Artificial Intelligence-real world applications

Intelligent Systems

By Cassandra Balentine

Part one of two

As artificial intelligence (AI) tools continue to develop they are integrated into more applications (apps) and improve varied business processes. One area rolling out AI capabilities is customer relationship management (CRM).

CRM software acts as the hub for all critical information needed for companies to grow. “Because it is the hub, it houses an enormous amount of information. The more connected a CRM system is to other apps, the more information it will store,” says H. John Oechsle, CEO, Swiftpage.

A lot of value is locked up inside of CRMs. Big data analytic tools are essential for unlocking that value, but they can only go so far. “AI is the last mile of the value equation. It is the foundation for a recommendation architecture that allows the software to work for you rather than you working the software, therefore AI plays a huge role currently and will be critical in the future of CRM,” explains Oechsle.

AI helps overcome limitations to CRM technology. “AI enables actionable insights to be presented to the CRM user without time-consuming data mining. It also allows businesses to surface new insights using adaptive modeling and machine learning,” says Jeff Nicholson, VP, CRM product marketing, Pegasystems.

Automating the Customer Experience
AI technology is prevalent in many consumer interactions and business processes and its continued potential is apparent.

Consumers interact with AI when chatting to Alexa or Siri, or when Amazon recommends a book you should read. “Making the transi-tion into the CRM/business apps world is inevitable. Forward-thinking organizations are already exploiting the technology,” says Paul White, director of customer engagement solutions, IFS.

Kevin Draggoo, product manager, Infor, explains that currently, AI’s primary role is to analyze data. “The underlying algorithms are still programmed by humans, but AI can analyze vastly larger sets of data in smaller timeframes than its human counterparts.”

AI brings a better understanding of customer behavior to optimize the sales process; utilizes automation to alleviate simple, time-consuming tasks from human administrators; and adapts with machine learning and natural language processing.

AI can learn to understand CRM users and their behaviors. “This might proactively compensate for the user’s tendencies to favor ac-tivities with an inherent bias, such as preferring emailing or texting over calling. It might also understand when a user is most likely to take on more challenging tasks during the day. This helps the CRM user become more effective and productive,” offers Jonathan Novich, VP of product, Bullhorn.

The broader role of AI is to help drive behavior. “With CRM, AI is making sales and marketing personnel more effective by churning through extensive data about a prospect or customer, narrowing datasets to focus on what matters most, and help identify the best approach for engaging a prospect. It is also helpful for sales and marketing teams as they work to understand which product a cus-tomer is likely to buy and determine the best tactics to drive the desired response,” explains Draggoo.

AI helps salespeople identify and prioritize sales opportunities. “With AI-based scientific segmentation, salespeople are guided to gaps in their customer’s purchasing history and both cross-sell and upsell opportunities can be directly served up to a salesperson, enabling them to spend less time searching for opportunities and more time focused on closing deals,” says Valerie Howard, senior product marking manager, PROS, Inc.

AI automates insights within CRM. “Humans are too busy doing the things computers can’t do—working with clients, solving prob-lems, managing their teams—so they don’t have much time left to do investigation work or analysis,” comments Geoff McQueen, CEO, Accelo. AI looks for insights, concerns, and patterns when people don’t have the time.

Howard says today’s buying and selling landscape has undergone radical changes in the digital era. Many refer to the Amazon Ef-fect, which has come to light in both the business to business and business to consumer selling environments. “Buyers are less in-clined to engage in long selling cycles and instead prefer fast, frictionless, and transparent transactions to buy products and services. They also want personalized quotes. That’s where AI plays an essential role in maximizing the value exchange in every unique cus-tomer interaction. It’s the personalization of every quote that enables selling to optimize and prioritize every interaction,” she com-ments.

AI is capable of learning over time to make better decisions, making it an effective player in CRM, business process management, and workflow systems. It can analyze data patterns, detect deviations, understand coherency and correlations between data, and suggest or even carry out recommended actions through automated task fulfillment. Tibor Vass, global solution director for business automation, Genesys, shares that AI can cooperate with humans or other bots to analyze and initiate corrective actions together. “We call this concept blended AI—when machines and humans work together to effectively augment each other’s capabilities.”

Benefits of AI
The use of AI in CRM brings improved automation for efficiency and productivity. With less time spent on administrative processes, employees have more time to work on customer relationships.

McQueen says the reality is the promise of AI is leveraged downstream of what most CRM software is today, in places where projects are delivered and work is done. “This post-sale dimension of the relationship is one where there is a lot more at stake. It is where profits are made or lost, clients are happy or dissatisfied, and where there’s a lot more interaction between members to achieve suc-cess.”

AI is also used to optimize pricing. It is a key component in ensuring organizations offer prices that are competitive, transparent, and preserve profitability. “Today’s digital marketplace demands that pricing be responsive to changing market conditions, and there is no more efficient way to do this than with AI,” comments Howard.

It excels at discovering patterns, predicting the future, recommending next actions, and automating busy work. “By focusing AI on these tasks, you can help every employee work faster, smarter, and more productively, allowing them to focus on driving value for your business in different ways,” says Marco Casalaina, VP product management, Salesforce Einstein. “Human traits like empathy, creativity, and intuition will always be valuable—especially when it comes to building and maintaining a relationship with customers. By leveraging AI to handle other tasks, humans are able to better focus on building a connection.”

Companies are able to process large amounts of data to predict the next best action in real time with each individual customer, re-gardless of channel with the help of AI. “This is applicable in most customer touchpoints, from customer acquisition and retention to cross selling and upselling,” says Nicholson.

It serves as a mechanism for diversifying a company’s approach to problem solving. “Many users look for one right way to do things through analysis. Instead, when presented with data and the outcomes, an AI-based approach may recommend paths that were not previously considered. This saves users much-needed time and gives them additional insights that can create stronger relationships or inform better business decisions,” shares Novich.

Howard says companies that utilize AI with their CRMs can deliver more with less and do so even faster. “In the digital era, personal-ized offers increase win rates. Real-time price optimization reduces quote turnaround times by as much as 80 percent by enabling auto-approvals for personalized, negotiated rates while still protecting margins. Opportunity identification and prioritization enables sales teams to make the best use of their most limited resources—time. “At this point, the most meaningful customer relationships and experiences depend on human to human interaction, and AI is not yet a viable stand-in for those interactions, but it can help drive and sustain them. It can automate day-to-day tasks that do not require human involvement and deliver insights that human sales ex-perts can use to forge and strengthen relationships that lead to new opportunities,” says Draggoo.

“The partnership between humans and their AI counterparts, when approached holistically, make the customer experience more effi-cient, accurate, and seamless, and offers optimal guidance on decision making. The resulting customer-centric solutions are built from the ground up with one primary goal—serving and thrilling customers by marrying machines with people into one system,” says Nicholson.

Vass suggests that by automating processes and handling most customer interactions autonomously, we can decrease human effort and save time.

“Monotonous and repetitive queries are deflected away from contact center agents, allowing them to focus on the more complex tasks,” says White. The same AI tools can be used to listen in parallel with human communications and provide automated prompts and suggestions to help the agent respond to the customers’ needs. “This can be a valuable way to crib less experienced agents involved in complex product lines or processes,” he says.

The benefits of incorporating AI technology into CRM ultimately improves customer relationships. “AI technology provides CRM users with more time, insight, and direction to provide the best possible service. Effective AI technology can help a sales team identify the right market for a specific product, the features a specific business or customer would find most beneficial, and cultivate a positive relationship with that person or business through the sales cycle and beyond,” says Oechsle.

Challenges of AI in CRM
As with any technology, challenges come with the benefits of AI. Silos, data quality, integration, and privacy are top considerations.

Nicholson warns that the lure is to rush in and build AI in silos—a chatbot here, an intelligent virtual assistant there, and intelligence on the website. “Customers don’t care that most were never really designed to work together and soon all that is left is a mine field of disparate silos that function in isolation, delivering lackluster results that fail to scale,” says Nicholson.

Customers get frustrated with inconsistent and disjointed experiences. “To maximize the benefit, organizations need one central AI brain that works across team silos so customers get the same experience no matter how or where they engage with the brand. The place to build your AI strategy is not within a channel or silo, but in the center, where it can be then extended and leveraged on any channel or touchpoint without replication. This is the power of a focused customer decision hub strategy,” he adds.

In addition to data silos, data quality is critical.

McQueen cites the testament, garbage in, garbage out. “This is the painful reality professionals face with business management technology today. The problem is that the people who are expected to enter the data, which is then used for reporting, see no yield for their efforts besides keeping their job. Conversely, the folks who benefit don’t actually do the work that’s required. This misalign-ment of motivation means that the data entry is poor when it depends on human effort and this means people can’t reply on the re-ports being produced—so the whole thing becomes a failure. To make AI work in CRM, first CRM has to do its job efficiently, which is why automation needs to be successful before you get into AI.”

Vass agrees, noting that too little or too much low-quality data in segregated databases can be a barrier for AI to make decisions in near real time. “Think about your typical daily decisions. You often make them based on incomplete or inconsistent data. But it is eas-ier for humans because we are equipped with intuition. In the case of CRM or workflow solutions, AI only works with the data that ex-ists and is available in reliable quality. It can’t work based on feelings only—at least not yet.”

Another data-related challenge involves privacy. “AI works best when it has a large dataset to draw from. However, the usual infor-mation that CRM needs to target can be limited. AI would be more effective if it could leverage broader datasets on demographics, markets, and industries across an entire customer base. This would allow for it to tell the user more than what has been successful with a particular customer in the past, but there are well founded concerns regarding how much data is too much and whether it com-promises an individual’s or company’s privacy,” says Draggoo.

Additionally, AI often relies on having a training set with correct answers. However, in the realm of relationships, a right answer can often be difficult to identify, so it’s important to scope the problem correctly before providing the corresponding answer or response. “In order to implement AI into CRM software effectively, many choose to narrow down the problem to solve the broader problem. For example, rather than rely entirely on AI to suggest how to engage with any of your customers, an initial implementation might focus on a specific set of use cases—identifying customers who meet a specific set of conditions and mapping a sequence of interactions that model appropriate responses and reactions,” says Novich.

Integration is another consideration. The challenge with AI and CRM is to integrate the technology in a way that provides significant value to the end user without becoming too bulky or cumbersome. “Users today want quick and easy access to their favorite tools and features. There is little time or patience for slow or clunky user experiences or extensive training. Any new AI technology advantages need to be clear to the user and allow the user to accomplish more in less time,” says Oechsle. He says if the technology offers im-proved technology but takes longer to navigate, they’ll likely never take the time time to truly commit to implementation. “The technol-ogy is coming, that’s not a problem. The challenge is to deploy any new CRM AI technology in a way that improves performance and doesn’t add heft or slow down workflow,” he adds.

Casalaina says traditionally, AI tools are not pre-integrated into CRM and companies require highly skilled data scientists to make it work. “Businesses would have to extract the data from their CRM system into an AI toolkit, and a data scientist would take the data, massage it, apply various algorithms it, and generate predictions. Then the business would have to find a way to communicate these predictions to its users,” he says. As AI tools are increasingly integrated into modern CRM tools, it eliminates this step.

It is a challenge to define the exact dataset needed to drive the desired behavior. Draggoo explains that you cannot anticipate every possible permutation a path to purchase could follow, the decisions involved, or how conditions will change, so AI must be able to adapt on the fly as data becomes available.

White believes the secret of success is not to be too ambitious with initial AI offerings. “An AI chatbot can answer a common question, like ‘what time is my appointment?’ or ‘is this product in stock?’ Achieving this initial step keeps the customer from having to wait in a queue for a live agent and is a very practical way to drive real value with minimal complexity. Remember that AI isn’t a channel in the omni-channel universe in its own right, it is a value-add capability for your messaging, self-service, or contact center channels—these need to be ‘right’ first,” he states.

AI and CRM
Technology provides pros and cons to the customer experience. When implemented correctly, AI improves automation and produc-tivity while strengthening the customer experience.

We continue this article with a part two, which looks at current integrations of CRM and AI as well as a look to the future. Read more in “Current and Future Uses for AI within CRM.” SW

Click here to read part two of this exclusive online series, AI and the User Experience.

Oct2019, AI Applied

Comments are closed.