1000 cold emails, 2 replies - what's broken?

gmbpro

Four months into a sales/biz-dev role, targeting US distributors for a manufacturing company. Roughly 1,000 cold emails sent manually via Gmail, one by one. Total replies: 2. Stuck and looking for outside perspective.

Here's the breakdown:

What I've tried: Pulled strategies from various online sources (YouTube, forums, blogs). Problem is my manager's advice often contradicts what I read online. His main directive: grab the prospect's website URL, drop it into ChatGPT, ask it to personalise and write the email. I've done that consistently - still only 2 replies.

How I send: Every email goes out manually through my personal Gmail inbox. No tracking of opens, bounces, or spam placement.

A few things I'm wondering:

Is manual Gmail sending inherently broken for cold outreach? No deliverability feedback loop at all.
Does "paste link into ChatGPT and let it write" produce genuinely personalised copy, or is it just templated fluff with a company name?
At ₀.2% reply rate, is the problem the copy, the list quality, the deliverability, or a combination of all three?

For context: international outreach (different country from the US), commoditised product category, manually researching and emailing distributors.

What am I missing?

splittester

Four months of that silence must feel brutal. but honestly, the core issue here is how you're presenting your brand to the world. sending from a single inbox is like trying to host a gala from a broom cupboard - it just screams "I'm not resourced." Even one other salesperson doing the same can poison the entire domain's reputation. The same goes for marketing blasting the same article to every contact on your list, it's noise, not nurture.

as for GPT doing your copy... its output has all the warmth of a spreadsheet. i've built a two-step rhythm instead: first, a quiet reconnaissance phase to gather real intelligence on the recipient - not scraping, just noticing what matters to them. Then I feed that intel into Claude with a very specific prompt that breathes cold email principles into something that feels like a handwritten note from a thoughtful colleague.

The hard truth? you won't get traction without investing in proper resources. AI API credits, a reputable sending tool, separate domains with warm inboxes. That's the infrastructure that lets your voice actually reach someone without their spam filter screaming "block."

chatbox

I've been in the weeds with email deliverability and outreach for years - GTM, tracking, the whole pipeline - and I've seen this pattern a lot. A few things jump out from your numbers.

Are you sending from a custom domain or a free Gmail account? If it's the latter, you're basically shouting into the void. Gmail's spam filters are aggressive for unsolicited bulk sends. Also, without proper tracking - opens, bounces, click rates - you're blind. It's like running a GA4 property without debugging. You can't optimise what you can't measure.
AI-generated copy is a trap for commoditised outreach. I've tested this myself. Most language models churn out something like "Hi {first name}, I noticed you do {generic industry thing}..." - and that personalised line rarely connects to what you're actually selling. The recipient spots it instantly. It's like using Screaming Frog to crawl a site and then writing a meta description that says "We provide services." No hook, no reason to click. For personalised cold outreach, you need real research - maybe pull data from Ahrefs or GSC about their site, then tie it directly to a specific problem your product solves.
Your assumption of interest is probably wrong. Most cold outreach works off a flawed logic: "They're in this industry → this problem exists → they want our solution." That's a gigantic leap. You'd be better off scraping a targeted list of people who've actually shown intent - blog readers, webinar attendees, or even referrals. Without that, you're just spraying. And without a decent email infrastructure (MX records, SPF, DKIM, DMARC, proper warming), your replies will stay at zero.

I'd bet it's a combination of all three: weak list, weak copy, weak delivery. You need to treat it like a technical audit - isolate variables, test one at a time.

tagmanager

I feel you - that must be incredibly disheartening after all that effort. You've got a few things working against you here, and it's probably a mix that's crushing your reply rate rather than just one issue.

First off, you absolutely need to verify your leads before sending anything. If they're bouncing, it wrecks your deliverability and domain reputation - and once that's damaged, it's hard to fix. then there's the copy issue. Pasting a link into ChatGPT and letting it write the email? That's a fast track to spam reports, not replies. So now you're dealing with two problems: poor targeting/copy and a rising spam rate that's poisoning your sender score.

With a reply rate that low, I'd bet you've got all three problems on your hands: deliverability, copy quality, and list targeting. You can't fix one in isolation - you have to address each head-on.

Start by cleaning up your deliverability - get your emails verified, warm up the domain if needed. Then do proper research on your prospects. After that, write a simple, relevant opener - one line that shows you've done your homework, followed by a genuine question to start a conversation. You don't need a sales pitch, you need a hook that says "I see you, and I've got something worthwhile to ask."

Hope that helps, mate. It's tough out there, but small fixes can turn it around.

liamcopy

I've been where you are.

1000 emails. 2 replies. That hurts.

You're bleeding time doing it all manually.

Try Gmass. It tracks opens, clicks, link visits.

Then set a follow-up trigger a few days later.

See who actually looked.

Manual outreach? That's a full-time job just finding companies and scraping emails.

Automate that part.

Free yourself up to actually call the people who opened.

"Hey, did you see my email? What did you think?"

Listen. Learn. Adapt.

That's how you turn 2 replies into 200.

emailwiz

I remember when I was in that exact same spot - hundreds of cold emails sent, maybe two replies that were probably automated. It's soul-destroying. What changed things for me was really getting the distinction between top-of-funnel (TOFU) and middle-of-funnel (MOFU) emails.

TOFU is about sparking curiosity, not selling. MOFU is where you actually start talking about value and next steps. Mixing them up means your emails land in the wrong mindset. Try segmenting your outreach: first email just opens a door, second email adds context, third email makes the ask. that one shift pulled my reply rates from near zero to something that finally felt like progress

prpro

The real issue here is almost always surface-level personalisation and lazy targeting. Dropping a website into ChatGPT usually coughs up something like "I see you distribute X products" - every prospect has heard that a hundred times. The model can't reliably pick up on what actually makes that business tick: buying triggers, specific pain points, the stuff that matters. The output reads templated even when it technically isn't, and that's what kills the response rate.

wordweaver

Honestly, 1,000 cold emails for 2 replies? That's not a strategy, that's a spam campaign with extra steps. I'd bet my lunch your targeting's the problem, not your copy. If you're pitching the wrong people or the pain just isn't screaming urgent, no amount of "hey [first name]" is gonna save it. Time to go back to the list before you blame the words.

cyberpunk

80% of your effort should go into lead quality, 10% into infrastructure, and 10% into copy. That's the fix - most people flip it and wonder why nothing lands. go back through your list with that in mind

leadgen

Ah, I've been there. Four months of grinding and barely two replies is brutal. When I first started cold emailing, I made many of the same mistakes. A few things that helped me turn it around:

First, check the baseline. What reply rate did the person before you get? If your industry doesn't respond well to cold B2B emails, maybe you need a referral-based approach instead.

Personalisation is key, but AI is garbage in, garbage out. Real personalisation means using signals from your leads - are they hiring? complaining? just fired a similar vendor? Did your manager actually review your email before you sent it?

Don't just ask GPT to "generate an email". You need warmth, a human tone, a clear value prop, and it must be short. Avoid the em-dashes that AI loves. I find it works best to have two layers - let AI set the tone and draft, then you go back and colour it a bit yourself.

How fresh are your signals? If you're just matching firmographics and the data is weeks old, that's your P0 problem. List discovery should happen on a cadence, not a one-time dump.

Last thought - can you speed this up? Even a 0.2% reply rate, if you can do it faster with tools (check with your boss), let the tool handle evaluation while you focus on referrals, community outreach, networking. Don't spend another four months grinding manually.

sophiasaas

Honestly, that reply rate screams issues across the board. i'd step back and audit three things: deliverability, list quality, and copy. biggest red flag is no tracking before rewriting emails - you've got to confirm they're actually landing in inboxes, not spam folders. a colleague once sent 500 emails with zero replies, turned out 80% were bouncing off filters. Once we fixed DKIM and SPF, reply rate went from 0% to 3% within two weeks. Worth checking before touching the copy.

conversionninja

The struggle is real. I have done cold emails and calls and it was really frustrating but eventually it gets better.

wordweaver

Oh honey, that ChatGPT personalisation is so obvious it might as well come with a flashing neon sign. Before you even touch the copy, check if your emails are actually hitting inboxes. Gmail's been quietly throttling cold outreach harder than a bouncer at a sold-out gig. List quality and deliverability kill reply rates way faster than any subject line tweak.

roichaser

Using GPT for cold emails is lazy, and the results prove it. I've seen too many people treat outreach like a volume game-blast 1000 templates and pray. That's not strategy, it's spam.

The pros who actually book meetings? they're spending hours tweaking subject lines, personalising each segment, and testing hooks. Newbies skip the thinking part because it's hard, then wonder why nobody replies. Took me about 500 emails and zero replies to learn that lesson the painful way

brandvoice

Honestly, your numbers aren't that unusual. Cold email has become brutally hard over the last couple of years. Most of the "playbooks" doing the rounds online were written back when the maths still stacked up.

Using a manual Gmail account for this is probably the worst approach. You've got no proper authentication for outbound, every spam report teaches Gmail's filters that your address is unwanted, and cold emails to strangers generate those reports by default. Half your messages might be landing in spam without you ever knowing.

That ChatGPT-written opener based on their website is also done for. Procurement people see that exact pattern several times a week. It might look personal and thoughtful in your sent folder, but on their end it's just another one of thirty generic AI emails.

To be honest, a 0.2 % reply rate to international distributors in a commoditised category is about what cold email returns now when nothing is actually broken. Inbox providers got much better at filtering, recipients just ignore this stuff, and the margin for error has basically vanished.

What works these days is the opposite direction: getting them to come to you. Content and SEO so distributors find your company, LinkedIn conversations where there's at least some human signal, trade associations where they've already opted into the category. Slower, yes, but you stop fighting the inbox.

dailygrind

Mate, 2 replies out of 1,000 sends up every red flag I've got. I've been on both sides of this - scaling outbound for a FinTech and cleaning up messes in startups where the founder just blasted a list they bought. Here's where I'd start ripping into it:

1. List quality - are you actually targeting decision-makers?
Run your list through something like Apollo or even a simple LinkedIn sales navigator export. I see so many people hammering generic job titles ("CEO" or "Marketing Director") at companies that don't fit their ICP. If your offer isn't solving a problem they wake up thinking about, you're noise. Validate intent: have they downloaded a relevant report, or visited your pricing page? if not, you're gambling.

2. Deliverability - are you landing in spam folders?
This is a silent killer. You can have the perfect copy, but if your domain hasn't been warmed up properly, or your SPF/DKIM/DMARC records are a mess, Gmail and Outlook are binning you before anyone reads. Check with a tool like GlockApps or just send a few emails to your own Gmail and look at "View original" headers for Authentication-Results. Also, how many recipients on the same domain? if you're blasting 50 people from @company.com in one go, you're getting flagged.

3. copy doesn't matter if steps 1 and 2 are trash.
But since you mentioned only 2 replies, I'd still audit your subject line and first sentence. If it's "Hi [Name]" with zero personalisation beyond the token, people smell the template. use a tool like Mailmeteor or Lemlist to see open rates - if they're under 30%, you've got deliverability or subject line rot.

Fix these two things first - I'd put money on list quality being the bigger culprit. When i scaled our cold outreach at the last gig, switching from scraped lists to intent-based data (e.g., people who visited our blog or matched a technographic filter) tripled replies within two weeks. Worth a shot

chatbox

Honestly, two replies from nearly a thousand emails screams deliverability or targeting before it's ever a copy problem. I've seen this exact pattern way too often.

Here's the usual process I'd run through:

Check your domain's email setup. Manual Gmail sending can work for tiny volumes, but if you're using your main domain without SPF, DKIM, and DMARC records - and you have no visibility on bounces - you're almost certainly hitting spam folders. I'd pull a Screaming Frog crawl on your domain to check DNS records, then verify in Google Admin Toolbox. This is step zero.
Scrub your list hard. A hundred bounces tank your sender reputation fast. Use a tool like NeverBounce or MillionVerifier before you send another batch. Cut every invalid or risky address. then segment by territory, distributor type, and actual product fit. Sending a generic list to everyone is burning credits.
Targeting over copy. If you're relying on AI-generated personalisation, it usually reads as "custom" without saying anything real. Pick one concrete observation about the distributor - a product line they recently added, a region they expanded into - and tie your offer to a specific outcome. Distributors respond to a focused angle, not "hey, want to chat?".
Test a tiny batch with a plain-text email. Strip out every image, link, and formatting. Single CTA. Short paragraphs. send to, say, 20 highly targeted contacts. if your reply rate doesn't jump to at least two or three from that group, the problem isn't volume - it's the list or the message.

quickest fix is always isolating the weakest link. With that silence at 1,000, it's almost never the copy.

nightowl

The issue is you're leading with the email, not the problem. Think of it like stacking inventory before you've validated demand-you're just burning capital.

what works is reverse-engineering the pain. I'll feed a bunch of competitor podcasts, interviews, and industry gripes into a LLM and ask it to rank the recurring frustrations by intensity, immediacy, and willingness to pay. then I cross-reference that against what I actually solve. Nine times out of ten, the gap is obvious-and my messaging writes itself.

surprising how many people skip that step. They're so focused on volume they forget to ask whether the problem even hurts right now

backlinker

honestly, the Gmail issue is probably your biggest bottleneck here. Sending from a standard Gmail account gives you zero visibility into deliverability-you've got no way of knowing whether those emails are landing in spam or vanishing into the ether entirely. the ChatGPT thing works as a rough starting point, but you need to rewrite it heavily. those AI-generated emails are painfully easy to spot, and most distributors will bin them on sight.

What kind of distributors are you targeting? that changes the entire approach

pixelperfect

I've been there - that feeling when you send hundreds of emails and hear nothing back is brutal. From what you've described, it's almost certainly a deliverability or list quality issue before you even worry about copy. Manually sending from a Gmail account based outside the US to American distributors with zero tracking is a recipe for hitting spam folders. And that ChatGPT-style personalisation often comes across as hollow - people can tell.

Try stripping it back: send just 20 plain emails, each with one sharp, genuine reason they should care. see if the reply rate changes. Sometimes less really is more.