Every founder I know who runs cold outreach has the same conversation with themselves around month two. The campaign isn't working. The subject lines must be off. The CTA might be wrong. Maybe the timing. Maybe a different opening line. They iterate for another three weeks. Reply rate stays under 1%.
The problem usually isn't the email. It's that the people reading it have nothing to do with the buying decision.
Most cold email lists are built on the simplest filter the tool offers — "job title contains Manager OR Director OR VP, company size 50+." That gets you a list. It doesn't get you the right list. Most of those people would say no even if your offer was perfect, because they aren't the ones who say yes to anything.
The Four People Who Decide Whether You Get a Reply
Walk into any B2B sale and there are four roles you encounter. They look different at every company, but the functions repeat.
The champion. Someone who feels the pain you're solving. They aren't necessarily senior. They're the person who, if you email them, will say "yes, this is exactly what we need" and then start fighting internally to get it bought. Champions are how deals actually close at small companies. Without one, your email is a cold pitch hitting an empty inbox.
The decision maker. Has authority to spend money. Often the founder or COO at small companies, a department head at larger ones. They won't research vendors. They sign off on what their champion brings them. Emailing them cold without a champion is almost always a waste — they delete it because they don't have the context to evaluate it.
The blocker. IT, legal, finance, or procurement. They aren't trying to buy anything. Their job is to say no to things that create risk. Emailing them about a marketing tool isn't going to land. Worse, you can lose deals where the champion was warm because the blocker pattern-matched your email as spam.
The end user. The person whose day-to-day actually changes. Their opinion matters at companies where end users have a voice (modern teams, design-led companies) and matters very little at companies where they don't (most traditional industries). They almost never have budget authority, but they can become a champion if you reach them right.
The mistake most cold email programs make: they email everyone in the "manager and up" filter and treat them as identical. They aren't. Each role wants a different message. Each role responds to a different angle. Sending the same email to all four guarantees three of them ignore you.
Why Most SMB Cold Email Targeting Is Broken
Apollo, ZoomInfo, Hunter, every tool in this category sells you the ability to filter by job title, company size, industry, geography. Sounds great. The problem is those filters describe the wrong thing.
Job title is a label, not a function. A "VP of Sales" at a 12-person company is the founder running the entire sales process himself. A "VP of Sales" at a 500-person company is a layer of management who hasn't touched a customer in three years. Same title, completely different relationships to the problem you're solving. Same filter pulls them both into your outreach. Your reply rate is meaningless when it averages across two completely different audiences.
Company size has the same issue. A 50-person company in pharma operates nothing like a 50-person company in trades. The filter pretends those are equivalent buyers. They aren't.
Geography is the one filter that's usually accurate, which is why every cold email program over-relies on it. "Companies in New Jersey" is useful. It's also, by itself, no closer to a buying audience than "companies that exist."
The honest version of B2B targeting: you have to identify the actual function of the person at that specific company before the email goes out. Not job title, not company size — function. Are they the champion who feels the pain? The decision maker who needs a champion's recommendation? Someone else entirely?
That research takes 5-15 minutes per prospect when done well. LinkedIn profile review, recent posts they've engaged with, the company's recent news, signals from the prospect's content (are they writing about the problem you solve?), tenure markers (are they two months in and looking for wins?).
A small team can do this for ten or twenty prospects a day. Which means a typical SMB sending 200 cold emails a week is sending 90% of them with zero personalization at the function level. Mystery solved on why reply rates sit under 1%.
What "Right Person" Actually Means
A right-fit prospect for a B2B small business outreach campaign is someone where three things are true at the same time.
They feel the pain you solve. Specifically. Not "they're in marketing so they probably want marketing tools." Their company has shipped 15 product updates in the last quarter and you can tell from job postings they're trying to scale demand gen but haven't hired for it. That's signal. "Senior Marketing Manager" is not signal.
They have a path to the budget. Either they sign off themselves, or you can name three people upstream they'd influence. If neither of those is true, the email is pure information transfer at best.
They're contactable in a way the algorithms don't punish. They have a real LinkedIn, their company has a real domain, your sending infrastructure isn't already burned with their inbox provider.
You can find this signal at scale, but the work is genuinely an order of magnitude harder than running an Apollo filter. Every prospect on a properly-built list comes from cross-referencing four to six data sources. Most SMB outreach skips this entirely because there's nobody to do it.
The Execution Gap: Knowing Isn't Enough
Here's the part that gets glossed over in every "improve your cold email" article. Even if you know exactly who to target and what to say, you still have to do it. Every week. For every prospect. Forever.
A founder running outreach manually hits a hard ceiling somewhere between 30 and 100 personalized prospects per week before something else falls apart. Past that, you either hire an outbound team (around $8-15k/month for a competent SDR) or you go back to mass unpersonalized email and pretend it still counts.
The reason most SMB cold email plateaus isn't that the work is intellectually hard. The right approach is well-known. The work is volume-hard, and execution stops the moment you get busy with everything else. Run a campaign for three weeks, get pulled into client work, miss two weeks of outreach, lose the momentum. Restart. The cycle is what kills the channel, not the technique.
This is where AI-driven outbound finally has something useful to contribute. Not "ChatGPT writes the email for you" (that hasn't moved reply rates much). The actual leverage is upstream — classifying every prospect by likely function, surfacing the actual pain signal, matching the message to the role. That's the work that took ten minutes per prospect manually. AI can do it in seconds, at volume, without skipping weeks because the founder got busy.
A One-Page Plan You Can Run on Monday
If you take nothing else from this post:
1. Pull your last 200 cold email contacts. Run them through the four-role test. How many were champions? Decision makers? Blockers? End users? Most teams find 70%+ were blockers or non-champions and didn't realize it.
2. For each company you're targeting, identify 2-3 functions. Not 2-3 titles — 2-3 functions. The champion is usually a senior IC or department lead pushing change. The decision maker is whoever signs off. Email both, not the same person twice.
3. Stop treating job title as a filter. Treat it as a hint. Use LinkedIn activity, company signals, and recent role tenure to confirm the function before the email goes out.
4. Personalize at the function level. A champion email leads with the pain and how it gets solved. A decision maker email leads with proof and ROI. Same product, different framing.
5. Track reply rate per function, not per campaign. A 1.2% blended rate could be 5% on champions and 0.1% on blockers. Measuring the average tells you nothing.
If you're doing this manually and you're a small team, you'll cap between 30 and 100 prospects a week before it falls over. That's the constraint. Not your copy.
The Version That Doesn't Require You to Do All This
theKrew is built around exactly this problem. The Market Researcher agent classifies every prospect by likely function before the campaign goes out. The Copywriter writes a different email for each role — champion-framed, decision-maker-framed, end-user-framed. The Data Analyst tracks reply rate per function and shifts the outreach budget toward what's converting.
It's not magic. It's the work above, run continuously, by software that doesn't have anything else to do.
$99/month is the entry tier. The reason it's that price isn't because the work is small. It's because once the system exists, it scales with no marginal cost — the same agents running outreach for one founder run it for the next thousand without operational overhead.
Bottom Line
If your cold email reply rate is under 1% and you've spent six weeks trying new subject lines, the bottleneck isn't the email. It's who you're emailing. Fix the targeting first. Personalize at the function level second. Pick your A/B test winners on something better than gut feel third. The copy is the last thing that matters, not the first.
The leads who don't reply aren't telling you to write a better email. They're telling you they shouldn't have been on your list in the first place.