Businesses are generating vast amounts of data across various fields. They are learning more about their customers, route optimizations, supplier information and more. Collecting data is not the issue; rather, it’s what an organization does with all of it. How they turn that big data into proprietary algorithms will determine the impact an organization can achieve. And, as we well know, the only way to process that much data is through some type of algorithmic-based software.
In a series on algorithms, mathematical optimization and business, we defined algorithms in supply chain and described how they could be implemented to achieve end-to-end business optimization.
To briefly recap, an algorithm is a kind of function that is used to predict and explain the cause of events. Specifically, algorithms help businesses take a large amount of data and make sense of it all. In the end, this helps them improve their decision-making process and, eventually, their overall business performance. (Click here to access the full article “What is an Algorithm.”)
For example, local government is often responsible for collecting trash. Removing the trash in an efficient manner is important if they want to keep residents happy, but trucks break down, workers call-off and weather often get in the way of that. An algorithm could sift through all sorts of data — such as, weather, time, traffic and available resources — to determine the best possible route for trash pick-up.
In the past, companies would hire analysts to find patterns and create data trends to determine the optimal decision/s. Depending on the data set, this could take weeks for a team of analysts to calculate. So algorithms were developed to run those same hand-calculations in a matter of seconds.
What good is an algorithm without application, though?
In recent years, organizations have been leveraging algorithms for predictive and prescriptive analytical applications that facilitate Integrated Business Planning and optimization. Without algorithms, optimization wouldn’t even be possible.
Targeting, segmenting and demand shaping by specific variables will become more refined as algorithms continue to improve. Peter Sondergaard, Senior Vice President at Gartner and Global Head of Research said, “Dynamic algorithms are the core of new customer interactions.” No surprise here: he’s right.
Organizations will continue to collect information about customers and what that collected information tells an organization. According to Brian Hopkins, from Forrester, “Customer-obsessed leaders will leapfrog their competition.” All of this information will require algorithms to define business processes and create unique, differentiated customer experiences.
Amazon is already doing this on a small scale. When a frequent visitor goes to Amazon, the home page is filled with items they may want to add to their cart. These items are all based on recent purchases and other customer variables. If you’re a frequent Amazon user, you’ve almost certainly fallen for this tactic, purchasing something that was so conveniently “suggested” to you. Talk about a smart way to influence buying behavior!
Organizations can add additional profitability through cross-sells and up-sells thanks to algorithms. They can shape demand by taking into account all supply and demand metrics. They can increase customer loyalty by being able to deliver a unique customer service experience. They can even save a life! For example, a hospital in Texas has been implementing algorithms to inform doctors which patients are at higher risk for heart failure. This ensures doctors take the necessary actions to keep high-risk patients as healthy as possible, like scheduling more frequent consultations and sharing educational material. Thanks to algorithms, doctors can prevent a health problem before it even occurs.
One of the largest expenses an organization incurs are employees, but they also happen to be the most critical aspect of a business. All of this is beginning to change as algorithms are being used to complete the daily tasks of many employees.
According to Gartner at the ITExpo, 20% of all business content will be generated by machines. (We’d all better start looking for a new job!) They also predict that by 2018 more than three million workers will be supervised by a “roboboss”!
Companies will become much smaller and more profitable as smart-machines replace employees. Some of this is already starting to happen in the supply chain industry. Companies are developing algorithms that will measure when inventory levels are low, evaluate suppliers on various qualities, rank them, and then place an order from the preferred supplier. The only task requiring a human is the last step: signing and dating the approval form!
The amount of staff needed to run massive enterprises will decrease, allowing organizations to reallocate resources to other channels that could more drastically improve profitability and revenue.
An organization’s profitability and success depend on their ability to recognize patterns or trends and then adjust their business quickly. How might an organization’s success be impacted if they have to wait weeks to run through all of the available data, rather than being able to do it in real-time?
Twenty years ago, a business would have been fine operating this way. Back then, almost all businesses operated this way since there was no popular, cost-effective software on the market. However, in today’s competitive environment, the ability to operate in real-time offers a significant competitive advantage.
For example, take two oil drilling companies. The recent burst has pushed many companies close to bankruptcy, nearly half of the U.S. shale companies may go bankrupt. The companies that survived were the ones who were able to spot trends and adjust their business before the low oil prices.
Oil refineries that saw the prices going down looked for alternative revenue streams and resource allocation. Perhaps they operated a fleet that hauled sand or water for the rigs, which they then used to serve other trucking modes. Perhaps they paused their growth so they could improve their cash flow since even profitable businesses sometimes declare bankruptcy.
Oil refineries cannot control the pricing environment, but what they can control is the efficiency of their operation. At these low prices and with profitability near zero, improving efficiency is the only competitive advantage an oil refinery has. And guess what? It’s only possible by operating in real-time.
Organizations are looking for ways to connect everyday items to the Internet so they can send and receive data. According to Gartner, next year, the spending will exceed $2.5 million per day on the Internet of Things (IoT).
We already see cars with internet connectivity, so don’t be surprised when the toaster starts connecting the Internet in order to calculate the optimal amount of time to heat your bread. (Imagine ... no more burnt toast!) Organizations will only be limited by their collective imagination.
Volvo recently announced a program to build 1,000 smart cars that would share information about road conditions. Imagine if this was applied to the logistics industry for truck drivers. Once a system like this is deployed, it could be used for predictive use.
For example, a system can be developed that uses on-going and historical data from millions of vehicles integrated with information such as weather to predict where slippery road conditions will occur in advance. This allows salt trucks and road workers to make the appropriate measures to prevent an accident that is often costly and deadly. Truck drivers could avoid certain routes based on weather or other various factors. Organizations would be able to keep their drivers safe and remain profitable.
Algorithmic business offers a large advantage to an organization’s operations, but there are still some challenges that will need to be addressed in the future, especially when it comes to demand manipulation.
For example, algorithms are not always sensitive to context. A customer’s response to a product or service is impacted on how well they slept the night before or their current mood. Until devices are created that can paint an entire picture, the use of an algorithm may backfire for an organization that is trying to provide a better customer experience.
Another problem with algorithms is the human element that is often difficult to account for. The procurement of supplies is not always as straightforward as evaluating a supplier and ordering. Sometimes deals that are not optimal financially are optimal for strategic alliances. Also, unless computers can find a way to negotiate with suppliers, it may not be able to achieve the lowest rate that would have been possible otherwise.
Of course, all of this can be avoided with the appropriate application of algorithms that achieve end-to-end optimization. These type of algorithms allow companies to know what to purchase in order to balance operations, financial and strategic aspects of the decision-making and planning process. Despite the challenges, “organizations that analyze all relevant data and deliver actionable information will achieve an extra $430 billion (on a worldwide basis) in productivity gains over less analytically oriented peers,” according to IDC’s Worldwide Big Data and Analytics 2016 Predictions.
With sufficient data sources, algorithms will be able to predict and offer quantifiable, actionable results — and in real-time. We trust that we’ve shown through this post that Peter Sondergaard was right when he said: “Algorithms are where the real value lies...”