This article is part of the series

Cutting Edge

A systematic approach to implementing cognitive technologies

By Cutting EdgeDeloitte

Companies can often set themselves back years by taking the wrong approach to implementing technologies. Through their client experience, Deloitte has found that when senior executives feel pressured to transform their business, they encourage the company to simply “do something cognitive,” and implement a program without proper planning or internal knowledge.

Continue Reading Below

These are typically high-level, contracted out programs that more-often-than not fail. In “Beyond Doing Something Cognitive,” Deloitte outlines a more structured and thought-out approach to adopting cognitive technologies that businesses should adopt to achieve better results requiring less time and money.

Opportunities and Challenges

Transformative projects are high-risk, high-reward. When the leadership of a company expresses interest in new technologies and a desire to transform business, a clear view of the goals and risks of the project should be established at the outset to manage expectations and set a course.

Problems often arise when a pilot program is overly ambitious—taking up too much time and money up front—or if the program is not designed with human users in mind, which will hurt the acceptance and adoption of a project by the end users.

Another common problem is choosing the wrong technology to solve a given problem. Robotic process automation (RPA), traditional machine learning, deep learning, natural language processing (NLP), and rules-based expert systems are all examples of the major types of cognitive technologies available today. They are among the options that should be studied and considered when trying to address a problem.

A Systematic Approach

To avoid the negative outcomes and potential pitfalls of a transformative implementation process, and to prevent the cynicism that might come with a technological failure, organizations can follow certain steps to a more systematic approach toward adopting cognitive technologies.

These steps, along with an individual coordinating the early stages of the process, can help an organization harness the potential that cognitive technologies present.

  • Educate senior management on cognitive technologies and their likely impact. To make informed decisions, executives need to be educated about the various types of cognitive technologies and what they can accomplish. The best way is internal education from someone in-the-know within the company, not just reading about it from outside sources or hearing about it from vendors.
  • Select the right technology for your business problem. Understanding the proper use and implementation of cognitive technologies is essential. Certain programs may require more than one type, so decision-makers should have knowledge on what tools they have at their disposal and what they do.
  • Form a “community of practice” of interested and involved employees.  Cognitive summits or a community environment where executives and employees can share experiences and exchange ideas about cognitive projects can help further learning.
  • Recognize that “low hanging fruit” projects tend to have a much greater chance of succeeding, even though they have less potential business value. The best projects to start with are ones that perform a limited task that incorporate some cognitive technology, and tackling more ambitious projects later. Deloitte has found that overly ambitious projects that push the limits are most likely to fail.
  • Build in expectations for learning and adaptation. Cognitive systems run on information, and therefore, by nature, will get better over time as they acquire more data. A system may not hit the ground running and be fully optimized at launch, so expectations should be adjusted. Success should be measured by the ability of the system and team to improve over time.
  • Get a portfolio of projects going. Deloitte suggests gaining experience quickly with smaller pilots or a collection of programs in various areas. Stay agile by using a “minimum viable product” approach and figuring out what works.
  • Discontinue some projects, scale up others. Don’t be afraid to admit failure and cut bait on technologies that don’t work. At the same time, those that seem to be working should be scaled-up. That’s why starting small is key, so you can abandon low-stakes projects without a substantial loss of time and resources while focusing on what seems to be successful.
  • Follow the changes in technology, and continue to educate leaders. Cognitive technologies and their capabilities are changing rapidly, and new vendors are emerging. Executives should get regular updates and stay on top of the latest trends.

While these steps may seem like more work at the beginning of the project, this approach will make it more likely to achieve results while ultimately requiring less time and money in the long term. A successful technology initiative can also positively influence how a company views change in the future, helping encourage and embrace technology transformations down the line.

Click Here for More Insights from Deloitte

Authors: Tom Davenport, John Houston