PDF Summary:The AI Edge, by Jeb Blount and Anthony Iannarino

Artificial intelligence is rapidly advancing, shifting industries and changing everything about the way we work. Some even claim that this technology will make sales professionals redundant. But in The AI Edge (2024), sales experts Jeb Blount and Anthony Iannarino argue that AI tools won’t replace sales professionals for two reasons: First, AI tools can’t replicate the human abilities that drive sales. Second, these tools need constant human oversight.

This doesn’t mean your job’s safe, though—AI may not replace sales roles, but sales professionals who effectively use AI tools will replace those who don’t. According to the authors, success lies in knowing which sales tasks to delegate to AI tools and which must remain in human hands.

This guide walks you through the authors’ key ideas. You’ll learn what AI tools can and can’t do, why the human touch drives sales, and how to effectively apply AI tools throughout the sales process. We’ll also supplement Blount and Iannarino’s ideas with opinions and advice from sales experts and AI specialists.

(continued)...

Sales Skill 3) Complex Problem-Solving

You demonstrate complex problem-solving by weighing multiple factors—such as goals, constraints, and competing priorities—and applying judgment and creativity to design tailored solutions that deliver the best outcomes. Blount and Iannarino explain that complex problem-solving drives sales by helping clients navigate uncertainty and overcome obstacles that make purchasing decisions difficult. This reassures them that you can guide them to a solution that will work for their specific situation.

(Shortform note: Research reveals that complex problem-solving methods help drive sales when clients have unique needs—when their goals, constraints, and priorities differ from those of other buyers. In these cases, there’s no established buying pattern that predicts which approach will meet their needs. Clients, therefore, evaluate how thoroughly you analyze their specific situation: The more detailed your examination of their circumstances, the more confident they feel about your proposed solution’s likelihood of success. However, clients with more common needs require less reassurance—they only need to see evidence of what’s worked for others in their situation.)

AI tools can’t solve complex problems because they can only recombine patterns from data they encountered during training. When asked to solve problems, these tools search their training data for similar situations and suggest approaches that worked in those contexts. However, when facing unprecedented problems or situations that don’t closely match their training data, these tools can only suggest generic solutions or cobble together responses from loosely related examples. While such solutions seem plausible on the surface, they rarely account for the unique combination of factors at play.

For example, suppose a client wants to purchase your software but has a limited budget and specific integration requirements. After evaluating these constraints, you might propose a phased implementation that prioritizes the most critical features first, and you might schedule additional support to ensure a smooth rollout. This solution helps the client adopt the software successfully despite their limitations. In contrast, an AI tool might suggest a standard implementation plan it has seen frequently in its training data. This plan won’t work in practice because it doesn’t account for the client’s specific constraints.

(Shortform note: Computer scientists explain that AI tools can’t solve unprecedented problems because they lack “out-of-distribution” detection—the ability to recognize when a problem falls outside their training data. They’re only aware of the data used to train them, and have no knowledge of anything that falls outside it. In other words, they’re unaware of what they don’t know, and this leads them to assume that they can solve every problem.)

Limitation #2: AI Tools are Unreliable

Now that we’ve clarified why AI tools can’t drive sales, let’s discuss why they’ll always require human supervision. Blount and Iannarino explain that because AI tools don’t understand meaning, they often misinterpret requests and generate information that sounds accurate but isn’t. Let’s explore these two flaws in detail.

Flaw 1) AI Tools Misinterpret Requests

The authors explain that AI tools misinterpret requests because they interpret language literally. When given an instruction, they follow it exactly as written—they can’t infer meaning, fill in gaps, or consider intentions or priorities. As a result, they often produce incomplete or irrelevant outputs that require extra time and correction. For example, say an AI tool is asked to summarize which sales strategies generated the most revenue from a quarterly report, but the request is simply, “Summarize this report.” The AI might produce a broad overview of figures and highlights without identifying which strategies drove results—completely missing the insight that matters.

For this reason, to provide meaningful insights or perform a helpful sales service, AI tools always need clear, detailed instructions that explicitly state the purpose of the request. For example, a more effective instruction would be: “Summarize the Q2 sales report by identifying which strategies generated the most revenue and explaining why they were effective.”

(Shortform note: Prompt engineers add insight into why explicitly stating purpose in your instructions produces better results. AI tools are trained on massive amounts of data that show many different ways to complete any given task. When you make a request without stating your purpose, the tools have no way to determine which approach matches your needs, so they select one arbitrarily. On the other hand, explaining why you need something—not just what you want—provides a contextual filter that helps them distinguish which possibilities in their training data are relevant to your situation and which should be ignored.)

Flaw 2) AI Tools Output False Information

Blount and Iannarino say that even with clear instructions, AI tools often produce false or misleading information. This is because they’re designed to generate convincing responses, not ensure accuracy. When they lack reliable data to fulfill a request, they don’t admit uncertainty. Instead, they assemble language patterns that sound factual, outputting statistics, case studies, or examples that look credible but are entirely fabricated.

For example, if asked to find a recent statistic showing how training software improves sales performance, an AI tool without access to real data might output, “Companies using training software saw a 27% increase in sales productivity,” and attribute the claim to a well-known research firm. The statistic sounds legitimate, but the study may not exist—the system may have invented it to provide a satisfying answer. Since relying on or sharing false information can undermine performance and credibility, AI outputs should never be used without verifying their accuracy against reliable sources.

(Shortform note: Researchers suggest that false or inaccurate outputs—otherwise known as hallucinations—are an inevitable consequence of contaminated training data. As previously explained, AI tools are exposed to billions of language examples during training. These examples often contain false claims, outdated information, biases, inconsistencies, or factual errors. AI tools can’t distinguish between reliable and unreliable sources during training, so they learn these flaws and perpetuate them when generating outputs.)

Part 2: Apply AI Tools Throughout the Sales Process

Since AI tools have many limitations, you might be wondering if it’s worth integrating them into your work process. Blount and Iannarino stress that not doing so will put your job at risk—you’ll miss out on the efficiency benefits they offer and eventually be replaced by those who use them effectively. However, they warn that relying too heavily on AI tools also puts your job at risk, as it may undermine the very capabilities that differentiate you from these machines. So, to safeguard your job and thrive in sales, you must create a balance—delegate tasks that AI tools are trained to excel at, and keep control of everything else.

(Shortform note: The advice to create a balance sounds straightforward, but it may be challenging to practice due to the increasing number of companies adopting an “AI-first” approach, which involves checking whether AI tools can complete a task before assigning it to employees. As a result, companies are removing the roles that enable employees to both develop skills that set them apart from AI tools and gain the experience needed to use these tools effectively.)

In this part of the guide, we’ll outline how you can achieve that balance and benefit from AI tools through three key stages of the sales process: targeting potential clients, approaching potential clients, and making a sale. Since both authors have built their careers training organizations to maximize sales, they emphasize how AI tools can support business-to-business (B2B) scenarios—where sales professionals sell to other companies rather than directly to consumers. Their advice, however, applies regardless of what you sell or to whom.

Stage #1: Targeting Potential Clients

To make a sale, you first need to target those who might be interested in what you have to offer. In B2B scenarios, this involves three steps:

1. Finding relevant companies: Identify companies that might need your offer by reviewing industry reports, marketing materials, or company announcements. For example, if you sell logistics software for manufacturers, you might look for mid-sized manufacturing companies in the process of expanding production capacity.
2. Identifying decision-makers: For each company, focus on the roles and departments most likely to use what you offer and determine who is responsible for making purchasing decisions. For example, you might seek out operations executives responsible for supply-chain performance.
3. Evaluating opportunities: Analyze the information you’ve gathered and decide which companies and contacts are worth pursuing.

Blount and Iannarino suggest you can use AI tools to automate the first two steps and save hours of manual work. You can instruct them to search industry databases, scan company websites, and compile contact lists based on roles or departments to gather all the information you need. For example, you might use this prompt: “Identify mid-sized companies with growing production capacity and list operations executives responsible for supply-chain performance.”

However, the authors emphasize that evaluating opportunities must remain your responsibility. This involves two steps: First, you need to fact-check the data and interpret its relevance. Then, you need to decide which companies and contacts are worth pursuing. If you skip these steps, you risk chasing companies that look promising on paper but don’t need your offer, or contacting people who don’t influence purchasing decisions. For example, say an AI tool flags a company that’s announced plans to expand. You’ll need to dig deeper to determine whether this expansion affects the company’s software needs, whether your offer will be relevant to those needs, and whether the listed executives have real decision-making authority.

(Shortform note: There are three frameworks you can use to evaluate viable sales leads, depending on the complexity of the sale: 1) B.A.N.T. helps you quickly determine whether a lead has a problem you can solve and the ability to work with you to solve it; 2) C.H.A.M.P. guides you in prioritizing leads with the most pressing business needs; and 3) M.E.D.D.I.C.C. provides a thorough approach for evaluating complex B2B opportunities.)

Why AI-Identified Opportunities Require Verification

Research on sales qualification in B2B contexts shows why salespeople must evaluate opportunities—and what happens when they take AI recommendations at face value.

AI tools identify structural matches—companies that could theoretically need your offer—but can’t assess situational factors that determine actual need. These tools assess company characteristics like industry sector, employee count, and technology use—to identify companies that structurally resemble your target profile. However, studies show these characteristics rarely correlate with purchasing needs due to situational factors that AI tools don’t pick up on, such as budgetary or leadership changes. Since salespeople already waste roughly 50% of their time chasing companies that won’t buy, they can’t afford to skip verifying AI recommendations.

AI tools analyze organizational charts to identify decision-makers, but can’t detect informal power structures that influence purchases. These tools focus on job titles that appear to have purchasing power, like “Director of Procurement.” However, people with such titles rarely have sole authority to approve purchases because buying decisions tend to involve six to 10 people working by consensus. Therefore, salespeople who contact only the person an AI tool identifies risk wasting their time on unviable leads.

Stage #2: Approaching Potential Clients

Once you’ve identified which companies and decision-makers to target, the next step is to make contact and quickly demonstrate why you’re worth talking to. You achieve this through three tasks:

1. Defining your approach: Review what you’ve learned about the company and the decision-maker and decide how you’ll make your message feel immediately relevant. (Shortform note: Key to defining your approach is figuring out what decision-makers already know about your offer. Seth Godin (Purple Cow) explains that this makes it easier to understand their perspective—how they judge your product. With their perspective in mind, you can frame your message in a way that connects to their existing understanding rather than trying to educate them from scratch.)

2. Crafting your message: Write a short, specific message that connects your offer to the decision-maker’s likely priorities. (Shortform note: Research backs up the necessity of crafting a short and relevant message: The average attention span is just 8.25 seconds—if you don’t engage people’s attention within this time, they’ll automatically switch their attention to something else and you’ll lose your chance of drawing them into a sales dialogue.)

3. Managing interactions: Keep track of who you’ve contacted, when, and what response you received so that each subsequent message feels timely and intentional. (Shortform note: You might find it easier to manage interactions if you use a customer relationship management (CRM) system. Many of these systems come with built-in features that reduce the need for manual tracking—for example, by centralizing contact information, recording previous messages, and automating reminders to follow up with potential clients.)

Blount and Iannarino suggest you can use AI tools to streamline all these tasks. You can instruct them to summarize company information, highlight patterns in what similar decision-maker roles value, draft messages based on tone or formality, and help you track communications.

However, the authors emphasize that communications must stay under your control. You need to make sure your messages sound like you, interpret the responses you receive, and judge how much persistence is appropriate. Otherwise, you risk sounding generic, contacting people too often, or pursuing opportunities that aren’t real. For example, say an operations executive hasn’t responded to your initial message. Instead of asking an AI tool to draft a generic follow-up, check whether the company has posted updates or investigate what project the executive is currently working on. Use that context to adjust your timing and wording—for instance, acknowledge the update and restate how your offer helps address this development.

(Shortform note: As AI technology advances, tools become better at personalized communication—the kind needed to make a good impression on potential buyers. For this reason, some businesses have found success letting AI tools handle some communication. For example, Salesforce used an AI SDR (sales development representative) to initiate conversations with 100 million leads, which led to a million deals being closed. However, other companies experimenting with AI SDRs didn’t see improved sales performance, which may be because they’re not equipped to handle sales conversations effectively.)

Stage #3: Making a Sale

After engaging clients in conversation, you then need to work toward closing a deal. This involves three tasks:

1. Tailoring a proposal: Learn about the client’s specific needs, then adapt your offer to address those needs.
2. Negotiating terms: Discuss pricing, timelines, and conditions with the client while addressing any concerns that arise.
3. Finalizing a contract: Create, review, and sign a legal agreement to make the sale official.

(Shortform note: It’s worth noting that these steps tend not to apply to business-to-consumer (B2C) scenarios, where transactions are based on convenience and speed: Businesses offer standardized products, prices, and terms so that they can cater to large numbers of customers with minimal friction. In contrast, B2B scenarios require customized proposals, terms, and contracts because clients evaluate each transaction based on the return they expect to get on their investment. In these cases, standardized offerings don’t work because they fail to address the specific needs and priorities of each business.)

Blount and Iannarino suggest you can use AI tools to support all three tasks. You can instruct them to organize conversation notes, draft proposal sections, suggest responses to common concerns, and review contract language for issues.

(Shortform note: Sales experts expand on how you can use AI tools to support the three tasks%20Loopio%3A%20Request%2Dfor%2Dproposal%20automation%20platform%20with%20AI%2Ddriven%20response%20intelligence%20tool). First, pinpoint the needs to address in your proposal by instructing an AI tool to scan your conversation history and identify your client’s specific priorities. Second, prepare to negotiate by using an AI tool to analyze all client interactions, highlight recurring objections, and identify counterarguments that have worked in similar situations. Third, accelerate contract completion by using an AI tool to track revisions and flag changes that could cause delays or introduce risk.)

While you can use AI to assist with sales tasks, the authors emphasize that you must make all strategic decisions. You need to decide what to emphasize in proposals, how to respond during negotiations, and which contract terms to accept.

For example, say your client mentions a problem that disrupts their production schedule. You’ll need to structure your proposal to emphasize how the software solves this specific problem rather than listing all its features equally. During negotiations, if the client doubts whether the software can resolve their issue, you’ll need to respond with relevant case studies or customer endorsements that demonstrate proven results. Then, if the client requests a contractual clause guaranteeing the software will resolve their production issues within six months, you’ll need to verify that this timeline is realistic before committing.

Tips for Closing

Let’s explore additional advice for strategically managing the three core sales tasks:

Tailor your proposal by asking several clarifying questions to better understand the client and their company. For example, you could ask them what their largest hurdles currently are as a company and then demonstrate how your offer could help them overcome these hurdles. Alternatively, you could ask them about their goals for the coming year and then show them how your offer can help them reach these goals.

Negotiate favorable terms by helping clients feel more certain. In Way of the Wolf, sales trainer Jordan Belfort explains that people usually waver on terms because they’re uncertain about something. You can prevent them from stalling the sale by establishing three certainties in their minds: First, that they will benefit from your offer. Second, that you’re trustworthy and likeable. Third, that you work for a company that’s reputable and reliable.

Flag contractual terms that create one-sided obligations. Legal experts warn that contracts often contain asymmetric clauses that favor one party at the expense of the other. For example, indemnification clauses may require your company to cover losses even when the client is at fault, payment terms might allow clients 120-day payment windows while requiring immediate delivery, and termination clauses might let clients cancel without penalty while locking you into ongoing service obligations. Therefore, negotiate reciprocal terms before signing—for instance, if clients retain the right to cancel with 30 days’ notice, secure the same right for your company.