Date posted: October 29, 2015
A recent article on Information Week points to how contextual awareness is redefining business intelligence. Contextual analysis— aided by mobile devices and the Internet of Things (IoT)— is providing enterprises what they need to reach consumers in relevant and meaningful and profitable ways:
Contextual analysis helps answer the “why” questions. With it, organizations are gaining deeper insight into the behavior of their customers, employees, equipment, situations, and trends. Marketers can deliver relevant messages and user experiences, and business leaders can make accurate decisions. Although contextual analysis isn’t a new concept, it’s being applied to a variety of cases to improve bottom-line performance.
The article points to a number of specific ways in which contextual analysis is helping business intelligence achieve new benefits:
- Decrypting purchasing behavior.
“Context is key because low engagement is not necessarily a suppression of intent to purchase,” said Dean Abbott, chief data scientist at customer marketing intelligence platform provider SmarterHQ, in an interview. “Context allows you to understand intent. You can refine the message and identify an audience that might otherwise have gone unnoticed.”
- Predicting effects on stocks
“There are ways to measure context in real time so you can predict much more than you could before,” said Luca Scagliarini, VP of strategy and business development at semantic intelligence platform provider Expert System USA. “Board members are usually board members of other companies. If you look at the votes each member expressed on similar issues in other companies, you can estimate the vote. [And] public opinion about a certain issue or important news [may influence how] the board votes around environmental issues.” Analyzing public company financial statements, such as annual reports, can provide insight into board members and their past voting behavior. Public sentiment can be monitored via newsfeeds, blog posts, and social media. All of that can help explain the context in which a decision will likely be made.
- Keeping Energy Flowing
When an oil field or an oil well isn’t producing, millions of dollars are lost in a single day. To avoid unnecessary disruptions, oil and gas companies are performing contextual analyses of their operating environments using unstructured content, including maintenance notes, drilling reports, and other sources, to gain qualitative insights about non-production time.
- Improving Return on Investment of eMail
The idea here is to deliver eMail that is contextually relevant at the time it is opened, rather than at the time it was sent, using various types of data such as geolocation, device type, time, and weather conditions, among other factors.
- Fine tuning cybersecurity
Without context, anomalous behavior can be misunderstood and misclassified. For example, if an employee downloads a 1GB file at midnight, the event alone suggests malfeasance. However, viewed in the contexts of the employee’s historical behavior, the behavior of the group in which she works, her role, and the behavior of the organization at large, it may become clear that the act had a legitimate purpose.
- Improving the quality of BI
“Without context, data is almost useless,” said Marius Moscovici, CEO of push intelligence software provider Metric Insights.. “If a number lacks context, it just leads to a lot of analysis to try to work it out. With context, you can take appropriate action right then and there.”
- Managing potential disasters
Consider this example: The city of Buenos Aires, Argentina, is analyzing contextual data from sensors to determine in real time which areas need immediate support. In 2014, the city and its citizens were preparing for the largest downpour in history. In such a situation, flash flooding can occur simply because trash and debris are clogging storm drains and city sewers. “In Buenos Aires, you have the ability to use sensors in water tunnels underneath the city to map the water flow, the rate of the water flow, and how fast it’s rising. But if you don’t have the contextual analysis of an inspector using a mobile device entering the information about the inspection, then you’re not going to be able to alleviate the issue. You need both sides of the equation,” said Dante Ricci, lead global public services and healthcare marketing and communications at SAP. Because the maintenance crews knew when and where to clear the storm drains and sewers, a lot of damage was prevented.
- Delivering more timely user experiences
Contextual mobile marketing has been on the lips of marketers since before the turn of the millennium, but the technologies and economics necessary to deliver it have taken many more years to develop. Finally, it’s possible to sense the location of an individual and her proximity to a possible destination, whether that’s a retail store or a coffee shop. Armed with that and other information about the customer, marketers can target an offer that is relevant to the person, place, and time. The same technologies can be used to alert employees to potentially dangerous situations or to turn app features on and off based on a context such as driving.
- More intelligent management
“Context greatly improves performance management. It can help us sort through piles of data around people, how we’re managing people, what we’re getting done at work, and what will happen in the years ahead,” said Kris Duggan, CEO of goal-setting platform provider Betterworks.
- New perception
Wearables are the new, new thing, capable of providing context that may not have been previously available. Instead of relying only on location to infer what might be relevant, it is possible to “see” what the wearer is perceiving when she is perceiving it. The benefit to marketers is finer-grained targeting and higher degrees of relevance, but the applications span everything from surgery to law enforcement to military operations. The same capabilities are being used in a variety of sports stadiums on the sidelines and in the stands.
Bottom line, contextual analysis can help organizations better understand why certain things are happening, not happening, or happening not in the way anticipated. What qualifies as relevant context varies based on the use-case, the environment, and the particular circumstances, but the use of context clearly furthers the application of business intelligence.
Date posted: October 27, 2015
I was sitting on a tarmac recently— plane delayed due to last minute maintenance— when an article in the Washington Post caught my eye: a new partnership between Boeing and Carnegie Mellon University might keep me away from such irritating delays in the future.
As part of a three-year, $7.5 million deal that will establish a new Aerospace Data Analytics Lab, Boeing and the Carnegie Mellon School of Computer Science will work on a range of new projects that will apply the principles of AI and big data to improving the quality of Boeing’s aerospace activities.
While the principal goal of the venture is to make sense of the burgeoning amount of data in the aerospace industry, one of its outcomes may be what the Post terms the self-healing airplane.
“The mass of data generated daily by the aerospace industry overwhelms human understanding, but recent advances in language technologies and machine learning give us every reason to expect that we can gain useful insights from that data,” says Jaime Carbonell, Carnegie Mellon professor and Director of the Language Technologies Institute, who will head up the new Aerospace Data Analytics Lab.
One example of how machine learning can be used to gain useful insights is the whole issue of airline maintenance. Think of airline maintenance the same way you think of maintenance for our car – you can follow the generally suggested guidelines for your vehicle (e.g. an oil change every 3,000 miles) – or you can use real-time data to see which planes needs fixing, when. By fixing planes before – not after – they need maintenance, an airline flying Boeing planes could gain a real competitive advantage over its peers.
This means determining when planes actually need maintenance instead of following historical maintenance schedules.
And while Boeing’s CIO speaks of using this initiative to “push the technology envelope,” I am thinking it would be great just to push these planes out of the airport on schedule— a victory for intelligence artificial or otherwise.
Date posted: October 21, 2015
How a manufacturing organization handles inventory management has a direct bearing on its likelihood of success. Do it well, and it can depend on having the materials it needs on hand when it needs them; do it poorly, and significant problems ensue.
To have the right products at the right place at the right time, retailers put a lot of time, thought and effort into inventory management, allocation, and replenishment processes. of their business process, then purchase the inventory and allocate it stores from a distribution center (DC). When inventory runs low at a particular store, the DC replenishes it. That’s the idea.
An article on Predictive Analytics Times speaks to how inventory distortion can derail the idea, and how predictive analytics are working to keep them on track.
Inventory distortion is a term used by retail technologies consultant IHL Group to describe the challenges faced in the typical inventory management process. Examples of what causes this would include:
- When products don’t sell well at specific stores, retailers are forced to markdown inventory in order to clear the inventory.
- When there is no more inventories in the DC, or even before inventory is replenished, retailers experience lost sales at stores where these products are still in demand.
- Customers resolve to purchase from a local competitor in order to fulfill the sale.
- Lost sales paint an inaccurate picture of lower demand, which in turns makes a retailer order and allocate less inventory to stores in the future. (This cycle repeats itself.)
According to IHL Group, inventory distortion costs retailers nearly $800 billion globally. Increasingly, notes the article, retailers are fighting this by using predictive analytics for inter-store inventory balancing:
In order to properly forecast demand, and build the optimal inventory management strategy a retail solution must be tailored to the specific business process of the retailer, which means that it must take all factors that affect demand into account. Predictive analytics technology will significantly improve demand forecast accuracy, and suggest better allocation and replenishment strategies. Moreover, there are specific tools that enable a retailer to further decrease inventory distortion. One such tool is inter-store inventory balancing.
Inter-store inventory balancing leverages predictive analytics to proactively analyze all the influencing factors of a retail supply chain, then recommends an optimal inter-store transfer schedule to move slow-selling products to stores where there is a higher demand for them. Retailers that use predictive analytics for this technique report significant benefits, with tangible results such as:
- A 25-40% decrease in inventory costs
- Sales increases of up to 20%
- Increasing turnover by a factor of up to 3.5x
This is just one way retailers are leveraging predictive analytics, but one that’s indicative of why its use in inventory management is destined to grow.
Date posted: October 20, 2015
“Few things are harder to put up with than the annoyance of a good example.”
—Mark Twain
Case studies continue to be a mainstay of the work I do. During the past 25 years, I’ve written more than 500 case studies. (Honestly, that number is probably low. I stopped counting a decade ago after reaching 350.)
At Aprilaire, their legacy system was outdated, expensive to update, and didn’t provide easy integration of work orders and preventive maintenance activities, or fast, easy access to information/reports. When they began working with FacilityDude, the company enjoyed better visibility across maintenance operations; ease of use and simple report generation; and significant time and labor savings.
The power of the case study is evergreen for good reason. Case studies are a potent marketing tool.
Date posted: October 19, 2015
How a manufacturing organization handles inventory management has a direct bearing on its likelihood of success. Do it well, and it can depend on having the materials it needs on hand when it needs them; do it poorly, and significant problems ensue.
Brookfield, MA-based TLC Group posted five top best practices for inventory management on its Knowledge Center blog; they’re worth another look:
- Categorize inventory
Portions of your inventory will move faster than others and, therefore, require a different management approach. Indeed, the 80-20 rule applies to the items you stock, so it’s a good idea to categorize inventory and set priorities accordingly. For example, you might want to make sure you have a bigger stock buffer for your fastest moving items. Slower moving items may call for less security. This best practices inventory management approach is sometimes referred to as ABC analysis. Sales numbers are often associated with ABC analysis, but profitability is another way to prioritize.
- Focus on demand forecasting
A company’s demand fluctuates due to seasonality, economic climate and other business trends. A solid forecasting capability can help you plan inventory and maintain appropriate levels, avoiding excess inventory or shortages. Companies can study past sales trends to determine likely future patterns and align inventory management policies to reflect those expected patterns. But there are other techniques to consider (e.g., qualitative assessment and the time series method).
- Apply automation
In particular, smaller manufacturing companies may track inventory via multiple spreadsheets and lack a unified view of stocking levels. A centralized inventory management system, a module included in many ERP products, provides a way to keep accurate inventory counts, deal with unexpected events, avoid overstock situations and boost inventory efficiency. An ERP suite can also automate such tasks as demand forecasting and ABC analysis. Such systems represent the systems side of best practices for inventory management.
- Look for underlying problems
Those who overlook history are bound to repeat it— this holds true for inventory management. If you find that a certain item is perpetually in oversupply, get to the heart of the matter. Consider performing root-cause analysis on excess and obsolete stock and understanding how those stocks are connected to action plans for combating future excesses.
- Consider alternative inventory models
Inventory management covers an array of models, so it makes sense to evaluate new approaches from time to time. If your organization finds it difficult to keep tabs on a certain item, approaches such as vendor-managed inventory (VMI) can help you share the burden.
Are you employing these practices? Others? We’d love to know what approaches you are taking to ensure efficient inventory management.
