Every founder I speak to right now is either using AI tools heavily, feeling guilty that they're not using them enough, or quietly suspicious that the whole thing is mostly hype. The honest answer is that all three instincts are partially right. AI is genuinely useful in sales and marketing — but not in the ways most people think, and the gap between good and bad implementation is enormous.
This is a practical guide to what actually works, what wastes time, and how to think about AI as a sales and marketing tool rather than a magic shortcut.
"AI doesn't replace the need for a clear message and a real relationship. It amplifies whatever you already have — good or bad."
Where AI Genuinely Saves Time in Sales
Pre-call research and account intelligence
One of the most time-consuming parts of good outbound is research. Understanding a prospect's business, recent news, hiring signals, and competitive landscape before a call used to take 20–30 minutes per account. AI tools can compress that to five. Tools like Perplexity, Clay, and even a well-prompted ChatGPT can pull together a useful account brief quickly — recent press, headcount changes, funding history, product developments.
The output still needs a human eye. You're looking for the insight that makes a cold email feel warm, not a list of facts to recite on a call. But the research phase is genuinely faster with AI, and faster research means more time actually talking to prospects.
First-draft outreach
AI is useful for drafting outbound emails and LinkedIn messages — but only if you treat the output as a rough starting point, not a finished message. The danger is that AI-generated outreach sounds like AI-generated outreach. Prospects have been trained to spot the patterns: the "I came across your profile and was impressed by…" opener, the three-bullet-point value prop, the "Would love to connect for 15 minutes" close. If your outreach sounds like everyone else's AI outreach, you've neutralised any advantage.
Use AI to beat blank-page paralysis and get a first draft on the page. Then rewrite it in your voice — tighter, more specific, more human. The goal is a message that sounds like it was written by a real person who actually looked at the prospect's business.
Follow-up sequences and objection handling
Sequencing follow-ups is one of those tasks that sales reps consistently underprioritise because it feels like admin. AI tools can draft the full follow-up sequence from the initial outreach — including the break-up email at the end — and do it in minutes. That means reps can spend their cognitive bandwidth on the actual conversations, not on remembering to follow up for the fourth time.
AI is also useful for preparing for objections. Feed it your common objection list and ask it to generate responses. The outputs won't be perfect, but they give reps a framework to practise with and build confidence around the difficult conversations.
Call summaries and CRM hygiene
Tools like Gong, Fireflies, and Otter can transcribe and summarise sales calls automatically, then push structured notes to your CRM. For small sales teams where CRM hygiene is always slipping, this is one of the highest-ROI AI applications available. It removes a friction point that causes reps to avoid logging calls, which means your pipeline data stays clean and your coaching conversations are based on what was actually said.
Where AI Genuinely Saves Time in Marketing
Content at scale — but not quality at scale
The most common mistake I see in AI-assisted marketing is using it to produce more content faster and assuming that volume compensates for quality. It doesn't. A blog that publishes twelve AI-generated posts per week that nobody reads is worse than useless — it dilutes the brand and trains your audience to stop opening your emails.
The right model is using AI to reduce the friction in producing good content, not to replace the thinking. If Gary has a strong point of view on something, AI can help structure it, expand it, and turn a rough voice note into a first draft. The thinking and the perspective have to come first. AI is the drafting assistant, not the thought leader.
Repurposing and distribution
This is where AI delivers genuine leverage with low downside risk. A strong piece of long-form content — a detailed blog post, a podcast episode, a webinar transcript — can be repurposed into LinkedIn posts, email newsletter copy, short-form video scripts, and Twitter threads using AI in a fraction of the time it would take manually. The source material is already good; you're just cutting it differently for different channels.
Set up a simple workflow: good long-form content in, channel-specific snippets out. Run each output through a quick human edit. That's a legitimate content operation for a two-person team.
SEO research and content gap analysis
AI tools, combined with SEO platforms like Ahrefs or Semrush, have made content gap analysis genuinely accessible to founders who don't have a dedicated SEO resource. You can identify what your target customers are searching for, what competitors are ranking for, and where there are opportunities to rank for high-intent terms — all without a six-week agency engagement. The output is a content roadmap grounded in actual search behaviour, not guesswork.
Ad copy testing
Writing ten variations of a Google or LinkedIn ad headline used to require either a copywriter or a significant time investment from a founder. AI makes generating variants fast and cheap. The creative quality is rarely extraordinary, but for testing purposes — which headline gets the click, which value prop resonates — you don't need extraordinary. You need enough good variants to run meaningful tests.
Where Founders Waste Time with AI
Mass AI-generated cold outreach
The worst use of AI in sales is using it to send more cold emails faster to lower-quality lists. The logic seems sound: if AI can personalise at scale, you can reach ten times as many prospects with the same effort. In practice, this produces exactly the kind of outreach that gets immediately deleted, reported as spam, and damages your domain reputation. Volume without precision is noise.
If you're going to use AI for outreach, use it to go deeper on a smaller list — not broader on a larger one.
AI personas and chatbots as a substitute for real sales conversations
There is a temptation to deploy AI chatbots and automated sequences as a way to avoid the awkwardness of early-stage sales. Don't. At the stage where most founders are reading this, every customer conversation is a source of product and market intelligence that can't be automated. You need to hear objections in real time. You need to understand why someone said yes — and why someone else said no. An AI chatbot doesn't give you that.
The Test That Matters
For everything AI produces — emails, content, ads, research — ask one question: would a smart, sceptical prospect think this was written by a real person who actually understands their business? If the answer is no, it needs more work before it goes out.
The founders who use AI best aren't the ones who use it most. They're the ones who use it to remove the friction from doing good work — and keep the judgment, the voice, and the relationships firmly human.