Stop using AI for cold email
It's the #1 reason to stop using ChatGPT and other AI to write cold email outreach: The "I did my homework -- let me show you..." strategy.
It's a spam trigger. Has been for years. And it's getting worse. All thanks to "artificial intelligence" tools focusing heavily on this spam pattern.
Making, and then stating, observations which prove you've done homework is now considered spam.
Does it still work? Yes but there's nuance involved. In this post I'll explain.
First, you've likely come to believe clients demand "show me you know me." Or that they are thinking, "PROVE you know details about my problem if you want my attention." While true, we need to think critically.
Historically, from 1999 to 2015 making observations about an organization or contact was enough to stand out in inboxes and earn replies. It seemed personal, genuine... like it took TIME for the sender to construct.
“Caught your recent talk at [Industry Conference] — your take on [Topic] stood out.”
“Heard your team on the [Podcast Name] episode about [Trend] — interesting perspective.”
“Your panel at [Event] on [Theme] really sparked some thoughts for me.”
“Interesting expansion into [New Market] — smart timing.”
Today, this strategy is all but dead. Because AI is automating this process to death. Recipients (including you!) have caught on.
Worse, picture a future where Buyer Agents (representing buyers' priorities and curiosities) will do battle with "intelligent" Selling and Marketing Agents.
Full stop. We're already living in this world!
Thus, the only way to break through is to be allowed entry. This requires you to do primary research — on and off the Internet. For example, speaking with people at conferences, on the phone.
Getting into inboxes (and earning meaningful conversations) demands genuine, human-created, AI assisted insights. A string of mere observations is no longer enough. Especially in Enterprise sales where cycles are longer.
AI personalization is not personal at all
AI is getting better at writing highly personalized emails — proving the sender knows multiple "business things" about prospects. Research-based AI tools are stringing together multiple observations in an attempt to automate sincerity, relevancy.
“Saw you’re hiring SDRs in Dallas.”
"Congrats on your recent funding — exciting times!”
“Loved your Q2 earnings call. Big moves.”
But that’s not personalization. That’s pattern-matching. Prospects know it. Their anti-spam tools do too.
What used to signal "effort" now screams "automation."
Fake sincerity. False familiarity. Nothing kills trust or spam bins you faster.
And it's all at your fingertips now via artificial "intelligence." Blech. Sorry, not sorry!
Increasingly, buyers see piles of "I know X about you" emails in inboxes. Worse, these "I know you" themed emails are destined for quarantine where Buyer Agents will automatically examine them for relevance against a set of buying or interest criteria.
Buyers will tell (or are already telling!) their AI Agent what to let through that interests or (directly) serves them.
Everything else is trashed.
So far, so-called AI tends to avoid breakthrough ideas, says Maury Kosh.
"It just regurgitates the most statistically probable response based on everything it's seen before. That's why every AI-written sales email sounds the same, every AI marketing campaign feels generic, and every AI strategy looks like something your competitors could have written.
If your goal is to be average — or if you are such a poor writer that AI does it better — then AI can be a good tool.
For AI to be really useful, you have to already know how to think critically and use AI to stimulate ideas, not write complete messages without any human oversight for you," says Maury, who's known as "The Prospect Whisperer."
Stop looking like a knucklehead
Stop using AI to write outreach emails. Because AI has learned, and feeds back, all the persuasive garbage being used by the masses of sales knuckleheads.
People who end up in spam bins.
Start here: Everything you're being told about cold outreach is wrong. All of it. I don't like speaking in absolutes. But it's true. AI is making you worse.
- Be specific about your offer
- Appear worthy of their time
- Look credible
- Share value
- Be helpful
AI will claim the way forward is to leverage these ideas. And to be personalized, relevant by making observations: Showing your homework.
Then it offers the same nonsense it offers everyone else you're competing with for inbox space — from in and outside of your sector! Phrases like:
“Congrats on your Series B raise — exciting times ahead!”
“Saw you landed on the Deloitte Fast 500 list — impressive growth.”
“Heard you on the [Podcast Name] episode — loved the part on leadership.”
“Happened to notice your CEO’s interview in TechCrunch — interesting comments.”
“Impressive release of [Product] at [Conference] last week.”
“Loved your Q2 earnings call where you discussed growth strategy — big moves.”
“Saw you ranked as a top 10 employer in your region.”
Insights are better
What does seem to, occasionally, stop people long enough to consider engaging are unique insights.
It's easy to observe something about a business or contact. It's difficult to create a unique insight using those observations.
It takes non-automatable effort.
Example: One can assemble an insight through a process including (but not limited to);
- conducting three non-sales calls with your decision-maker's direct report (primary research)
- drawing genuinely unique conclusions using a 10K and pouring through investor call transcripts
- attending trade shows in person, creating unique observations
Generate observations, then, begin putting pieces of the puzzle together to form critical thought — a daring conclusion.
Right or wrong, these insights can cause decision makers to pause — thinking, "How does he/she know that about us/me?!" and/or "That took time to figure out."
These reactions recognize the effort taken to create the insight. Also, noticing X and Y... and connecting it to fact Z... then, considering a market condition... could lead to a conclusion the prospect has not realized BUT SHOULD.
Your research is spam
Bottom line: Showing you've done homework is a show stopper. Especially when your email comes from outside the prospects' organization. AI is only making the practice less effective to engage in worthwhile conversations.
Same with congratulatory spam. Congratulating based on some kind of trigger even or observation is an established pattern.
I'm not saying don't ever do it. There are no absolutes. Only tradeoffs. It depends on your situation. It depends on what you're selling and to whom. Overall, recipients are on to the ruse.
"Show me that you know me" was a great rule to live by when reaching out. Yes, it still is. But it's getting complicated.
You need (human generated, AI assisted) insights -- not mere observations.
This patterns automated spam. Faked sincerity. Don't be led to believe people cannot see through it!
Don’t observe and spit it back at them. Instead, make a few calls, do the primary research. Put a few puzzle pieces together; then, offer an original take.
Read the 10K, the earnings call transcript and talk to three insiders. Then, draw a conclusion only YOU could draw.
Offer critical thought.
Not “I know this about you.”
Instead, an insight you discovered through research -- that AI could not.
There is a useful, helpful way to use AI... in support of outreach research and many other tasks.
How do you react? What's your experience lately?