Direct sellers adopt artificial intelligence across departments to improve recruitment, training, compliance and retention
By: Stephanie Ramirez
Companies are taking customer data and feeding it through a machine-learning algorithm, turning patterns into predictions.You don’t have to understand statistics and probabilities, only how the technology will benefit the company.
—Michael Bayan, CEO, DirecTech Labs
Artificial intelligence (AI) and machine learning are not new concepts to many direct sellers. The pandemic, however, has led to a sharp rise in the adoption of digital and technological advancements. Novel AI applications are opening new pathways and potential for improvement for the channel.
Face-to-face encounters have driven the social selling channel for decades; therefore, as the pandemic continues to require social distancing, executives must find new ways to train their field, retain customers and support their salesforce.
The Key to Artificial Intelligence Is ‘Intelligence’
Standard demographic data points may no longer be the best sources for evaluating what a customer wants and where they are in the buying process. Recent studies indicate that demographic data is flawed, according to Michel Bayan, CEO of software company DirecTech Labs.
“A company’s target customer might be 25-to-35-year-old women who have an average income of $50,000 a year, but that doesn’t say anything about what they want to buy,” says Bayan. “If you are making broad strokes with your reach, those things may be effective, but standard demographic data is becoming obsolete.”
The good news, according to Bayan, is that direct sellers have always been good at the “I” of AI because humans know the other humans they want to sell to, and direct sellers are excellent at customer acquisition.
“There’s a huge advantage economically and socially as direct selling companies don’t have to pay anything to acquire the customer. They pay a commission after the customer has been acquired,” says Bayan. “But, how do you keep customers and prevent distributor burnout?”
Reaching people digitally through data collected via machine learning is a sophisticated and deeply involved process. Many companies have been successful with inbound interest, but the methodology for distributing those leads to the salesforce is where AI can really add value.
“Using AI, companies can now really look at how a person came to know their product,” says Bayan. “ ‘What can I learn about this individual, demographically, psychographically through data that’s already been collected?’ Knowing these things, I can match them with a distributor who matches that predicted personality profile, so that I increase the probability of them liking each other and developing a long-term relationship.”
Teaching the Algorithm to Find Illegal Posts on Social Media
Companies who have processed online compliance violations may be sitting on a mother lode of data—data that can be used to “train” machine-learning models to recognize an improper health or income claim or message online.
Few companies have more of this core data than Momentum Factor, whose FieldWatch platform monitors the internet for many of direct selling’s largest companies. “Over the years and by working with hundreds of clients, we’ve acquired a ton of anonymized data,” says Christopher Mitchell, director of products at Momentum Factor. “With that, we’ve been able to build a huge database to train our algorithm to automatically identify what makes a claim look like a claim.”
Machine learning, Mitchell says, is critical to being able to monitor a massive social media landscape, where so many direct sellers share their opportunities and products.
“For instance, a post may contain the word ‘guaranteed,’ which by itself is not problematic, but can put a company at risk if the word ‘income’ is also in the post,” he says.
“We’re constantly training our data acquisition system to quickly sort through the millions of results and identify illegal claims with a goal of 95 percent accuracy. It saves companies an enormous amount of time and expense.”
Improving Direct Sales Training with AI
AI has recently facilitated a groundbreaking new tool for sales training. Dr. Stefanie Boyer, a marketing professor at Bryant University, is a Direct Selling Educational Foundation (DSEF) Fellow and co-founder of RNMKRS.org (sounds like “rainmakers”), the first virtual, mobile, and gamified sales competition that uses an AI-powered app to train students in the sales and closing process.
Funded in part by the DSEF, the RNMKRS app has been integrated into the intercollegiate sales competition that Dr. Boyer directs.
“Before the app, we had sales competitions at the university where we bring in employers to act as judges,” explains Boyer. “They roleplay as the buyer in a sales meeting and give the students live feedback. It makes the students incredibly nervous because they are right there with CEOs and VPs who are evaluating them on their sales process. This is the first sales class for many of these students, so it was not ideal for them to learn at this stage in such a high-pressure situation.
“Also, we were limited by the number of students who could compete, by the amount of space required and the number of executives available to judge,” Boyer continues. “So, we decided to create this tool that would allow us to expand the competition.”
She says that by using a gamified app bot to run a training course, trainees can have fun with the process and are more likely to work on their own shortcomings in order to “win” a training session.
“From a student’s perspective, someone starting out in sales needs practice because they’re going to fail often—and that’s a good thing. That is how they learn. Students need to receive coaching with feedback in order to be successful. This app gives them an opportunity to go fail in a safe place so that they can pick themselves up using feedback from the app before they go out into the real world.”
When asked how else the AI-based training would help a student be successful at direct selling, Boyer replied, “The app teaches students how to build rapport and own a meeting by setting an agenda. In order to understand the customer, it’s crucial to not only know how to ask questions, but also to be able to ask questions during a conversation to discover more. Often, clients don’t buy because we don’t ask enough of the right questions to understand their goals and challenges and how to overcome these obstacles. The pitch should be customized to the needs of the clients, and sellers should be prepared to handle pushback and ask for the sale. Our app teaches students how to do all of these things with ethics and concern for the client.”
Behavioral Profiling and Predictive Targeting
Predictive targeting, as its name suggests, predicts and recommends how to target each experience, and to whom, without any need for manual analysis.
What many companies are now doing is taking the customer data that has been collected and feeding it through a machine-learning algorithm. The algorithm can turn these patterns into predictions, allowing businesses to determine metrics such as an employee’s value to the company or what the probability is that they are in the last 30 days of their lifecycle. Using these data points, a company can put together a behavioral profile.
“Based on that profile, a business can predict whether someone is in their last 30 days, and they can also connect that person—again, using the data—with the right person in their upline, based on shared values or the influencer rating of that person, to reach out and help that individual with their next win,” says Bayan of DirecTech Labs. “In one example situation, if the company can help that person to recruit one more person, the probability of him or her leaving goes from 94 percent down to 34 percent.”
All of this prediction and profiling happens automatically. The algorithm processes data from millions of user events to create actionable data so executives can accurately predict what their conversion rate is going to be or what percentage in customer turnover to expect, or customer-specific predictions for upsell potential. The machine finds patterns that are far too granular for human analysis. All of this can lead to better performance, increased sales and improvements to a company’s valuation, adds Bayan.
Using AI for Competitive Intelligence
Arguably, many industry analysts believe it’s impossible to be successful if businesses don’t have at least some degree of insight into their competition. Previously, this has been done manually. However, companies were investing large amounts of time and resources aggregating competitive intelligence data; by the time everything was gathered and analyzed, market shifts could have occured.
According to the Marketing AI Institute, automation of competitive intelligence is one of the biggest trends sweeping across businesses today.
Much like AI for customer acquisition and retention, machine learning can be used to harvest valuable information about competitors and their customers through a process called web-scraping. Web-scraping tools are able to track a competitor’s digital footprint, both on and off their website. The process has been around for more than a decade, but only recently gained popularity as its sophistication has increased.
The customer data points can be sorted and analyzed to create valuable conclusions based on readily available information scraped from the web. There is no off-the-shelf web scraping tool, however. Web-scraping is both an art and a science that requires deft programming skills along with mathematical and scientific expertise to be successful.
The Marketing AI Institute also says that the practice, once confined to the largest and most mature companies, is now more available to smaller organizations, giving them the leverage to potentially displace industry leaders.
Individualizing Sales and Recruitment Promotions
Despite these findings and proven applications with AI, there is still hesitation from many company leaders to trust the technology, according Bayan.
“There are a lot of unique things about AI that make it not immediately intuitive, but that doesn’t mean it’s not effective or useful,” he says. “You don’t have to understand statistics and probabilities. You really only need to understand how the technology will benefit the company, and to trust the people providing you with the data.”
According to Bayan, direct selling companies traditionally plan large-scale promotions to boost sales and recruitment. He says, again, it’s not the customer acquisition piece that direct sellers struggle with, it’s the churn rate. “Many companies will hit these huge recruiting booms, and then as soon as that boom ends, they nose dive. They are constantly losing people and have to replace them.”
By shifting away from one-size-fits-all promotions and moving to what Bayan refers to as “destiny moments,” using all of the demographic and psychographic data collected, companies can reach out to people individually by sending them unique auto-generated messages.
Bayan adds that with the assistance of a predictive AI model, established using a company’s existing data, this type of unique messaging can be sent out simultaneously to thousands of people in different languages all over the world every day. A system can be developed to learn behaviors based on the data, make predictions and then send these messages at the AI-predicted moment in time that a person needs to be reached. Establishing an individualized approach using AI in this manner can reduce churn and increase recruitment, he says.
“For every 1 percent that a predictive AI model can reduce monthly turnover and keep it that way for 12 months, a company will see a minimum of 4 percent to 6 percent increase in sales by year’s end,” Bayan adds. “There’s no other metric in the company that you can decrease by 1 percent that will give you a 5 percent increase in sales.”
By using AI to customize and individualize the interactions between company and distributor as well as between company and customer, a more productive and efficient business cycle continues to emerge. The power behind these algorithms is the amount of data captured in their use.
Thus, AI technologies will continue to improve as they “learn,” leading to new avenues for both horizontal and vertical growth for direct selling companies.