If you thought AI (artificial intelligence) had peaked in sales, you’d be wrong. In fact, AI is at the beginning of a long curve that promises to impact sales as much as big data did. In spite of some formidable deployment challenges, Forrester has estimated a third of sales leaders will invest in AI tools that recommend actions for reps to take – and a fifth will invest in AI to aid first-line managers in coaching of reps.
COVID accelerated both virtual selling and AI. “The pandemic was a catalyst that accelerated buyers and sellers to a new norm,” according to Liam Halpin, VP of sales North America at LinkedIn. “The belief that B2B buyers won’t trust us without face-to-face is no longer valid.”
Online sales interactions can be automatically recorded – generating the rich data AI needs to work properly. Natural language processing (NLP) can convert spoken dialogue to text, and AI can look for buying signals from prospects or analyze the effectiveness of reps’ conversations. There is even software that can analyze facial expressions to look for prospect moods – even distinguishing expressions used in different countries and cultures.
For example, Affectiva’s media analytics solutions understand complex emotions and cognitive states by analyzing facial movements. Its software uses computer vision and deep learning to look for significant changes in velocity and signs of nervousness.
“AI can spot if they change tone, slow down, or look confused,” notes sales author Victor Antonio. AI analysis of sales calls can be used by managers to coach, by trainers to teach, and by sales leaders to change sales strategies.
“Call analysis can tell us how effective the rep is,” explains Gartner consultant Greg Hessong. “We can use that for training in calls, product demonstrations, and principal meetings.”
By interpreting recorded sales contacts, AI can automate entry of data into CRM systems – saving reps time and effort and building the database needed by AI’s analytical algorithms to improve sales, Hessong adds.
And this is just one instance of AI’s potential for task automation that can help reps. Any repetitive task that involves numbers, text, voices, or pictures that can be translated to text – and that is done by clear rules – may eventually be sped up and done cheaply by AI algorithms.
Supported by sufficient data, AI should also aid up-selling and cross-selling strategies – showing the best items to offer and best paths to offers. For instance, how long after an initial sale a rep should wait for “decision fatigue” to wane before adding an up-sell option.
Unlike B2C, B2B markets have complex solutions and need human intervention. Communication with a lead or prospect may need to switch from digital to human at times – and vice versa. This switch may occur back and forth several times, with AI handing off to the rep, the rep handing back to marketing, then AI passing back to the rep for the final moves. So when should a company make that switch? With sufficient data on past switching practices and sales results, AI can help make those decisions.
AI, if fed by rich historical data, can also help reps optimize prices by doing the complicated tradeoffs between maximizing revenue and profit on the one hand and maximizing probability of close on the other. No human can really do these tradeoffs as precisely – especially when they may depend on so many variables, such as product, customer profile, geographical location, season of the year, even day of week or holiday occurrence. “AI can do it in the blink of an eye,” Antonio says.
Hessong agrees. “AI can tell us which is the most effective price range to win the deal, what is the highest margin price, and how price-sensitive the prospect is so we can be careful in negotiations or proposals.”
Pricing is just one element of a sales strategy. If sufficiently supported by historical data on past practices and successes, AI can advise all the way through the sales process on the best next steps to take, when to take them, what channel to use, which product features to emphasize, and how to convey them. Hessong calls this “prescriptive analytics,” because it prescribes the best means of achieving favorable results.
Prescriptive analytics could be used pre-sale (to improve the odds of a win) or post-sale (to maximize cross-sells and up-sells or to minimize account churn).
Vendor Neutral consultant Steven Wright agrees: “If you have data on all your steps, you can analyze the effects.”
Aygun Suleymanova, VP marketing at SetSail, says a modern sales dashboard can track many sales behaviors – and improve and predict performance. “Today, a sales dashboard is a tool to help you manage the performance of every individual on your remote team at just the right level.”
Behavioral dashboards help sales managers change course or deploy and estimate the impact of new strategies. “If you can track every sales rep’s prospecting and closing activity across key behaviors, you can make more informed decisions on who needs more reinforcement in which area,” Suleymanova argues. An individual rep coaching dashboard enables managers to track every team’s key behaviors using AI. Any tool that improves current reps’ sales success can of course be used to improve onboarding of new reps and training of all reps.
Wright warns that AI will do better at increasing conversion rates and revenues than at speeding up sales cycles. “Faster sales is a chimera. Customers take a certain amount of time to make decisions, and you can’t speed that up.”
AI can help sales managers and sales leaders make more accurate sales forecasts – a major aid to CFOs and operations execs. On top of relating current prospect data to historical data on pipeline progress to yield robust, statistically-based forecasts, Antonio argues AI can sharpen the probability of a close by using Internet data on prospect actions – for instance, visiting a rival’s Website. But Wright cautions that regulators may eventually hinder that kind of tracking across the entire Internet.
Wright says AI should be able to spot when many individuals from the same company visit the seller’s Website – helping to form an integrated picture of the potential buyer and its interests.
“AI can be a huge support to sales forecasting,” Hessong argues. “Managers often rely on reps, who use their intuition to forecast the probability of a win, which is often not accurate.” AI exploitation of well-kept prospect data, in light of historical pipeline conversion data, can remove that subjective element and yield better sales forecasts. And by forecasting lifetime values of probable wins, AI can convert a short-term sales forecast to an important part of a long-term revenue forecast.
Hessong also foresees AI improving territory allocation. Traditionally, territories have been allocated by simple geographic methods, or to equalize numbers of accounts. To ensure these territories are valid and accurate, however, AI can adjust them based on equalizing tier 1 accounts or other industry characteristics.
Or AI could recognize that some reps are more experienced and better than others and allocate territories based on maximizing the company’s total revenue – given the ability of the best reps to sell more to the best prospects.
AI can itself help maximize each successful account’s lifetime value by analyzing its retention probability and what steps may be taken to improve retention. “Predictive analytics can help us understand customers – who is likely to churn and what are the churn risks,” Hessong explains. “Then you can focus on them immediately.”
Suleymanova says AI can also help determine the best commission structure in sales compensation. SetSail software can collect sales tracking data essential to evaluating different commission structures. “First, you examine and analyze past structures that have been enacted. What has worked best in the past? What challenges arose?” Then SetSail monitors key metrics, sets monetary rewards for top reps, and motivates the team. “Monitoring these key metrics will keep you up to date on the processes that are working and not working for your reps,” Suleymanova says.
For customers, the big benefits of AI should be much more productive interactions with sales reps, less frustration in the buying process, and much speedier movement through the initial (easily automated) stages of learning about the product – even if the tough parts of the buying decisions still take time and major effort.
Does all this sound too good to be true? Well, short-term, it certainly is. AI is data-hungry, and companies must build the unified databases of historical sales process steps and lead and prospect characteristics to support powerful capabilities.
And Scratchpad founder Pouyan Salehi stresses that AI, like any new sales tool, must simplify and lift burdens from reps’ work – not distract and burden them.
Gartner recommends companies focus on two important and narrow use cases for initial AI implementation. “Be very specific about what you expect AI to do,” Hessong advises. “It might be guided decision-making to help reps make better decisions, or it might be task automation for day-to-day activities.”
In terms of maturity, feasibility, and business value, Gartner ranks potential AI use cases as:
That’s a whole lot of help from AI.