Gartner Blog Network

How to Manage Algorithms in Digital Business

by Peter Sondergaard  |  December 16, 2015  |  5 Comments

Software algorithms are at the core of all products and services in the digital economy. As I explained in a previous blog post, proprietary algorithms will come to define the function and value of almost everything we will use in the future.

Whether it’s an algorithm that keeps us at a safe distance as we navigate traffic in the fast lane, keeps the ever-taller high-rise apartment building upright in high winds or offers the best seasonable accompaniments to compliment the meat we ordered online, algorithms increasingly impact everything in our lives.

The same is true for CEOs – whether they lead a bank, an industrial manufacturer or professional services firm, and for leaders of government institutions. For this reason, in the digital economy, CEOs must focus the same attention on managing their organization’s proprietary algorithms just as they would safeguard traditional analogue products and intellectual property.

To prepare for a future defined by algorithms, CEOs must do three things:

  • Inventory your algorithms
  • Assign ownership to manage those algorithms
  • Qualify which algorithms are public and which are private

As they say, there’s no time like the present.

Inventory Algorithms

First, do you know that your business already has many algorithms you rely on? This includes algorithms in products and services, in business processes and via external partners, whether in software or as part of products from sub-suppliers. Most businesses reply on a core set of proprietary algorithms and the number is growing. Since they are core to most products or processes, it is paramount for you to do an inventory. Without a good understanding of what you have today, you may be giving strategic components of your future business away without ever realizing it.

Assign Ownership

The recognition of algorithms as strategic assets leads to the need to assign ownership and responsibility for coordinating corporate activities. The emerging role of chief data officer (CDO), who is responsible for data and information, is a natural candidate. Consider extending their responsibilities to include algorithms.

If you choose this route, ensure your CDO focuses on algorithms internal to your organization as well as those used in relationships with external partners of software or products.

Since algorithms define many of the products and services that your partners provide, they will also define the safety and quality of your products or services. This impacts the definition of responsibility of your chief risk and security officer (CISO). S/he will become responsible for not only risk and security, but also quality and safety as products are increasingly defined by software. This effectively erases the distinction between software-based security versus quality and safety. For example, how can an autonomous car, defined by software algorithms, have several different executives responsible for security? Ultimately someone needs to do the overall coordination and be responsible to the board.

Algorithms — Public or Private?

CEOs must define which algorithms are private and which algorithms can be made public. Selling, licensing or even giving away your algorithms can enhance your influence in future digital business ecosystems. Tesla and Goldman Sachs have decided that giving away some of their algorithms is beneficial to the future of their business because it creates an ecosystem that ultimately benefits them in the future.

So now is the time for you to audit, assign ownership and begin the debate within your leadership team about which algorithms should be kept secret and those that should help create new public ecosystems. Treating algorithms strategically is your responsibility. Make this your New Year’s resolution.

Category: big-data  symposium  trends-predictions  

Peter Sondergaard
Senior Vice President and Global Head of Research
25 years at Gartner
29 years IT Industry

Peter Sondergaard is a senior vice president in Gartner, where he is the global head of Gartner Research. Mr. Sondergaard is responsible for people management and the direction of the global research organization, which includes Semiconductors, IT Infrastructure and Operations, Communications, Software and Services Management, Business of IT, Research Operations Management, and IT provider and end-user organizational roles.

Thoughts on How to Manage Algorithms in Digital Business

  1. Ravi Kant Dixit says:

    Valid inputs Peter. The first step is the hardest though, most organizations would struggle to inventory their data assets, let alone their algorithms.This disruption will impact not only Marketing, IT, Procurement, Services or Supply chain, but HR and organization structure as well.

  2. I know about this – 20 years too early, Mathcad took the task of documenting algorithms, even protecting them through XML. The product was acquired by PTC as the necessity for documentation is obvious. Yet, this is a task separate from other design and as commented earlier, very often neglected by many. However, with increased complexity and as patents will rule the future, this is increasingly relevant. See for more

  3. We have customers who are doing this using our data governance platform. It comes up most often in the context of complex analyses that are done to drive different types of corporate behavior. Often, the motivation to begin is much more simple than you describe… it is just that between the data, the models, the processing jobs that they invoke, there are a lot of moving parts. Organizations need to validate that everything is as it should be.. otherwise the the outputs are not trusted.

  4. Paul Castro says:

    Peter, many people talk about the best way to explote internal information by algorithms that compile the sources of information about every use case (for example churn, loss of service), but really is very complex develop and do the algorithm because the companies that have more than 15 or 20 years in the industry, maybe start reviewing and checking all the sources of the information, defining internal projects to consolidate them, migrating or updating the legacy data, redefining the architecture, etc. I know that the CDO role is important for the future, but in this context how we can manage the government of the information and the ownership?

  5. Chris ONeill says:

    All true. And because leading companies will inventory and work to protect their proprietary algorithms, there will also be a need for independent translation networks connecting competitors in order to protect entire industries and the economy from the systemic risk of the failure of a single company.

    Examples of this are competing banks joining credit/debit card networks like VISA and MasterCard and competing airlines connecting their systems via Global Distribution Systems like Sabre.

Comments are closed

Comments or opinions expressed on this blog are those of the individual contributors only, and do not necessarily represent the views of Gartner, Inc. or its management. Readers may copy and redistribute blog postings on other blogs, or otherwise for private, non-commercial or journalistic purposes, with attribution to Gartner. This content may not be used for any other purposes in any other formats or media. The content on this blog is provided on an "as-is" basis. Gartner shall not be liable for any damages whatsoever arising out of the content or use of this blog.