AI slop has infected cold outreach - here's how I spot it

digitalnomad

I've noticed buyers are now filtering cold emails the same way they scroll past LinkedIn fluff. No one's saying it out loud, but we all know the telltale signs after two lines.

"Loved your recent post about scaling revenue ops..."
"Quick question..."
"Thought this might be valuable..."

Then it's just generic advice that could apply to any company with a website.

Here's the kicker - by old standards, some of these emails are actually decent. Clean copy, halfway personalised, sent to the right person. But they all share the same shape. And when everyone uses the same AI tools and outbound playbooks, that shape becomes the spam signal.

I don't think it's just about deliverability or targeting anymore. More and more buyers can smell the template before they even read the offer.

What still gets my attention? One specific thing that proves a human actually looked at me or my company. Not fake personalisation like "I saw you're hiring" - that's just another template. I mean a weirdly specific observation, a sharp one-liner, or something visual/custom that clearly wasn't pumped out of a prompt and mass-produced.

For me, the smartest use of AI is for research and finding the right hook - not for writing the actual message. Let AI dig up what matters, then have a human add the bit that feels real.

Even when a real person does the research, if it feels like the insight is just a setup for a pitch, I'll still hit delete. A genuine observation helps, but it has to feel like a conversation, not a trick.

analyticsjunkie

now? Mate, that horse bolted back when ChatGPT first dropped. My inbox is a graveyard of generic LinkedIn drivel.

nightowl

yeah, the shape is the giveaway now.

even if a real person did the research, i’ll still delete it if it feels like the research is just a setup for a pitch.

one real observation helps, but it has to feel like a conversation, not a sales trick.

ranktracker

Yeah, I've been seeing the same. even folks with what should be a solid few years of experience are sending out messages that read like they were copy-pasted straight out of ChatGPT. What gets me is the lack of distinctiveness - the sentence structures, the overused transition phrases, the generic "value proposition" that says nothing specific about the recipient's site. I've been digging through my log files lately and it's the same pattern: a flood of outreach that all sounds identical. Makes you wonder if they're actually writing with AI or just rewriting themselves into a homogenous mess trying to sound "professional". either way, it's losing impact fast.

nightowl

I've noticed the same pattern. In outreach, tone decides everything. Most AI defaults to being polite, which kills response rates. The trick is teaching it to mirror the specific market you're after. For high-ticket or B2B, a touch of edge cuts through the noise. Training matters far more than the tool itself.

bigcoffee

I've run enough cold campaigns to know that deliverability isn't even the top problem anymore. Buyers have evolved a sixth sense for spotting templated outreach. It's not just AI slop - any automated sequence lacks the one thing that actually converts: genuine human effort.

High personalisation? Most people confuse "Hi {FirstName}" with personalisation. Real personalisation means referencing a specific project they're working on or a challenge they've posted about. Takes 5 minutes per lead. That's manual.
Open rates mean nothing if the reply rate is <0.5%. I track reply-to-open ratio as my North Star. Anything below 2% means the message is garbage, regardless of copy.
The fundamental error: cold campaigns try to shortcut relationship-building. You're asking for a commitment without investing anything. That window is closing fast.

If you treat someone like a number, they treat your email like spam. The setters who will survive are the ones willing to go back to manual, thoughtful outreach - even if it means fewer touches per day. Quality > volume, and the numbers back it up.

socialbutterfly

Completely agree. The problem is not that AI writes badly, it is that everyone ends up sending messages with the same structure, so the template itself becomes the signal to ignore. I have found AI much more useful for research and pattern spotting than for drafting the final outreach, because the last 10 percent of human judgment is usually what makes the message feel real.

neonlights

Honestly, this tracks perfectly with what I've been seeing in cold outreach metrics.

Activation rate for AI-generated sequences? Absolute floor. People are numb to it. The irony is we're now competing against robots who learned their pitch from other robots.

Someone in another thread was complaining about the same thing - got hit with an AI outreach that was "personalised" with LinkedIn data. The open rate was fine, but reply rate? Zero. Because the signal-to-noise ratio is so bad now that real buyers actively filter out anything that smells automated.

The most unhinged part is watching companies automate their "human-first" outreach with GPT wrappers, then wondering why pipeline velocity drops. You're basically running a robocall campaign against robocall fatigue.

that "think piece" about AI killing creativity? guarantee it was A/B tested for headline engagement and optimised for scroll depth. The call to action was probably generated by the same model that wrote the complaints. Fully automated circle jerk.

For growth teams, the lesson is brutal: if your outreach doesn't feel like a real human actually thought about the recipient, you're just adding to the slop pile. Activation isn't going to improve until we stop treating personalisation as a checkbox.

We're not back. We're just generating more noise to drown out the noise.

gmbpro

I get 5-6 of these 'quick questions' per day now. awful.

growthgeek

That AI slop has infected hiring too - had a recruiter send me a personalised message that was clearly churned out by GPT. No shame in using tools, but when the "personalisation" reads like a generic template, it's just noise. at least in outreach you can track reply rates. In recruitment, they're just guessing and wasting everyone's time

brandvoice

I've noticed this too. The template smell is real even when the research is solid.

One thing that worked for me was ditching text entirely for initial outreach. I started sending quick personalised videos instead (like 30-45 seconds max). Just me, talking to camera, mentioning something genuinely specific about their company or a recent post.

It's harder to scale, which is actually the point. The barrier to entry filters out most of the noise. And when someone sees a face instead of another wall of text with their name mail-merged in, the pattern interrupt is massive.

I still use AI for the research bit like you mentioned. Finding recent hires, funding rounds, product launches, whatever. But the actual outreach is just me recording something that sounds like I'm talking to a mate, not pitching.

Response rates went from around 3% with well-crafted emails to consistently above 20% with video.

The downside is you can't send 500 a day. But honestly that constraint has made our pipeline healthier because we're only reaching out when there's an actual reason to.

What kinds of specific observations have actually landed for you when you've sent text outreach?

splittester

'Genuinely curious', 'real talk' - I could list dozens. Each one a flat, lifeless mask.

tagmanager

Oh, the single-channel spam issue is exactly it. Everyone's blasting the same AI-generated email into the same inbox - of course it feels like slop. I've seen the same thing happen with SMS when brands just fire off the same automated text to everyone on their list.

What actually works is a multichannel approach where each touchpoint feels human and distinct. A LinkedIn note referencing something specific they posted, a short email referencing that same post, then - and this is where I see the best results - a personalised SMS with one clear question. The channel shift itself grabs attention. We sequence these through Overloop too, and reply rates hold up because prospects feel like they're being spoken to as individuals, not batch-processed.

The tool handles the coordination so you can focus on relevance, not volume. But if the sequence is the same AI template across all channels? You're just multiplying the slop.

adcraft

Actually, that's been happening for months. The real issue isn't the tool, it's the lazy prompts.

gmbpro

What's your current reply rate baseline? and which specific personalisation data points are you tracking? Without a benchmark, any "AI slop" is just noise