Use Smart Machines to Acquire High-Value Customers

Harnessing AI to acquire customers with high lifetime value
Data & Artificial Intelligence

For decades, the marketing and sales functions have automated processes to streamline tasks such as lead management, order processing, and routine customer service. Now, with artificial intelligence (AI) platforms that sense and infer — even think — on behalf of humans, marketers and salespeople are discovering tools that have the potential to help them make better, more informed decisions about the type of customers they acquire.

Read how AI is pushing the boundaries of innovation for the marketing and sales functions, particularly those activities designed to attract and retain the most profitable of customers.

Sales and marketing organizations are harnessing
AI to help them acquire customers that are most
likely to produce high lifetime value.

Not All Buyers are Created Equally

The art of qualifying prospects has become as much science as art. Thanks to marketing automation from providers such as Marketo, Oracle, and Salesforce, business leaders have been able to automate the qualification of prospects. While far from perfect, users of marketing automation report that these tools’ lead-scoring capabilities (which measure the propensity of a lead to close) have enabled them to acquire the type of customers that are most likely to become repeat, loyal buyers.1 And, as study after study confirms, loyal customers buy more, are more profitable, and tend to refer others like them. All of this reduces selling costs and, more importantly, helps the business attract and retain the type of customers that are more likely to generate high lifetime value.

B2B Sales and the Human Touch

In many cases, marketers report that sales automation is able to bring purchases to a close with no human intervention. This is especially true in consumer sales. But, when it comes to B2B selling, longer, more complex sales cycles require relationship building from account executives that conduct in-person, individual sales calls tailored to meet the needs and buying practices of large enterprises. Moreover, these calls require support from expensive pre-sale resources. 

AI sales assistants use analytics and natural language processing to silently record personal selling in the background. Applying artificial intelligence tools to the organization’s aggregate sales call activity (which is stored in the Cloud) helps individual account managers differentiate which of his or her prospects will likely go on to become loyal, and which will not — by recommending a next-best course of action. For example, scores engagement based on how long prospects respond to a question, while also measuring how excited (or skeptical) they were about the question itself. For example, one company’s evaluation of its AI-analyzed calls revealed that the following questions were the most effective at sustaining positive engagement on an initial sales call:

  • “Which processes work well with your onboarding approach?“
  • “Do you mind sharing what doesn’t work well?“
  • “What would you do if your CEO doubled your budget?“

The analysis also revealed that, while rapid-fire questions raised excitement, they also contributed to stressful conversations that hindered the productivity of those interactions. 

How Do AI Sales Assistants Work?

While marketing automation streamlines sales workflows, it is dependent upon timely data entry from each individual salesperson. By enabling AI and analytics to extract insights from customer dialogue and responses, salespeople can improve the desired outcomes of their next-best actions without the burden of entering their own (and often bias) interpretations from a sales call. And, like any machine learning system, AI assistants get smarter and more reliable over time. 

Currently, sales managers in organizations such as AdRoll are using insight from technology to coach and train the global salesforce. Another B2B seller, FarmLogs (which helps farmers monitor and measure their crops), uses insight from the software to quickly provide coaching to the salesforce after analyzing daily sales calls. Alex Terry, CEO of Conversica, uses AI tools to provide his sales team with questions and account management behavior that is instrumental in developing trusted relationships. Marketing automation provider, Marketo, uses the technology to accelerate its own sales cycles.2

Accelerated Sales Cycles and Qualified Leads

Figure 01: Customer Engagement Leads in AI Integration 

A recent survey reveals the type of AI applications organizations have integrated with or plan to integrate with their existing solutions.

Base: n=80 Gartner Research Circle Members, excludes “not sure”
Q05: What types of artificial intelligence applications has your organization integrated or planned to integrate with existing application(s) or solutions(s)?
Source: Gartner, A Chief Data Officer’s Guide to an AI Strategy, Figure 4; 26 July 2017.

In the second edition of their book, Sales Growth: Five Proven Strategies from the World’s Sales Leaders (which distills interviews with 200 sales leaders from the world’s most successful organizations), authors Thomas Baumgartner, Homayoun Hatami, and Maria Valdivieso de Uster report that those sales organizations that have been early pioneers of AI have experienced an increase in qualified leads and appointments of more than 50 percent, sales cycle reductions of 65-70 percent, and overall cost reductions of 40-60 percent. The research is an early indicator that AI in sales can help generate more qualified leads, while additionally executing their associated sales cycles in less time. These same organizations report that AI has helped them acquire customers that have gone on to become loyal.3

Engaging moments are the key to closing deals faster. If you’ve ever sold anything to an enterprise or participated in a large B2B sales pursuit, then you’re familiar with that moment in the sales cycle where a buyer excitingly and willingly delivers the type of information you need to generate a winning proposal. Such engaging moments are preceded by the right question(s). In a Gartner survey, customer engagement was the leading application for AI integration into existing solutions.4

To help the salesperson generate an engaging moment, AI sales assistants regularly sift through hundreds, even thousands, of questions from the organization’s sales calls to identify those that are most effective at encouraging the prospect to deliver the type of information required to qualify a deal. Engaging moments are also great indicators of a good opportunity. Sales cycles that generated an engaging moment were 60 percent more likely to usher a deal to closure than those that did not.

Companies that use AI in sales report: 50% more qualified leads and appointments, 40-60% overall cost reductions,
and 65-70% reduction in sales cycles

AI Helps Marketers Learn from Customers

Since the introduction of tools such as NPS (net promotor score), research has shown that the quick resolution of customer issues is often the fastest way to convert an unsatisfied customer to a brand advocate.5 Once again, scalability is the challenge, especially for large enterprises that face thousands of customer interactions every day — or even every hour. 

Consider a global gaming company that handles tens of thousands of support tickets per day after releasing a new game. Or, a business software company that needs to analyze thousands of live chat logs to detect customer experience gaps. In situations like these, marketers retain customers by using text analytics to gather insight from surveys, reviews, social media posts, and call center transcripts. But, training these systems to understand a company’s domain-specific lexicon requires weeks, even months, not to mention the complications brought on by slang, misspellings, sarcasm, and emotion.

Finding the Needle of Knowledge in the Haystack

One AI tool from Luminoso (spun out of MIT’s Media Lab) has been designed to solve the problem by drawing upon text analytics, natural language processing, and machine learning to detect trends, insights, and experience gaps from a broad range of customer interactions. Rather than training systems to learn each new client’s domain expertise and surface nuances, it uses a proprietary mix of semantic networks (e.g., an open-source knowledge graph it calls ConceptNet) to understand the meaning and relationships of words. In one instance, the software spotted a couple of complaints about a broken payment process and enabled the organization to generate a fast fix, curbing what could have been tens of thousands of dollars in lost revenue. 

As we’ve seen with other AI solutions, the benefit is speed. Luminoso enhances analytical depth to quickly decode unfamiliar words and shorten what data scientists call time-to-insight. Additionally, it supports 13 languages, including English, Spanish, Japanese, Korean, Chinese — even Arabic. Luminoso users — from Hulu, Roche, and Scotts Miracle Grow to Johnson & Johnson — report using artificial intelligence to classify customer support tickets, understand key drivers behind NPS, and prioritize which issues to address first. 


By 2020, 25% of customer service and support operations will integrate smart technology, virtual customer assistants across engagement channels.

Mike Rollings and Thomas W. Oestreich A Chief Data Officer’s Guide to AI Strategy. Gartner, 2017

Personalize Thousands of Simultaneous Sales Inquiries

Another AI tool known as Amelia, from IPsoft, parses natural language to understand customers’ questions in record speed (handling up to 27,000 conversations simultaneously and in multiple languages). When installed in a call center, Amelia’s algorithms absorb (in a few seconds) the same knowledge that would require weeks or even months for human staffers to memorize. When Amelia is connected to customer relationship management systems, she becomes even more productive.  

Amelia (named for American aviator Amelia Earhart) is described by her inventors as a pioneer in cognitive computing, a type of AI that learns by experience or instruction, much the way a human would. Hence, Amelia reads and digests the same training information as her human counterparts, but in a matter of seconds and without intensive programming. She also learns on the job by observing the human interactions of her coworkers and customers to independently assemble her own process map of what is happening. Amelia then stores her knowledge in the cloud, drawing upon it to determine how to resolve similar situations on her own.6

Gartner forecasts that autonomics-based managed services and cognitive platforms could easily reduce the cost of information technology (IT) solutions by 60 percent by automating repetitive tasks currently handled by humans. It could signal good news for IT workers anxious to give mundane tasks to technology, thereby freeing them to devote time to efforts of higher strategic value. Some experts say these AI tools could also radically alter today’s outsourcing models.7

By automating repetitive tasks, cognitive platforms could reduce the cost of managing IT solutions by 60%, thereby freeing technologists to devote more time to high-value initiatives.

Our Recommendations

Plan now for the impact on your recruiting strategy 

AI sales assistants are proving to be quite effective at predicting how a sales cycle will unfold. The move is promising for large sales organizations that are often burdened with long purchase cycles that are expensive to support. Such transformation is good news for sellers but has many implications for the hiring and managing of account representatives, changing the way business leaders recruit and train their people. While a sales-driven, empathetic personality remains critical for relationship building, future salespeople will need to be skilled in understanding and interpreting data generated from AI assistants — skills that will need to be integrated into job descriptions and recruiting efforts. 

Position AI as augmented intelligence

In customer support, artificial intelligence can parse frequently asked questions while accurately predicting customer inquiries. As these customer support solutions develop more learning skills, they will develop hyper-fast response rates and help organizations resolve issues before they become a crisis (or use knowledge to avoid them in the first place by proactively supporting issues that typically slow new customers down). Marketing and salespeople can then use this knowledge to develop satisfied customer relationships faster.

Artificial intelligence is not a panacea

At this point in its evolution, AI cannot make judgments around ambiguous situations or completely understand the nuances of human emotions (skills that are important in selling). Nevertheless, salespeople will need more analytic skills to derive insights out of data and use them to run territory campaigns. Much of what AI delivers will require sales and marketing to interpret and understand it. These teams will also need to manage escalation when AI is stumped. Managers may be responsible for executing the escalation, but sales and marketing will need to navigate the overall process. 

Build business cases around problems that AI can solve

Marketing leaders, such as the chief marketing technology officer (CMTO), that are responsible for technology roadmaps should quantify the cost of activities such as selling to unqualified prospects, engaging in longer-than-average sales cycles, or supporting low-value customers — all activities that AI tools are equipped to tackle. On the revenue side, the CMTO can work with marketing and sales functions to quantify the bottom-line impact of selling to customers that are more qualified and that go on to become loyal, repeat buyers. Armed with the potential impact on both revenue and cost, marketing and sales can set goals to test and pilot emerging AI platforms.

  1. Bisconti, Ken. “How to Acquire the Right Customers.” IBM Watson Customer Engagement, 2015.
  2. Gibbons, Serenity. “10 Most Innovative Movers and Shakers in Sales Leadership.” Entrepreneur Magazine, 2017.
  3. Baumgartner, Thomas, Homayoun Hatami, and Maria Valdivieso de Uster. Sales Growth: Five Proven Strategies from the World’s Sales Leaders, 2nd Edition. Wiley, 2016.
  4. Rollings, Mike and Thomas W. Oestreich. A Chief Data Officer’s Guide to AI Strategy. Gartner, 2017.
  5. MacDonald, Stephen. “Why Customer Complaints Are Good for Your Business.” SuperOffice, 2017.
  6. Ankeny, Jason. “Meet Amelia, the AI Platform That Could Change the Future of IT.” Entrepreneur Magazine, 2015.
  7. Gartner Research. Top Strategic Predictions for 2017 and Beyond: Surviving the Storm Winds of Digital Disruption. October, 2017.

More Insights


49 Cities.
5 Continents.
1 Vision.