Building an Algorithmic Business

Intel Corporation explores how, for an enterprise to become an algorithmic business, the challenge is as much about the organization as the technology.

 

What is an algorithmic business? Is this yet another marketing term, or does it have relevance to today’s enterprises? After all, algorithms have been around for decades, so why do we even need such a term?

The answer is that having algorithms in your software does not make you algorithmic, in much the same way that having gym equipment in the garage does not make you an athlete. Algorithms apply mathematical rules to data sets and deliver a useful result, but it is what you do with this result that matters.

Consider the following examples.

  • In online retail—recommendation engines use algorithms to suggest products based on past purchases, browsing behavior and the actions of ‘similar’ customers. While the recommendations may not always fit the need, they nonetheless help a customer think about what they are looking to buy.
  • In healthcare—diagnostic algorithms are being used to support medical professionals, creating a second opinion and ensuring that they have covered all the bases, for example checking interactions between drugs based on patient history.
  • In transport and logistics—algorithms are analyzing diagnostic data and then feeding it into spare parts management and re-ordering so that parts are available in advance of an engineering fault being detected.

In each case, algorithms are doing more than simply number crunching: they are acting in direct support of certain roles (in the above, customers, medical staff and inventory managers). The people involved are empowered to make better decisions, faster and with reduced risk.

So, why aren’t more businesses algorithmic? All too often the answer, as we discuss in our paper, Five Steps to Delivering the Data-Driven Business, is that organizations simply use data analytics to aid reporting, which feeds a hindsight view of business activity: last month’s sales or customer service data may be interesting, but may already be out of date.

For insights to empower the business, they need to be delivered at the point of need and in a timely fashion, that much is certain. Equally important, however, is that they need to be provided into organizational structures and processes that are capable and mature enough to do something with them. For example, there is little point offering the day’s retail sales data, if pricing changes can only take place once a week.

So, the algorithmic business involves a direct relationship between algorithms on one side and an empowered business on the other. This isn’t something you can buy from any vendor but needs to be developed from within. Which is why, as a company with a broad portfolio of analytics technologies, we spend much of our time helping our customers understand where they are on the analytics journey, then increasing engagement between business and technology teams so they can achieve their analytics goals.

Building an algorithmic business cannot happen overnight, but requires mindset change from top to bottom. Strangely, this does not have to be hard, given how as consumers we are all used to, and even expect, algorithms to assist us in our online lives. At Intel, we are driving tech innovation – from the edge to the data center and the cloud – to prepare businesses to thrive in the age of data. Given that analytical technologies are now mature enough to tackle most data challenges a business might have, the epiphany all organizations need to go through is that the algorithmic business is about people engaging with their data.

Information can already be at your fingertips, to empower you and your organization. However, becoming algorithmic – and as a result data-driven – cannot happen by accident. It requires all layers of the business to make an active choice.

Learn more about how advanced analytics can help you transform your business, and what you can do to make it happen, by reading this new white paper from Intel.