Can Your Organization Avoid AI Failures?

By on

Artificial intelligence (AI) initiatives come as businesses enthusiasm for the technology grows, and AI initiatives go as the excitement fades in the face of false starts and little proof of value.

It doesn’t have to be that way. Early AI adopters are actually ramping up their AI investments and launching more projects — and getting positive returns for their work, too. Multiple organizations have gained a financial return from their AI investments.

At the same time, some industries are lagging. In particular, the life sciences and health care sector have made high investments in AI but are experiencing low returns. And while the industrial products and services sector has seen high returns from low investments, others that have made low investments have seen low returns. That includes financial services and insurance, consumer products, and government/public sectors.

There must be some reason for the distance that separates companies making equally low investments and seeing big returns vs. those that don’t. It may be attributed at least in part to the struggles that even businesses with AI success stories have had to overcome. According to a Deloitte survey, topping the list of issues are:

  • Implementation challenges
  • Integrating AI into the company’s roles and functions; and
  • Data Management concerns such as accessing and integrating data.

As the numbers show, there can be measurable returns for AI adoption, but companies must set themselves up for success from the start.

Think Before You Leap

Unfortunately, businesses don’t always do that before they act on AI. There may be pressure for many companies to do something with the technology fast, due to concern that their competitors are moving full steam ahead.

They may have a general idea that they want to use AI in their call center, for instance, but they don’t know what that looks like.

Tracy Malingo, Senior Vice President of Product Strategy at “actionable intelligence” solution vendor Verint, sees that all the time as she talks to companies about using Verint’s AI services and technology to power their customer engagement strategies. “People don’t know how to be successful at that,” she said. 

In order to scale and drive the customer experience, an organization must initiate AI in a way that defines smart use cases from a customer service standpoint. They must have a roadmap for measuring progress, too, she said.

That’s the path to avoiding what Malingo calls “AI fatigue.” That’s what happens when people wind up using AI just to try to get cost-savings or address something that they think is a customer pain point — not something that actually is.

Get the Evidence to Confirm the AI Build

The road is better paved, she said, when a company’s data, whether structured or unstructured, that relates to customer logs, voice calls, emails, chatbots — basically anything that connects with the conversations that currently occur between humans and customers — is closely examined before anything moves forward.

With a proprietary AI system and services like those offered by Verint, that information is the keystone for tracking end-to-end workflows, sentiment, and other characteristics. That knowledge makes it possible to bubble up different data-driven ideas or interests, she said.

The intent is to get an idea of whether a business actually can automate a particular aspect of customer engagement through machine learning — that is, find a real customer use case for putting self-learning AI to work.

“We do these types of analysis to show the customer what they can and should be automating,” she said, drawing upon her company’s library of knowledge related to what other customers have been successful at doing with AI for customer service. Verint, Malingo notes, has 15 years of experience building enterprise virtual assistants, with a huge library of curated business interests, intents, and ideas.

With multiple business models already defined that boil corporate data into these intents and ideas, “we can leverage that for more accurate and predictable results.” Sometimes, the results show that a particular AI use case the customer is interested in just won’t be a great success.

Verint provides consulting services for building evidence-based blueprints and then gives customers the option to purchase starter sets of prepackaged language models focused on a particular use case. Results of that AI effort then are verified through established KPIs to chart whether all is going well. That’s a step to moving to the next use case to be validated. The company, she said, bundles its tools, technology, and best practices into consumable deliverables to inform the AI blueprint and to measure it.

Transformational AI

The use of AI to solve a well-defined problem can go further than that. AI can actually be transformational to a business, Malingo said.

“This is about making humans more efficient and effective, multiplying them as a workforce, taking intelligence and driving different insights for business transformation.”

An example of that transformation via conversational AI, according to the company, can be found in Alaska Airlines’ use of the company’s multi-modal, multi-channel, multi-language platform to launch an intelligent virtual assistant for its self-service web operations. The intent here was to give travelers a personalized online experience that would connect them quickly and accurately to the services they need. It also used the virtual assistant to help employees navigate internal information across its highly populated knowledge base so that live agents would be able to accurately and quickly provide customers with a premier level of service.

The opportunity for transformation rests on bringing high-level data experts to the table when it comes to AI project discussions. Originally, Malingo said, the people who were interested in its solutions were operational teams looking to cut costs. Now it’s the company’s innovation group or machine learning specialists, as well as business leaders.

The need for education is there as companies want to move past initial failed deployments or to get started on the right foot. “It’s easy to fall in love with the technology,” Malingo said. “But I don’t think projects will ever be successful without a business partner to drive that.”

Image used under license from Shutterstock.com

We use technologies such as cookies to understand how you use our site and to provide a better user experience. This includes personalizing content, using analytics and improving site operations. We may share your information about your use of our site with third parties in accordance with our Privacy Policy. You can change your cookie settings as described here at any time, but parts of our site may not function correctly without them. By continuing to use our site, you agree that we can save cookies on your device, unless you have disabled cookies.
I Accept