How to Write Cold Emails with Generative AI (and How Not To)

How To Write Cold Emails With Generative AI

• Why use generative artificial intelligence to write sales emails?
• How does generative artificial intelligence write a cold sales email?
• Where do sales reps go wrong writing cold sales emails with generative AI?
• How to write cold sales emails with generative AI

Among all the challenges B2B sales reps face, one stands out: crafting cold sales emails. It's the key to unlocking doors—but let's face it, it's easier said than done. It’s time consuming, requires intense focus, and can slow your workflow.

That’s why many reps are using generative artificial intelligence to write cold sales emails. But while it can be a helpful tool, unmanaged use can hurt your reply and conversion rates. So let’s look at how to leverage generative artificial intelligence to scale up your outbound motions without sacrificing the personal touch.

Why use generative artificial intelligence to write sales emails?

Accelerated content production

The faster sales reps can generate personalized, contextual sales sequences, the better. Whether split testing copy, targeting different verticals, or reaching out to different personas, there are many reasons to regularly produce new cold sales email copy. generative artificial intelligence can speed up this process, enabling reps to focus on strategy, prospect research, and relationship-building. 

Seamless editing assistance

Often reps struggle to write engaging, readable copy. generative artificial intelligence can help refine and polish sales email copy, providing recommendations for voice and tone, readability, grammar, and other factors that can impact its success.

Maintaining brand voice consistency

When every sales rep is generating their own email copy, it’s hard to maintain brand voice consistency. A generative artificial intelligence model trained on your brand voice guidelines can recognize nuances in language and adjusting cold sales email copy to match.

Personalization at scale

Depending on the sophistication of your model, generative artificial intelligence can increase personalization, pulling in specific insights from prospect websites to help craft a unique response for each individual recipient. 

How does generative artificial intelligence write a cold sales email?

Whether we’re talking HubSpot’s free AI email writer, MailChimp’s Email Content Generator, ChatGPT (the O.G.), or the range of free and freemium tools out there, generative artificial intelligence is becoming more common for writing sales emails. 

But if you don’t know how the technology works, you can’t make good decisions around its use. So here’s a quick summary of how generative artificial intelligence actually writes a sales email.

First, generative artificial intelligence doesn’t “write” anything. We’re actually talking about a complex process where the AI’s neural network analyzes massive datasets to absorb the nuances of language, context, and industry-specific terminology. Rather than generate anything “new,” generative artificial intelligence recognizes patterns and is able to mimic human-like language.

For your AI bot to do this effectively, it requires several “ingredients”:

• Access to industry-specific data to analyze key terminology and concepts for your vertical
• Examples of quality cold emails—and the ability to distinguish “good” from “bad”
• Well-developed prompts that give specific instructions on output parameters
• Feedback on the output, so the model can stop the bad outputs and double down on the good

Parsing out all this information requires a lot of human input. So while generative artificial intelligence can be a major timesaver, it can only function well when there’s a human actively managing it. It’s helpful to think of AI as a co-pilot or assistant, rather than the primary “writer” of your cold sales emails. 

Where do sales reps go wrong writing cold sales emails with generative AI?

Based on data from McKinsey, only 14% of business leaders report using generative AI in their sales functions. Why so low? Among the top areas of friction are the accuracy of generative artificial intelligence outputs and employee adoption—which both point to under-management of generative artificial intelligence among their human salespeople. Here are some specific ways that can play out. 

Over-reliance on automation

Automation is helpful, except when it isn’t. Writing a cold sales email with generative artificial intelligence is never “set it and forget it.” Rather, you can use the AI to accelerate production, stepping in to rewrite awkward phrases, correcting factual errors, and guiding the model toward improved outputs with real-time, specific feedback. 

Neglecting prospect context

If you’re using a public-facing generative artificial intelligence model—whether a free option like ChatGPT or a paid option like Jasper—it won’t understand the nuances, context, and specific verbiage of your industry. Failure to incorporate these into your sales emails can result in lower response and conversion rates. 

Assuming factual accuracy

One of the fundamental weaknesses of generative artificial intelligence is a high rate of factual inaccuracies. While this shouldn’t discourage you from using AI, it’s important to know what you’re getting into. Check all factual statements the model provides so you can maintain credibility with prospects.

How to write cold sales emails with generative AI

1. Research your prospects thoroughly

A theme you may have noticed in this post is contextualizing and personalizing generative artificial intelligence outputs. To do that, you need detailed insight into your prospects. Detailed prospect research can not only help train your generative artificial intelligence models, but can help you better write and rewrite its outputs. 

2. Craft detailed, precise prompts

Generative artificial intelligence output is only as good as the prompts provided to it. So after you’ve collected detailed prospect insights, take the time to write a prompt that speaks to the value you want to highlight, the pain points you want to address, and any other attributes of the email (e.g. sense of urgency, clarity and readability). 

The good news: crafting a prompt often takes less time than drafting email copy. So in the end, you’re still saving time. But if you rush through the prompt-writing process, your email performance will start to tank. 

3. Adjust outputs in real time

Generative artificial intelligence often doesn’t get it right the first time. Rather than accept the first output, review it and correct any misalignments with your prompt, target audience, brand voice, etc.. Not only will this improve the quality of this particular output, the model can retain that information for future requests. 

4. Correct factual errors

If there are factual errors, correct them. This is where your human expertise comes into play—because you’ll only catch the errors if you know the space, industry, and product you’re selling. 

5. Personalize and contextualize the output

If you’re working with a standard, public instance of a generative artificial intelligence model, personalization will be limited. After you have a good first draft, make sure to personalize and contextualize it to individual prospect contexts. 

You should already be doing this anyway with your manually written sales templates. So simply integrate generative artificial intelligence outputs into that same workflow. 

6. Edit for clarity

Often generative artificial intelligence produces sentences that are long, convoluted, and difficult to follow. Read these outputs and correct anything that would be confusing or off-putting to the client. 

7. Read the whole thing before sending

Then, before hitting the send button, read the whole email one more time. Ensure it flows seamlessly, addresses the prospect's needs, and aligns with your overarching communication strategy. This last check helps catch any overlooked issues and ensures a polished final product.

Final thoughts on writing sales emails with generative artificial intelligence

Generative artificial intelligence can be a helpful tool in crafting sales emails—if you know how to use it well. Not only do you need to know the strengths and weaknesses of generative artificial intelligence to leverage them accordingly, but it’s important to understand your prospects so you can provide context and personalization to your emails.

If you’re unsure who these prospects are, Coldlytics can get you a list within 24 hours. We’ll collect whatever data points you specify, giving you the level of insight necessary to succeed with your cold email campaigns and outbound sales motions.