The rise of LLMs like ChatGPT have sparked excitement across industries, and sales is no exception. On paper, the idea of having an ever-available, tireless assistant to brainstorm ideas, draft conten
The rise of LLMs like ChatGPT have sparked excitement across industries, and sales is no exception. On paper, the idea of having an ever-available, tireless assistant to brainstorm ideas, draft content, and handle repetitive tasks sounds like a dream. But in my opinion, at its best, today LLMs are a productivity multiplier. They can churn out personalized email drafts in seconds, create value proposition summaries tailored to specific customer personas, or even simulate prospecting calls for sales training. Imagine the time saved when a tool can take raw data—like customer demographics or industry trends—and transform it into polished messaging ready to be deployed. No more staring at a blank screen or recycling stale templates. ChatGPT injects freshness and speed into the process, enabling salespeople to focus more on building relationships and closing deals rather than slogging through administrative tasks. But itʼs not just about saving time; itʼs about expanding horizons. LLMs have the ability to think laterally, connecting dots in ways that even experienced sales professionals might overlook. It can suggest additional angles for pitches, uncover insights from complex data, or generate questions that spark meaningful dialogue with prospects. In a profession where differentiation is often the deciding factor, such creativity can be the edge that sets a salesperson apart. However, with great power comes great responsibility (as we know from Spiderman 🙂🕷). One of LLMʼs most glaring limitations is its dependence on the quality of input. Ask it a vague or poorly framed question, and youʼll receive a vague or irrelevant answer. This means salespeople must learn the art of crafting precise, context-rich prompts. Itʼs a skill that isnʼt immediately intuitive and can take time to master. Without it, the toolʼs potential is diminished, and its outputs may cause more confusion than clarity. Then thereʼs the issue of tone and nuance. Sales is a deeply human activity, one that thrives on empathy, intuition, and the ability to read between the lines. While LLMs excel at producing grammatically correct and logically structured responses, they often struggle to replicate the subtlety and warmth that characterize authentic human interactions. An AI-generated follow-up email might tick all the technical boxes, but it can easily miss the emotional mark, leaving prospects feeling disengaged or even alienated. Sometimes even small mistakes and errors make a positive difference. This make clear that it needs careful review and editing to ensure the message truly resonates. Confidentiality is another challenge that cannot be ignored. Feeding proprietary or sensitive client information into an open AI platform like ChatGPT poses significant risks. In industries where trust and discretion are paramount, such missteps can have serious consequences. Organizations must tread carefully, leveraging secure, internal AI systems or anonymizing inputs whenever sensitive data is involved. The toolʼs utility is undeniable, but so too are the risks of using it recklessly. Some Obvious and 5 not so Obvious Ways to Use LLMs like ChatGPT A great starting point is asking LLMs to summarize a dense sales deck into key takeaways for specific personas. Imagine youʼre pitching to the CTO of a retail company. Drop your deck into the AI and say: “Summarize this sales deck into three key value propositions for a CTO in the retail sector.” In seconds, youʼll have a distilled version that cuts through the noise and hones in on what matters most. This prompt is a lifesaver when time is short, but beware—itʼs only as good as the clarity of your original material. If your deck is overloaded with jargon, the output may mirror that confusion. Think of it as a sharp knife—it needs a steady hand to wield effectively. When it comes to follow-ups, few things are more tedious than drafting emails that strike the right balance between friendly and professional. Instead of staring at a blank screen, you can tell ChatGPT: “Create a follow-up email template for a prospect who downloaded our whitepaper on supply chain optimization.” The result? A polished, structured email thatʼs almost ready to send. For instance, it might generate a line like: “I noticed you downloaded our whitepaper on streamlining logistics—are there specific challenges youʼre trying to solve?” While itʼs a huge timesaver, donʼt trust it blindly. AI emails can sound stiff, so sprinkle in a touch of humanity to make it sing. Advice your LLM on which tone to be used (and which not). Objection handling is a cornerstone of any sales strategy, and ChatGPT shines when you ask it to prepare for tough questions. Try this: “List 5 common objections for cloud migration services and craft responses to address each one.” You might get answers like, “Iʼm concerned about the upfront costs,” with a response template explaining cost-savings over time. This prompt is like having a brainstorming partner who never gets tired. Still, the AIʼs suggestions can feel generic unless you customize them with your unique selling points. Some more and not so obvious ways to use LLMs in Sales Imagine analyzing a LinkedIn profile to tailor your outreach. Feed the profile into ChatGPT and ask: “Analyze this LinkedIn profile of XYZ, CTO at a healthcare startup, and suggest 3 personalized ways to approach him.” It might respond, “John frequently posts about scaling cloud infrastructure—highlight your experience with healthcare-specific scaling challenges.” This prompt opens doors to hyper-relevant messaging but relies on the quality of the LinkedIn profile. If itʼs sparse, the results will be, too. For competitive analysis, a LLM can save hours of research. Letʼs say youʼre up against Salesforce in a pitch. Ask it: “Draft a competitive analysis comparing our AI-powered CRM to Salesforce based on these features: pricing, user-friendliness, integrations.” It could highlight how your CRM offers more intuitive integrations for mid-sized businesses at a lower price point. The results are a fantastic foundation for your pitch deck but shouldnʼt replace deeper, hands-on research. Practicing for calls can feel awkward, but ChatGPT can simulate conversations. Try this: “Simulate a prospecting call with me as the seller and a VP of Procurement at a logistics company as the buyer.” The AI might push back with questions like, “How does your solution integrate with our existing systems?” While itʼs a fantastic tool for practicing delivery and overcoming objections, it lacks the spontaneity and unpredictability of a real human. Use it to refine, not predict. Keeping up with industry trends can be exhausting, but AI can help you craft hooks based on whatʼs happening. For instance, ask: “Generate a list of 5 trends in the fintech industry that we could use to create a compelling hook for our API integration services.” You might get, “Rising demand for real-time payment systems” as a key insight. Itʼs like a brainstorming buddy with a finger on the pulse, but keep in mind AIʼs knowledge cutoff dates—it might miss the latest developments. Finally, for those looking to expand customer accounts, cross-selling prompts can be a revelation: “Write a playbook for cross-selling cloud storage solutions to existing customers who use our data analytics platform.” The AI might suggest tactics like bundling storage with analytics upgrades or offering discounts for multi-product contracts. While the ideas are often clever, they may lack the nuance needed for large, complex deals. Treat it as inspiration, not a blueprint. It is my deep conviction that Generative AI isnʼt here to replace your intuition, charm, or hustle—itʼs here to amplify them. Use you prompts in sales wisely, and youʼll find that GenAI doesnʼt just keep you competitive. It could give you the edge youʼve been looking for. But it also could be a frustrating exercise, depending on the quality of your prompts, your expectations and the quality of the underlying LLM.