Deep Dive: Machine learning is better with (sales)people
How can B2B companies ensure their tech investments intentionally expand, rather than marginalize, their salespeople’s contributions?
In an earlier edition, I worried out loud that the universal adoption of data and digital tools would diminish contributions from employees’ very human brains and argued that B2B innovators should nurture thinking processes that machines cannot replace. Wanting to dig deeper, I listened to several episodes of the excellent podcast, Me, Myself, and AI. One episode, Less Algorithm, More Application: Lyft’s Crain Martell, helped considerably. Moreover, it got me thinking about the future of sales, and specifically the potential negative impact of the current application of artificial intelligence (AI) and machine learning (ML) to improve sales productivity. I worry that the core approach for improving sales is relentlessly reductive—replacing the application of a salesperson’s knowledge and experience when AI/ML tools can do it faster, better, and cheaper. Martell’s insights made me wonder if there was a way to deliberately expand the contribution of salespeople, perhaps by redefining traditional sales roles to align with overall business model innovations. Doing so would create new goals that can be enabled or enhanced by AI/ML, and thus achieve outcomes that are comprehensive and additive rather than reductive.
Realizing the human potential of machine learning
Listening to the discussion with Craig Martell, I learned that the incredible power of artificial intelligence is not gained by designing differentiated algorithms but by applying data and machine learning to optimize business processes against a specific business objective. I thought about how his insights might be helpful for B2B innovators and began to see both promise and problems for how every B2B company learns about its customers and the role its salespeople fulfill. I’ll explain in a moment, but first, consider what Martell has to say:
When I think about AI, I find the algorithms mathematically fascinating, but I find the use of the algorithms far more fascinating. Because, from a technical perspective, we’re finding correlations in extremely high-dimensional nonlinear spaces. It’s statistics at scale in some sense, right? We’re finding these correlations between A and B. And those algorithms are really interesting, and I’m still teaching those now, and they’re fun. But what’s more interesting to me is, what do those correlations mean for the people? I think every AI model launched is a cognitive science test. We’re trying to model the way humans behave. Now, for automated driving, we’re modeling the way cars behave in some sense, but it’s really [that] we’re modeling the right human behavior, given these other cars driven by humans. So for me, the goals of AI—I look at them much more from the humanities perspective, although I can nerd out on the technical side as well.
The power of Martell’s insight is that it is human-centric. It’s about leveraging machine learning in the pursuit of doing business as humans for humans. In that sense, followers of Martell’s vision would aim to improve sales reps’ contributions to their core (selling) and collateral (advocating for customers) responsibilities. But there is a problem. Very often, B2B companies deploy AI tools to make salespeople more productive in a reductive sense—by assuming activities that salespeople do not do well. AI tools suggest products and solutions that salespeople might forget (or neglect) to offer. AI tools suggest prices for one-off purchases or complex deals in a way that diminishes pricing based on a salesperson’s gut feeling or “rush to close” a deal. Moreover, the machine learning powering these tools makes educated guesses about how a customer will respond, getting better over time, and thus providing data-dominated customer learning. All of this improves productivity by shrinking salespeople’s contributions.
Full disclosure: I am a layperson, not an AI/ML expert. Doubtless, I am mangling the design and implementation of artificial intelligence and machine learning capabilities. But my comments are an accurate representation of how B2B leaders describe the value they perceive through the application of AI/ML to achieve sales goals. And there’s the rub. Leaders set the company’s vision and strategic metrics. Effective leadership can help an organization get the most from data technology. Destructive leadership can influence technology’s use in a limiting, undifferentiated direction, and ultimately waste capital, people, and opportunity.
And so, as an aid for leaders, it is worth considering Martell’s insights and attempt to apply them for achieving sales goals in a way that is not relentlessly reductive but instead leverages a salesperson’s knowledge and experience to achieve even better outcomes. Improving sales productivity is a laudable objective, but I would suggest a supercharged purpose—improve sales results and, at the same time, leverage salespeople to achieve strategic business objectives around transforming your business for the digital age.
Making the most of your salespeople
Martell offers three admonishments for leaders to aid in designing and deploying AI/ML tools. I’ve tweaked them a bit to align with our discussion about getting more out of salespeople’s knowledge and experience:
Data must be gathered, but how it is gathered matters. In the design phase, salespeople can add perspectives on satisfying customer needs while driving sales outcomes. Every data set must have a goal, and salespeople can help set objectives. Moreover, salespeople can provide a rich, qualitative understanding, and answer questions, including: Who are our customers? How do we serve them? How do we create value? How do we measure success?
Every organization must be re-engineered for data. How will salespeople, sales managers, sales operations teams, and sales leaders consume data? How will it impact sales process execution, territory definitions, customer assignments, and sales compensation plans? How (and when) will all roles in a sales organization intercept AI/ML tools? How can AI/ML tools be considered as digital workers and part of the sales organization?
Leaders must lead. Salespeople can help inform leaders about the dangers of AI/ML going awry through data-based bias or misapplication of the tools. Salespeople may be assigned to work on development teams, helping to build out data sets and significantly label data to appropriately indicate customer personas for the best possible machine learning. AI/ML is not magic, but salespeople can help build data sets that lead to differentiation and competitive advantage.
More than anything, Martell’s insights convinced me that organizational change, including carefully designed roles and processes, is essential for leveraging the maximum power of artificial intelligence and machine learning. This forces us to ask: What is the salesperson’s future role? What new competencies are required? How will sales processes be re-engineered to incorporate human and digital workers? Every company should consider their foresight for the future of their industry and the evolution of the company’s business model. Considering my conversations with B2B innovators, I would suggest three future roles for salespeople:
The captain of your omnichannel. Omnichannel strategies and platforms are about providing a consistent experience across every channel. AI/ML is an essential capability for directing customers to the right channel at the right time and for providing the correct information and buying functionality. At Lyft, Martell’s machine learning team offsets the limitations of AI/ML by providing human escape hatches and (literal) course corrections. Users experience these features as the Lyft app asks for confirmations and route selections. In the same way, salespeople can act as the captain of your omnichannel. Customers should be able to ask for human contact with a salesperson at any time. Salespeople should review the operations and performance of the omnichannel platform to identify problems and service opportunities, and judge performance against goals.
A referee for your communities. In another edition, I argued that community development capabilities are a missing and essential strategic capability for B2B companies. Marketing will likely lead the charge in defining and managing communities, just as marketing owns segmentation and targeting today. But salespeople must play a critical role. Communities require diversity to help enable product and service design, and salespeople can help ensure that the best customers are included and play an active role. Community interactions will provide data for AI/ML tools, and salespeople can help avoid coding bias. Salespeople can help nurture relationships within a community, solving problems, sharing stories, and memorializing progress. Cross-functional participation in communities is essential, and salespeople should not dominate. But salespeople have the people skills and customer experience to ensure every community is a standout success.
A colleague in your cohorts. I have suggested that B2B companies can better leverage salespeople’s knowledge to create customer value through a market-based approach. I offer several options beyond the existing (and largely ineffectual) approach of giving advice away for free in the pursuit of sales and share. One of those solutions is to establish cohorts—intimate working groups of customers, suppliers, vendors, and cross-functional company team members. In the cohort concept, everyone contributes knowledge and experience, and the team works toward solving a specific problem, pursuing an opportunity, or developing a program, policy, or tools. Salespeople can be part of a cohort and contribute as a colleague. Depending on the cohort’s task, salespeople may not be in charge of the cohort, but they can provide strong leadership.
Ideas for innovating B2B
The purpose of sales is and always will be to sell. Nothing in this edition suggests otherwise. But sales organizations often stand apart from their companies’ organizations to focus on the sales mission and avoid interference. The increasing use of AI/ML tools will act to more closely integrate salespeople and processes with the rest of the company. This is most certainly so if AI/ML tools are implemented in an additive, not reductive, way. Salespeople will act in cross-functional leadership roles and directly enable the company's evolution toward a modern, digital-age business model.
To help B2B innovators, I suggest several actions for including salespeople in their planning and execution of business model innovations. These are important additions to the suggestions above:
Identify new ways for customers to hire you. The best path to the future is to help your customers get to their future in the digital age. Christensen provides an invaluable framework for identifying innovations built around the concept of “getting hired” by customers. Salespeople are an essential part of your jobs-to-be-done culture.
Measure trust. Every company tracks sales, share, and profitability. Customers don’t care. Customers care about your product’s (and services’) measurable performance against promised features, benefits, and services. But in the digital age, trust matters. As business goes more and more virtual, human connections are diminished. Trust is a human condition and must be measured as an indicator of solid customer relationships.
Task your human resources department. HR is attempting to redefine itself as more consequential, helping companies survive in the digital age. You will find one excellent study by Monitor Deloitte here. I would push HR to a purpose beyond establishing a digital-age culture and managing employee engagement. I wonder if HR can own what it means to do business as humans for humans. Can HR play a leading role as an advocate for human-centricity on your innovation team?
Don’t forget about sales compensation plans. Effective sales compensation plans are designed to fit what it takes to sell the company’s products and services in the context of customer personas. Sales plans are highly leveraged, or not. Overall, sales rewards are aligned with the revenue a salesperson can create. Salespeople demand an appropriate compensation plan, enabling them to do suitable sales activities and reap the rewards of sales outcomes. But sales compensation plans are a prison if your company seeks to change sales roles and processes for the digital age. They lock in the status quo. Understand how your sales plans work today and figure out a transition plan to get them to what they need to be for the future.
Join the journey
I encourage everyone to be an active listener to the Me, Myself, and AI podcast. Let me know what you learn. Let’s discuss implications for B2B innovation and the overall B2B revolution. I invite you to join the quest. Don’t be a stranger. Share your thoughts in the comments section below. If you prefer, reach out directly at mark.dancer@n4bi.com.