You asked AI to write a sales email. It came back polished, polite and completely forgettable. You sent it anyway. Nobody replied.

If that sounds familiar, you're not alone. Most small business owners have tried AI for sales by now. The problem isn't the tool. It's how it's being used.

Why a few prompts won't fill your pipeline

A prompt here and there can be handy. It might tidy up a proposal or suggest a subject line. But it rarely brings in new customers or gives you hours back each week.

The reason is simple. AI only knows what you tell it. Ask it to "write a cold email to a marketing manager" and you'll get the same email everyone else gets. Give it your offer, your best customers, the prospect's situation and how you actually talk, and the output changes completely.

The gap between those two results is what separates AI that helps you sell from AI that just makes more noise.

What works: finding and researching the right customers

Research is where AI earns its keep fastest. Before a call or a first message, it can help you:

  • Describe your ideal customer clearly, based on the clients you already enjoy working with
  • Build a shortlist of businesses that fit that profile
  • Summarise what a prospect does, what's changed for them recently and where you might help
  • Prepare a few sharp questions for a discovery call

What used to take an hour of tab-hopping can take minutes. You still make the judgement call. You just start from a much better place.

What works: outreach and follow-ups that get replies

Generic emails get ignored because they read as generic. The fix is context, not cleverer wording. Outreach that works tends to share a few traits:

  • It's specific. It mentions something real about the prospect, drawn from your research.
  • It sounds like you. Give the tool examples of emails you've written that got replies, and ask it to match them.
  • It's short. One clear point and one easy next step.
  • It follows up with something new. Not "just checking in", but a useful idea, example or question.

Use AI to draft, then edit like a human would. That last read is where a decent email becomes one worth answering.

What works: cutting the sales admin

The biggest time savings often aren't in writing at all. They're in the admin around selling: call notes, CRM updates, follow-up reminders, meeting prep.

AI can now connect to many of the tools small businesses already use, such as email, calendars and documents. That means it can help prepare you for tomorrow's calls, turn rough notes into a tidy summary, or draft the follow-up while the conversation is still fresh.

No new software and no technical skills are needed to get started. It's about plugging AI into the way you already work.

What's a waste of time

Just as useful is knowing where not to bother. Common traps include:

  • Mass-producing cold emails. More volume of bland messages just means more people ignoring you.
  • Prompting with no context. One-line requests get one-size-fits-all answers.
  • Sending without reading. AI gets details wrong sometimes. A single mistake about a prospect undoes the effort.
  • Chasing the perfect prompt. Tweaking wording for an hour is still an hour not spent selling.
  • Adding tools you don't need. If it creates more work than it saves, it isn't helping.

AI won't replace the relationships, judgement and follow-through that win deals. Used well, it gives you more time for them.

See it live

Rather than read about it, come and see it. I'm running a free 30-minute Lunch and Learn with Enterprise Nation on Monday 23 November 2026, 12:00–12.30 UK time, online via Zoom. I'll work through real sales challenges sent in by members, live in Claude, step by step.

Bring your lunch and your questions. You can send questions in advance to events@enterprisenation.com.

Book your free place on Eventbrite →