The Future of Lead Engagement is Now, Thanks to Artificial Intelligence
Carl Landers, SVP of Marketing and CMO at Conversica, illustrates how companies are leveraging artificial intelligence (AI) and machine learning in the sales process. Article by: www.martechadvisor.com
AI is everywhere. From fraud detection to online retail recommendations and self-driving cars, artificial intelligence (AI) applications are increasingly prevalent in our day-to-day interactions and business dealings. Companies are even using it to attract, nurture and retain customers.
Amazon, Google, and Apple provide voice assistants that allow us not only to check the weather but also to make purchases, update our calendars and adjust the lights. Netflix uses AI-powered recommendations to suggest the next best show to watch, airlines like Delta and KLM have deployed AI-powered chatbots to answer common questions and manage booking transactions, and Google is using AI to provide smart, automated email inbox sorting and replies for both personal and business applications.
Similarly, companies use AI to identify sales prospects, nurture customer relationships and, ultimately, drive sales. This approach to lead management is built on a traditional funnel that drives suspects to engaged leads, to Marketing Qualified Leads (MQLs) and then into Sales Qualified Opportunities (SQOs), leveraging technology to improve quality and conversion rates throughout. With AI, these companies can more effectively apply business rules and lead scoring and utilize AI-powered “bot engagement” to increase revenue. For example, one enterprise technology company uses predictive analytics to identify the leads most likely to convert and routes those directly to the sales team for follow up. The lower-scoring leads are engaged by AI-based automated sales assistants that identify the hidden gems in the lower scoring leads, boosting the MQL-to-SQO conversion rate at a much lower customer acquisition cost.
Business Rules and AI for Better Leads
On a recent webinar, Charles Eichenbaum, Microsoft’s Director of Marketing Technology & Applied Artificial Intelligence in the company’s Global Demand Center, introduced a three-pronged approach to serving up the best leads to their salespeople, including the use of business rules to filter out obviously bad information, predictive lead scoring to rank the rest, and conversational AI bots to automate outreach and conversion.
Business rules work at the top of the funnel, where a large number of inbound leads are processed. These straightforward rules scan for contacts with bad data quality, such as a missing name or phone number or an invalid email address, and weed those out right at the top of the funnel.
Lead scoring is then employed to rank leads, and AI now plays a significant role in this area. In the past, the process incorporated rules- and points-based lead scoring, assigning a point value to a prospect who downloaded a piece of collateral or an e-book or attended a webinar, for example. When that prospect passed a certain point threshold, the lead would be sent to a sales rep for follow-up. However, although a prospect may have taken a number of actions, like downloading content and attending webinars, that activity didn’t necessarily correlate with sales success. As a result, the cost to process and follow up with those leads isn’t always commensurate with the value returned in terms of sales pipeline: the same follow-up resources are extended on poor propensity leads as on high propensity ones. In an organization with large inbound lead volume, a better way to score and follow up is needed, as not every single lead should be contacted by a human.
Companies are thus employing more sophisticated means to determine high propensity to buy leads, including predictive lead scoring solutions. In predictive scoring, machine learning algorithms are used to correlate available information with past success and assign a score or grade to indicate which are most likely to buy. The inputs to the scoring model include demographic, firmographic, technographic, behavior and intent signals, and the output is the degree of correlation between these attributes and those of an ideal prospect. Based on the predictive model scoring, leads can then be ordered in priority for human follow-up, ranked best to worst. But the organization still needs to decide which should get the human touch and which should be nurtured until their score increases.
Even with the most sophisticated predictive models, high propensity to buy individuals can still end up with low scores, for example, if they use a personal email address, leave out their title or company name, or don’t engage with the highest scoring content. And this is the perfect opportunity to apply automated, yet personalized, outreach to identify the hottest leads that aren’t already in front of the sales team.
Companies are deploying conversational AI assistants to process a large number of leads, more than any human-powered organization can, by automatically reaching out to determine interest and readiness to buy. The virtual assistants can help nurture a given lead, from a prospect to an MQL, by asking the right questions, and leads can be readily reprioritized for follow-up so that no one falls through the cracks and every contact made by the sales team is on the most profitable prospects. If a lead responds to the AI assistant by saying, “please contact me right away,” that lead gets moved to the top of the queue, but if the prospect says, “maybe next month,” it becomes a lower priority and gets automatically reengaged at the time requested. This approach ensures that the sales organization is constantly supplied with leads that are most likely to convert.
Lead Scoring Enhancements
Today’s most sophisticated marketers are using machine learning and AI to continually enhance the lead scoring process. They have evolved from a rules-based approach, using experience and human judgment to determine a “good quality” lead, to a data-driven approach that models historical behaviors of high-converting leads and a lead scoring environment that uses real-time outcomes to continually tweak the scoring model. The traditional lead scoring process is being improved in the following ways:
1) Stack ranking: leads are re-ranked based on their probability to convert, resulting in higher MQL quality
2) Scoring models: scoring models are self-learning and optimized by segments such as product area and geography
3) Re-scoring: leads are re-scored in real time after each new activity, providing sales with the most updated information
4) Standardization: lead scores are the same across different product groups and sales teams, giving everyone a common foundation and understanding
In real-world terms, using AI to optimize lead scoring increases the likelihood that they will convert. Where AI really shines is by achieving a fine-grained and nuanced understanding of customers’ responses and interest level in order to feed good information into the system. This delivers an improvement over previous rules- and intuition-based approaches.
Conversational AI Assistants: Outreach Experts
Using AI-powered assistants for lead outreach enables greater coverage of the range of scored inbound leads because the assistant can engage one lead or a thousand leads with little incremental cost. So, low scoring leads that would have been just nurtured or recycled can now get personalized and human-like outreach to identify those ready to engage with sales. AI assistants never give up, never have a bad day, and never go on vacation, so every prospect gets the same consistent message, personalized to them. These assistants are also used with dormant, cold or recycled leads that were engaged at some point but lost interest. When the cold prospect replies to the assistant’s check-in message, that lead is dynamically re-prioritized and assigned to a rep, reintroducing that prospect into the sales funnel at the time that’s right for them. There is no wasted effort by the sales team on leads that aren’t ready and no unnecessary outreach to leads when they aren’t ready.
Better Results and Best Practices
The sum total of these AI-powered sales efforts is a substantial increase in leads converting to MQLs and then to qualified opportunities by sales. In the Microsoft example, the boost ranges from 30 to 60 percent depending on the product line. Many of these AI-discovered opportunities are significant and would have otherwise gone unnoticed.
When thinking about how to deploy AI to boost your lead conversion rates, start by determining your desired business outcomes and then identify where AI can be deployed to help. Capabilities improve every day, but there’s no need to wait. Companies are getting real value today and learning how to better integrate AI into their sales and marketing practices now. You don’t need to hire a data science team or build a product from AI frameworks. As many companies have discovered, even narrowly focused off-the-shelf AI solutions can drive a big productivity increase. The most important takeaway is to get started now. Improving lead engagement and conversion is an age-old challenge that is now dramatically easier with the application of robust AI solutions. Automating some portion of routine business conversations with AI is not only within reach but a reality today.