Safko. Customer acquisition costs can quickly pile up for enterprise companies with thousands of customers.

By better understanding the needs of customers and using in-memory analytics to react quickly to these demands, companies can be assured of increasing sales and fattening up the bottom line.

Next Steps:

• Subscribe to our blog to stay up to date on the latest insights and trends in in-memory analytics.

Print post