The salesperson did everything right. In fact, they did it better than they ever had before. They researched the challenges within the customer’s organization. They made the perfect connection to their value proposition. They learned all about the customer’s business dynamics, their key people, and pressing industry trends. All this work was tied into a highly contextual and targeted outreach. Best of all, the seller did very little work, instead spending a short amount of time on ChatGPT and inserting some of the learnings into their organization’s AI-assisted outreach engine.
Now the way the storyline is supposed to go, we’d expect a remarkable ending: a historical sale with a raving customer feeling like a vendor finally “got it.” But in reality, the customer, expecting to speak with a real expert on their industry and challenges, is shocked when the seller only offers a standard talk track. The stark reality is that generative AI had set an expectation that the seller could not match. Instead of having an insightful sales conversation, the customer feels embarrassed they took the call and vows to never speak with that vendor again.
Stories like this are arising more commonly as generative AI takes a foothold in business transactions. Call it what you want — shallow knowledge or a façade of expertise — but the reality is that incredible amounts of highly relevant outreach efforts are happening with customers, all couched in relevant language and insightful commentary, only to fall apart in the actual sales dialogue.
Discussion of generative AI has maintained a fever pitch since OpenAI released ChatGPT-3.5 in late 2022. But enterprise-level adoption has remained spotty. Our Q3 survey of 113 CEOs found only 9% doing anything more than small-scale pilots, and only 26% even running small-scale pilots. Meanwhile, their frontline teams are rapidly incorporating generative AI into their workflows. In one of our live trainings with 50 frontline sales professionals, nearly three-quarters told us they expect AI to play a significant role in their work in the next 12 months, and 94% want leadership to formally integrate it into their sales programs. This gap in programmatic usage and understanding is leading to customer-harming behaviors, which we believe hold more downside than more commonly cited issues such as data or security risks.
A New Way to Work
In our consulting and advisory efforts at SBI Growth Advisory, we see far too many sellers treating generative AI like just another new piece of technology. Technologies automate otherwise complex work. They’re tools that present shortcuts.
Most organizations are approaching AI in the same way that they have approached technology evaluation, purchasing, and adoption for the past several decades. They focus on minimizing risks (in this case, AI hallucinations and data privacy/security risks) and driving standardized adoption as a change-management exercise. Introduce the tool, offer training on its usage and benefits, and have champions promote the benefits and value.
Treating generative AI in this manner mistakes its true power in fueling a new way of working. Generative AI will be used ubiquitously, across software programs and outside of them as workers apply the insights and outputs generated from it into their day-to-day workflows. In this way it is more like the internet than it is a specific technology that can be purchased and deployed.
Harnessing generative AI in a sales setting requires creativity and interaction. Learning to effectively leverage it requires creativity and lateral thinking. Academics would call this a difference between “knowledge transfer” (raising awareness and merely providing resources) and “tacit learning” (learning rooted in application, experience, and practice). Leveraging AI to develop true customer business understanding, not just to acquire convenient shallow knowledge, is the key to minimizing its downside.
Encourage Creativity Over Codification
Many organizations are building libraries of generative AI prompts. Simply scrolling LinkedIn will reveal a cottage industry now selling bespoke prompts to fuel sales teams and efforts. This is not to say that thoughtful prompts are a bad idea — indeed, they can spark thoughtful exchanges with generative AI many humans had not previously thought to ask. But expecting a specific cadence of prompts limits the potential of generative AI.
In sales roles, our research at SBI has consistently demonstrated that sellers who are fanatical about preparation ahead of their sales calls perform drastically better than their peers. This includes both the amount of time they spend preparing, as well as the amount of research and learning they conduct ahead of interactions. Our latest research shows that sellers who embrace an anticipatory approach to their customers’ objections, challenges, and points of indecision have 12% faster sales cycles, and 11% larger deals than their counterparts.
Leveraging generative AI as a channel for curiosity, learning, and preparation is key. Inputting a simple prompt, based on the stage of the sales process, may yield useful information. It may result in “the hook” that captures the customer’s interest. But it decisively will not yield deeper business acumen on the customer’s situation.
On-Demand Business Acumen
An exercise our sales training division is now using with sales teams harkens back to basics of consultative selling. It involves conducting a “five whys” exercise when using generative AI for customer research. This requires understanding things like the issue a customer is dealing with, why that issue matters to their business, why they may be falling short in their current approach to it, and why this may present an opportunity to help the customer. The seller can learn to engage and collaborate with AI chat to keep asking “why” until they deeply understand the issue. This exercise does more than just quickly provide background information about the buyer and their organization; it provides massively useful context and learning for the seller. The seller can ask these questions within the confines of generative AI and come to the meeting with a point of view, instead of learning the customer’s business during initial discovery.
In this regard, generative AI has the potential to crack one of sales leaders’ longest-standing frustrations: sellers lacking business acumen and deep customer business understanding.
This use case for generative AI spans well beyond simply asking such platforms to write an email in XYZ tone or to summarize a call recording. Certainly, these are time-saving administrative tasks, but they decisively are not improving the business acumen, customer understanding, and ultimately the effectiveness of sales professionals.
Often the job of augmenting business acumen falls on sales managers. However, SBI’s latest research finds that only 44% of sellers receive regular business acumen coaching from their managers. Having thoughtful, critical discussions with generative AI can yield a remarkably detailed understanding of the business issues, their likely causes (often with specific points of evidence from that customer organization), and it can help sales teams appropriately position their solutions into this business context — all without extensive manager involvement.
A New Working Relationship
Left to experiment on their own, many salespeople have already begun to use generative AI tools in their daily sales motion, but often with very inconsistent and sometimes negative results. Even in organizations that block access to generative AI tools, we find individuals are using their personal devices or other means to access them.
As we’ve mentioned, the instinct of many leaders is to provide direction and codification when trying to drive adoption of new tools. This is particularly true in sales, a function rooted in processes, playbooks, scripts, and templates. However, driving adoption of generative AI as a new way of working — leading sellers toward creative use of generative AI, and helping them to use it as a coaching and learning aid — requires a very different approach. Here are four steps to get you started.
1. Show examples of what can be done, not what must be done.
Busy sellers need to have examples and case studies of how AI can be used effectively in their sales motion if they are to truly understand the art of the possible. Too often these are presented as use cases to replicate, rather than only a starting point to get them thinking about the potential applications. It is not desirable, or possible, to provide an “answer book.” Provide use case examples in your sales playbook or sales coaching sessions, and invite brainstorming on what could happen next. What else does this use case make them think about? How else could they adapt their workflows? And, critically, how would it help them to reach their targets? Share examples of creative chats with generative AI. Leverage it in deal reviews and “war-rooming” sessions to help surface a new angle or urgency driver for the customer’s business or industry.
2. Establish guidelines, not commands.
There is an art and science to effective prompting of generative AI. Provide sellers with guidance on how to structure a prompt to include the appropriate context, voice, and desired output. This can be in the form of “callouts” in the sales playbook, supplementary documents, or screen-shared videos showing them how to engage the tool as a learning aid. Emphasize that interacting with generative AI is not about getting “the answer” but rather about engaging in a dialogue to brainstorm and iterate to a useful response. Teach sellers how to have a discussion with generative AI: to correct it when an answer strays from something helpful or asking it to consider another approach. Direct it to sources closer to the customer, such as public filings, the customer’s website, product releases, presentations their executives have shared at conferences, or other sources from which the AI can learn.
3. Focus training on practice, not information transfer.
Sales professionals need to understand the basics of how generative AI works and considerations with using the tool, such as accuracy, confidentiality, privacy, ethical use, hallucinations, and potential overreliance. But if your training stops there (which it often does), you have likely discouraged them from evolving their way of working rather than gotten them excited about its potential.
Make the vast majority of any training session a live practice. Give them a real customer scenario and have them work through it with facilitation and group problem solving. Then give them time to apply it to one of their own accounts in the moment. Generative AI can really come to life when sellers can see it produce responses in real time that are applicable to their daily sales activities. Simple examples such as using AI to understand how their customer makes money, what the key personas care about when considering a purchase, or what objections buyers might have to their solution can be extremely educational. Giving sellers a chance to then apply these prompts to their own accounts and challenges allows them to “see the magic” for themselves.
4. Use manager–seller one-on-ones for joint learning, not feedback.
AI is changing the job of the sales manager as much as it is changing the job of the seller. Reporting, pipeline management, seller performance monitoring, and many administrative activities are all becoming more streamlined with the use of AI. Managers should introduce a regular cadence of generative AI “show-and-tell” with their sellers, sharing how each has used it over the previous weeks, what they learned, and how it has impacted their work. Rather than asking for reporting and numbers focused on adoption and compliance, focus on creating a climate of iteration and a continuous improvement mentality as the organization adapts to new ways of working.
Generative AI use is another learned skill leaders need to help their teams develop. It gathers and applies knowledge. It makes human judgment, curiosity, creative thinking, and deep understanding more important than it’s ever been. Driving deep intelligence with generative AI requires approaches that run counter to simply deploying a new tool.
