What Most Cold Email Agencies Get Wrong (From Someone Who Runs One)

By Mitchell Keller, Founder & CEO, LeadGrow. Managed 3,626+ cold email campaigns. 6.74% average reply rate. 2,230+ meetings booked in 2025.

The Agency That Emailed Their Own Client

True story. We were competing with another agency for a client's full gig. One week into the trial, the other agency sent the client itself a cold email about the client's own offer.

They were pitching their prospect's product back to them. From a list they were supposed to be sending on behalf of that same prospect.

Bar is low.

That's an extreme example. But the problems it reveals aren't extreme at all. They're common. Most cold email agencies operate with a fundamental misunderstanding of what makes outbound work. They have the tools, the volume, and the infrastructure. What they're missing is the thinking.

We've managed 3,626+ campaigns at LeadGrow. Booked 2,230+ meetings in 2025. Average 6.74% reply rate. We've also cleaned up after a lot of agencies that didn't work out. The patterns are consistent. Here's what to watch for.

Red Flag 1: "We'll 10x Your Pipeline"

If an agency promises specific results before they understand your market, your ICP, your average deal size, or your current conversion rates, they're selling a template.

Real outbound doesn't work that way. Results depend on dozens of variables: market size, competitive density, offer strength, deal value, sales cycle length, buyer sophistication. An agency that guarantees 50 meetings in 90 days without knowing any of these things is either lying, inflating numbers with unqualified leads, or planning to blast your market so aggressively that they'll burn your domain reputation in the process.

What to look for instead: honest expectations. "We typically see 2 to 5 qualified meetings per week for companies with $5K+ deal sizes and a clear ICP. Depends on the market. First 30 days are testing. We'll know what's realistic by week 4."

That's less exciting than "10x your pipeline." It's also true.

Red Flag 2: Maintenance Men, Not Technicians

Your toilet is flooding. Water everywhere. Two people show up.

The maintenance man says: "I do it all. Plumbing, electrical, drywall, whatever you need."

The technician does a walk-through and says: "The problem isn't your toilet. It's the pressure valve. I noticed your hookup is wrong. Does your shower run scalding when you flush?"

You say omg yes.

"I can fix it today."

Who do you go with?

Most cold email agencies are maintenance men. They offer a menu of services: email setup, list building, copy, A/B testing, reporting. They do what you ask. If the campaign underperforms, they adjust the subject line and try again.

A technician diagnoses. They look at your targeting, your offer, your positioning, your competitive landscape, and your buyer's situation. They identify problems you didn't know existed. "Your reply rate isn't low because of your copy. It's low because you're targeting influencers who can't buy. Your copy is fine. Your audience is wrong."

The difference shows up in results. Maintenance men optimize within the current approach. Technicians identify whether the approach itself is right.

When we analyze underperforming campaigns that come to us from other agencies, the copy is usually decent. The problem is almost always upstream: wrong audience, wrong offer, or wrong frame. The agency was polishing a message that was aimed at the wrong people.

Red Flag 3: Optimizing Copy But Never Fixing Targeting

This is the single most common agency mistake. They'll A/B test subject lines, rewrite openers, change CTAs, adjust email length, test humor vs formal tone. All within the same audience that isn't working.

Targeting is 80% of the result. Copy is 20%.

When you target hyper-specifically (one situation, one segment, one pain), reply rates average 21.7% across our campaigns. When targeting is broad ("VP of Sales at SaaS companies"), the same copy gets 2.9%. Generic targeting with no situational context drops to 0.4%.

That's a 7.4x difference between hyper-specific and broad. And a 54x difference between hyper-specific and generic. No amount of copy optimization can close those gaps.

What to look for: does the agency talk about targeting before they talk about copy? Do they ask about your ICP's buying situations, not just their job titles? Do they test different audiences, or just different messages?

An agency that runs 24 copy variants on the same audience is testing paint colors. An agency that runs 4 copy variants across 6 audience segments is testing what actually matters.

Red Flag 4: The Handoff Problem

This is where most agencies fail silently. They generate replies and hand them to you. "Here are your warm leads. Good luck."

The space between a reply and a booked meeting is where most deals die. And most agencies don't own that space.

Here's what happens when a prospect replies "yeah, tell me more":

• If you respond in under 5 minutes, booking rate is 3 to 4x higher than responding in an hour.
• If you respond the next day, you've likely lost them. The context is dead. They've moved on.
• If you book the meeting but don't confirm twice before the call, your show rate drops to 50 to 60%.
• If the prospect no-shows and nobody follows up within 30 minutes, that meeting is gone forever.

Most agencies measure success by replies. But replies aren't revenue. Booked, confirmed, shown-up meetings are revenue. The agency that stops at the reply is delivering half a result.

What to look for: does the agency own the response process? Do they manage response time? Do they confirm meetings? Do they recover no-shows? Do they push qualified contacts to your CRM? Or do they just drop a spreadsheet of "interested leads" in your Slack?

Our show rate at LeadGrow is 9 out of 10 booked meetings. Industry average is 50 to 60%. The difference isn't luck. It's process: respond in under 5 minutes, confirm twice before the call, recover no-shows the same day, push everything to CRM in real time.

Red Flag 5: "Personalization at Scale"

Every agency says they do personalization at scale. Most of them mean: "We merge {first_name} and {company} into a template."

That's not personalization. That's mail merge. It's been around since 1995.

Real personalization isn't about surface-level variables. It's about diagnostic specificity. Does the email demonstrate understanding of the prospect's actual situation? Not "I noticed {company} is growing." That proves nothing. It proves you can use LinkedIn search.

Diagnostic personalization sounds like: "You posted 4 SDR roles last month. Most teams at that stage hit a wall when each rep's workflow is custom." That describes their situation. It proves you understand the implications of their hiring activity, not just the activity itself. This is the diagnosis over personalization approach in practice.

The difference in results is significant. Surface-level personalization with generic messaging produces 1 to 3% reply rates. Diagnostic messaging that accurately describes the prospect's situation produces 8 to 15%+.

What to look for: ask the agency to show you an example email for a prospect in your market. If it reads like it could be sent to any company in your industry with different variables swapped in, that's surface-level. If it references a specific situation that's happening at that type of company right now, that's diagnostic.

Red Flag 6: No Testing Methodology

Ask any cold email agency: "How do you test?" The answer reveals everything.

Bad answer: "We A/B test subject lines and optimize the winner." That's testing within one angle. It tells you which phrasing works, not which positioning resonates.

Good answer: "We test 3 to 4 positioning angles simultaneously in the first two weeks. Different frames, different situations, different value props. Then we identify the winning angle and optimize within it."

We test 24 to 48 variants in the first month of every engagement. Day one, we launch 12 tests: 3 situations, 2 offer lengths per situation, 2 frames. Within 4 days we know winners and losers. We take the principles from winners, apply them to losers, and test again.

By month 2, we know exactly which angle works for your market. Month 1 is a sprint to find the winning offer. Month 2+ is scaling the winner while continuing to test new angles with 20% of volume. Our Sprint, Test, Scale framework breaks this process down in detail.

What to look for: how many variants do they test in month one? What are they testing? (Angles, frames, and situations matter more than words and phrasing.) How quickly do they identify winners? And do they continue testing after finding a winner, or do they "set it and forget it"?

The 12% Positive Reply Floor

This is the metric most agencies don't track and most clients don't know to ask about.

Positive reply rate is the percentage of replies that indicate genuine interest (as opposed to "not interested," "please remove me," or "wrong person"). It's the clearest signal of whether the offer is landing.

Our rule: do not increase send volume unless positive reply rate is above 12%.

Below 12%, the message isn't working well enough. Scaling it means sending more emails that don't resonate, which wastes your market, damages your sender reputation, and produces meetings with people who aren't a good fit.

Above 12%, you have something. Scale it.

• 20%+ positive reply rate: Scale aggressively. Found something. Pour volume.
• 12 to 20%: Scale significantly.
• 8 to 12%: Scale steadily.
• 6 to 8%: Scale cautiously. More testing needed.
• Below 6%: Don't scale. Sprint for new angles.

Ask your agency what their positive reply rate threshold is for scaling. If they don't have one, they're optimizing for send volume, not results.

What to Actually Look for in a Cold Email Agency

Here's the evaluation framework we'd use if we were hiring someone else to do what we do.

1. Do They Diagnose Before They Prescribe?

Before proposing a solution, do they ask about your ICP, your deal size, your close rate, your previous outbound attempts, your competitive landscape? Or do they jump straight to "here's our package"?

An agency that asks questions before quoting is one that customizes the approach to your situation. An agency that has a standard package for everyone is a maintenance man.

2. Can They Show You Real Emails With Real Reply Rates?

Not sanitized case studies. Real emails. "This is the email we sent. This is the audience. This is the reply rate." If they can't show you at least 3 to 5 real campaigns with specific metrics, ask why.

3. Do They Own the Full Pipeline?

From first email to confirmed meeting on your calendar. Not just "we generate replies." If there's a handoff where they drop leads and you're supposed to follow up, you're buying half a service.

4. What Happens in Month One?

If month one is "we set up infrastructure and start sending by week 4," that's fine. Infrastructure matters. But what's happening during those 3 weeks? Offer testing? List building? Situation mining? Or just waiting for domains to warm up?

At LeadGrow, the first 3 weeks are infrastructure build plus offer workshop plus list building plus copy drafts. By week 4, we're sending. By week 6, we have consolidated market intelligence. The warmup isn't a delay. It's parallel preparation.

5. How Do They Handle Underperformance?

Every campaign has rough patches. The question is what happens next. Does the agency diagnose and pivot? Or do they just send more volume?

Look for agencies that have a clear escalation path: "If positive reply rate is below X% by day 21, we pivot the offer. If it's below Y% by day 45, we re-mine the ICP. If we can't find a winning angle by day 60, we'll tell you honestly."

The agencies that never lose aren't honest about their results. Every market has tough spots. Transparency about underperformance is a sign of integrity, not weakness.

6. What's the Minimum Market Size They'll Work With?

Cold email requires volume. If your total addressable market is under 30,000 companies, cold email might not be the right primary channel. An honest agency will tell you this upfront. A dishonest one will take your money and run out of prospects in 90 days.

Minimum viable TAM for cold email as a primary acquisition channel: 30,000+ total addressable companies. Monthly contactable pool floor: 5,000 to 8,000 accounts per month. Both must be met.

The Real Cost of a Bad Agency

The financial cost is obvious. $4,000 to $8,000 per month for 3 to 6 months with nothing to show for it.

But the hidden cost is worse. A bad agency burns your market. Prospects who received bad cold emails from your domain (or a domain associated with your brand) have already formed an opinion. When you come back with better messaging later, they've already decided you're spam. That impression is nearly impossible to undo.

A bad agency also trains your internal team to believe "cold email doesn't work." When the real problem was targeting, offer, or execution, not the channel itself. The conclusion becomes "we tried outbound, it failed" when the accurate statement is "we hired the wrong agency."

That's why the evaluation matters. Getting it right the first time isn't just about saving money. It's about preserving your market and your team's belief in the channel.

Questions to Ask Before Signing

Ten questions. If an agency can't answer these clearly, keep looking.

• What's your average positive reply rate across all campaigns? (If they don't track positive reply rate separately from total reply rate, that's a flag.)
• How many variants do you test in month one?
• What's your threshold for scaling volume? (Looking for a specific positive reply rate floor.)
• Can you show me 3 real emails with real reply rates from campaigns similar to mine?
• Who responds to replies? What's the average response time?
• What happens if the campaign underperforms at day 30? Day 60?
• Do you own the meeting booking process, or do leads get handed to us?
• What's the minimum market size you'll work with?
• How do you approach targeting? (Looking for situation-based, not just demographics.)
• What does your infrastructure setup look like? (Dedicated domains, warmup process, inbox health monitoring.)

The answers don't have to match our approach exactly. But they should be specific, honest, and demonstrate that the agency thinks about outbound as a system, not just a volume game.