#cold email

I review outreach tools for a living, more or less. Over the last year I have built a routine for stress-testing a cold email draft before it goes out at any volume, and I have settled on six free tools that earn a place in the pre-send check.

A few of these are well-known. A few are less famous than they deserve. None of them require a paid plan to get useful answers. The point of this list is the routine, not the brand names; the right tools change as the market changes, and any equivalent free tool in the same category would do.

Photo by www.kaboompics.com on Pexels

1. MXToolbox for the technical layer

You cannot diagnose content problems if the deliverability layer is broken, and most senders never check. Run your sending domain through MXToolbox before you write the email. It returns SPF, DKIM, DMARC, blacklist status, and a half-dozen other technical signals that mailbox providers actually look at.

Why this is first: a perfectly written cold email from a flagged domain goes to spam, where nobody opens it. Content optimization is wasted effort until the technical layer is clean. If MXToolbox surfaces a problem, fix it before sending anything.

2. Spamhaus for reputation

The most consulted spam reputation database in the world. Most major mailbox providers check against Spamhaus before delivering an email. If your sending IP or domain is listed, you are not getting through, regardless of the content.

Worth checking the IP and the domain separately. A clean domain on a flagged IP (shared hosting situations, mostly) is just as bad as a flagged domain on a clean IP. Both have to pass.

3. Mail Tester for one-shot scoring

Sometimes you do not want to interpret SPF records and DMARC policy reports. You just want a number that tells you whether your specific draft is going to land. Mail Tester generates a unique address, you send your draft to it, and it returns a 0-to-10 score based on what a real spam filter would see.

A score of 9 or higher means the technical layer is clean for this specific email. A score below 7 means there is a problem you need to fix before sending at any real volume. The tool tells you what dimension is dragging the score down, which is usually more useful than the score itself.

4. EvvyTools for content-layer scoring

Once the technical layer is clean, the content has to do the work. This is where this free outreach tool earns its place in the routine. It scores a draft against seven content dimensions: subject line, opener, value proposition, CTA specificity, length, spam triggers, and personalization. Returns specific rewrites for the lines that drag the score down.

What it catches: the unconscious template patterns that creep into outreach after you write a few similar emails in a row. An opener that reads as "I hope this finds you well." A CTA that asks for a thirty-minute meeting up front. A subject line that does not survive mobile truncation. The kind of things that are obvious once pointed out but invisible from inside your own draft.

What it does not catch: voice problems. A draft that scores 90 on the seven dimensions can still read as robotic. The scoring tool is a diagnostic, not a writing partner.

Honest comparison: tools in the same category I have tested include Lavender (paid, more polished), Smartwriter (paid, narrower focus on personalization), and the inline scoring built into most outreach platforms (worth very little for content quality, mostly checks deliverability). For the free tier, the EvvyTools scorer is the most useful one in my routine.

Photo by Jan van der Wolf on Pexels

5. Hemingway Editor for readability

A clean prose check. Hemingway Editor flags long sentences, passive voice, complex constructions, and reading-level grade. Useful as a final pass on a cold email body to make sure the language is at a sixth-to-eighth grade reading level, which is where most outreach lands best.

The free version runs in the browser and is enough for a single-draft check. The desktop version adds save-and-export, which is convenient but not necessary for outreach use.

What I would warn about: Hemingway is opinionated about voice. It flags any passive construction, even when the passive is the right call. Treat the suggestions as informational, not mandatory.

6. The CAN-SPAM compliance guide

Not a tool, but a document worth reading once a year. The CAN-SPAM Act sets the legal floor for cold email in the United States: clear sender, valid postal address, working unsubscribe, no deceptive subject lines. The Federal Trade Commission maintains the canonical compliance guide.

Most outreach platforms comply with most of this by default, but not all of them. The unsubscribe requirement is the one that catches people, because some automation tools either hide the link or skip it on follow-up emails. Read the guide once and you will know what to verify in your specific sending stack.

The routine

In order, before any campaign: 1. MXToolbox the sending domain. 2. Spamhaus the sending domain and IP. 3. Send the draft to Mail Tester, aim for 9+. 4. Run the draft through the EvvyTools cold email scorer. 5. Hemingway the body if the scorer flags length or complexity. 6. Spot-check CAN-SPAM compliance on the unsubscribe link.

The whole loop takes about fifteen minutes once you have done it a few times. It catches the technical and content problems that account for most reply-rate failures.

For background on what the seven content dimensions actually mean and how they interact, the longer guide How to Diagnose a Failing Cold Email Before You Hit Send is worth reading once. Other writing utilities I have reviewed live at the same site's tools directory if you want to compare other diagnostics.

What I am not running

A few tools that get recommended in cold email guides but that I have removed from my own routine over the last year: - Subject-line generators. The output is generic; faster to write your own. - Word-by-word spam scanners. False-positive heavy; the broader scoring tools handle this dimension as part of the bigger picture. - Most paid "personalization at scale" services. The output reads as templated to anyone paying attention.

The pattern: tools that automate the writing tend to produce average outputs. Tools that diagnose existing drafts tend to produce useful signal. I keep the diagnostics and skip the generators.

The reviewer takeaway

A free, six-tool routine that catches most of what kills cold email reply rates is genuinely available. The senders who run something like it consistently outperform the senders who skip it, by margins that show up in the data within a few weeks.

Build the routine once. Run it before every batch. The number of broken campaigns it prevents pays for itself in the first month.

A note on the tools I rotated out this year

Tool reviews are only useful if I am honest about what I have stopped using and why. A few specific tools that have been in past versions of this routine:

A specific subject-line generator I will not name dropped out because the outputs became formulaic; every line read like the same machine wrote it, which is the opposite of what I want from a tool meant to differentiate my outreach. The patterns got obvious within a hundred emails.

A reputation-monitoring service got swapped out for MXToolbox plus Spamhaus because the paid features I was using were just nicer dashboards over the same underlying data. The free combination got me the same answers in two more clicks.

A paid scoring tool got replaced by the EvvyTools free option, because once I compared the recommendations side by side on the same drafts, the free tool's suggestions were just as useful for my use case. I keep paying for tools when they earn it. This one stopped earning it.

I have been testing outreach tools for the better part of two years now, and there is one failure mode that shows up in almost every cold email I get reviewed by readers asking what went wrong. The subject is fine. The opener is decent. The value prop is real. Then the last sentence asks for a thirty-minute call with a senior decision-maker who has never heard of the sender, and the email goes to the trash.

The CTA is where the most outreach dies. Not because senders forget to include one, but because they include one that asks the reader to value something at more than the reader has any basis to value it. Worth writing about because the fix is small and the senders who get it right separate themselves from the senders who do not.

Photo by wal_ 172619 on Pexels

The asymmetry the sender does not see

The sender thinks of the meeting as a free thirty minutes. The reader thinks of it as thirty minutes they cannot get back, with a person who has not yet earned the time. That asymmetry is the whole problem.

You have to think about CTAs the way investors think about portfolio allocation. The reader has a finite pool of meeting time. Every meeting commitment is a withdrawal from that pool. The reader's default position on any new outreach is "no, my pool is fully allocated." Moving them from "no" to "yes" requires either a very high expected return or a very low cost.

Cold email senders almost always optimize for the wrong side of that equation. They try to raise the expected return ("we deliver a 40 percent reduction in costs") when they should be lowering the cost ("reply with a thumbs-up if this is worth fifteen minutes next quarter"). The thumbs-up costs the reader nothing. The thirty-minute meeting costs them a meaningful slice of their week.

What a small ask looks like

The CTAs that consistently convert in cold email all share the same property: they require a one-second reply.

"Are you the right person to talk to about this, or should I follow up with someone else?"

"Reply with a thumbs-up if this is worth a fifteen-minute call next quarter."

"Would Thursday at 2pm or Friday at 10am work better?"

"If this is not for you, who should I be reaching out to?"

None of these ask for a meeting up front. They ask for a signal that the sender can build a meeting on. The reader can answer in five seconds, and the answer is binary or near-binary. Five seconds is a low enough price that even readers who think they want to ignore the email find themselves replying anyway.

The senders who ask "would you be open to a 30-minute call to walk through our solution" are asking the reader to pre-commit to half an hour. The senders who ask "is this worth a brief call sometime next quarter" are asking the reader to commit to nothing more than an opinion. The difference in conversion is roughly an order of magnitude.

The escalation ladder

If the small ask works, the next email can escalate. This is the part most senders skip.

A clean outreach sequence looks something like: 1. First email: a small-ask CTA (thumbs-up, routing question, opinion). 2. Reply received: a medium-ask follow-up (a fifteen-minute call, a Loom video, a specific question). 3. After the medium ask is answered: the full ask (the thirty-minute call, the demo, the proposal).

Each step earns the next. The reader who replies with a thumbs-up has signaled they are interested enough to engage; the next email can ask for slightly more. The reader who never replied has told you the value prop did not land, and sending the thirty-minute call request anyway is just spam.

Senders who collapse the ladder into one email get one chance at conversion. Senders who spread it across three emails get three chances, each one informed by the reader's response to the last. The math is obvious in retrospect.

How this shows up in scoring tools

The cold-email scoring tools I have tested all score CTA specificity as a separate dimension from value prop strength, which is the right call. Many drafts I have run through these tools have strong value props and weak CTAs, or vice versa. Scoring them together would hide which problem is which.

The cold email scorer at EvvyTools is the one I have reached for most often this year. It scores CTAs against the seven-dimension framework: subject line, opener, value proposition, CTA specificity, length, spam triggers, and personalization. The scorer flags CTAs that ask for too much (a common failure mode) and suggests smaller-ask alternatives, which is exactly the rewrite most drafts need.

What it gets right: the CTA specificity dimension does not just flag long CTAs as bad. It evaluates whether the ask is proportional to the value the email has earned. A short CTA can still ask for too much ("Book a 30-minute demo here") and a longer CTA can still ask for the right amount ("Reply with a thumbs-up if this is worth a follow-up").

What I would change: the scorer is more useful as a diagnostic than as a rewrite engine. The specific rewrites it suggests are usually decent but not necessarily what I would write. I treat it the way I treat a code linter: it tells me what is wrong, I decide how to fix it.

Tools in the same category I have tested: - Lavender (paid, more polished, narrower set of dimensions). - Smartwriter (paid, more focused on personalization than CTA quality). - The free tier on most outreach platforms (worth almost nothing for content scoring, mostly checks deliverability).

For the cost (free), the cold email scorer earns its place in the toolkit.

The fix is one line

If your CTA is asking for a meeting and your reply rate is below five percent, replace the meeting ask with a smaller ask in the next batch and see what the number does. Most senders see a measurable improvement within fifty emails.

The framework that backs this up (the seven dimensions of cold email diagnostics, what each one means, and how they interact) is laid out in the longer guide How to Diagnose a Failing Cold Email Before You Hit Send. Worth reading once if you do outreach at any volume.

For deliverability checks (the layer that decides whether the email reaches the inbox at all), MXToolbox and Spamhaus cover the basics. CTA scoring lives at the content layer; deliverability lives at the technical layer; both have to work.

The reviewer take

Cold email CTAs are one of those rare cases where the fix is genuinely small and the result is genuinely visible in the data. Asking for less converts more. The reader gives you the signal you actually need (interest), the sender saves the bigger ask for the second email, and the reply rate goes from a trickle to a real number.

The senders who get this right are not the ones who write more sophisticated CTAs. They are the ones who write smaller ones. Every diagnostic tool I have run drafts through agrees: the CTA dimension is where the most outreach loses the most reply rate, and the rewrite is one of the cheapest in the entire sequence.

Everyone Has an Inbox Everyone who uses the internet has an email address. Not everyone is on LinkedIn. Not everyone uses Instagram, or TikTok, or whatever platform is currently being declared the future of marketing (even Reddit). But everyone has an email address. Sometimes more than one. Email marketing has been pronounced dead more times than I can count, and for the better part of thirty…