What I Learned Sending 200 Cold Emails in 30 Days as a Freelancer

I kept a spreadsheet for thirty days. Every row was a cold email I sent. Column one was the prospect name. Column two was the subject line. Column three was the body. Column four was the day sent. Column five was the result: opened, replied, converted, or nothing. I did not have a sophisticated reason for tracking it this way. I was just tired of guessing whether cold outreach worked and wanted actual numbers. The data I collected was messier and more useful than I expected.

Two hundred emails over thirty days. Eleven client conversations. Three paid projects. One repeat client who came back six weeks after the initial sequence ended. The overall reply rate was eleven percent. The conversion rate from reply to paid was twenty-seven percent. But those averages hide most of what actually happened, which is where the useful information lives.

I did not go into this as a skeptic. I went in as a believer who had spent three months sending maybe five or six cold emails per week and getting nothing. The belief had degraded into mild cynicism. The spreadsheet was an attempt to either fix the process or prove that cold email was not worth the time. The data turned out to be neither of those things. It was more complicated and more instructive.

The setup before I sent anything

I spent three days on list building before sending a single email. Most people skip this step and it shows in their results. I built a list of fifty prospects I wanted to work with specifically. Not random names. Not scraped lists. People I had identified as running companies I could actually help, based on their job title, company size, and the visible problems the company appeared to have. I had about twenty names at first and kept adding throughout the month.

The subject lines I initially used were what I would describe now as cautious. “Quick question about [company name]” was the opener for roughly the first forty emails. It felt safe. It performed respectably. Average open rate across those first forty was around fifty eight percent. Reply rate was about eight percent. Nothing dramatic. But the reply rate was real, and two of the people who replied became active conversations. That gave me enough signal to keep going while I experimented on the other half of the list.

I should say something about the offer. I was pitching writing and content strategy work for B2B SaaS companies, a niche I had been building for about eight months. The specificity of that offer is I think the reason the early reply rate was not zero. Vague offers do not convert, and a lot of the cold email advice on the internet is written for vague offers.

The first two weeks the data was depressing

Days one through fourteen produced eighty seven emails and four client conversations. Two of those conversations came from a single subject line that I stumbled onto accidentally in day nine. Two weeks of sending and my conversion numbers were: eighty seven emails, eleven opens marked by email tracking, nine actual replies, four conversations, zero paid projects. I was not losing money. I was just not making any.

The two conversations came from what I started calling the “observation opener.” I had run out of subject line ideas and started testing ones that referenced specific things I had noticed about each prospect. “The onboarding email on your pricing page has a broken link” went to one company. “Your comparison page still mentions the old feature set” went to another. Both opened. Both replied. One of them turned into a three-week project four weeks later. The other said no but referred me to someone else.

The pattern was obvious in retrospect and invisible while I was inside it. Emails that led with the sender were performing worse than emails that led with the recipient. The subject lines that referenced something specific about the prospect were outperforming the generic ones by a significant margin. And the emails that assumed nothing except that the recipient had a specific problem I could name were getting replies that the longer, more polished versions were not.

The biggest surprise of week two was that my best-performing email by reply rate was also my shortest. Forty two words. One specific observation. One implication. One question. It had a twenty one percent reply rate out of nineteen sends. The average across the rest of my list for that period was seven percent. That number changed how I wrote every email for the rest of the month.

What changed in week three

Week three was when I stopped writing emails and started running experiments. I had enough data from the first two weeks to know what kind of subject lines opened and which ones did not. I had a rough sense of what the people who replied had in common. I used that information to redesign the second half of the campaign, and the difference was immediate.

The change was not complicated. I stopped trying to write a good email and started trying to write one specific thing. Every email after day fifteen opened with a concrete observation about the specific company, not about their industry or their role. Every email stated the implication of that observation in one sentence. Every email asked one low-friction question as the closing move. No case study in the first email. No call to action requiring calendar access. Just a question worth answering.

The reply rate on week three emails was nineteen percent, up from eight percent in weeks one and two. The open rate held roughly steady at around sixty percent, which told me that subject lines were not doing most of the work. The email body was. The structure I had settled into was the three-part observation hook that you have probably read about in other cold email advice. What I had not understood before the data was that knowing the structure and reliably executing it are two different skills. The spreadsheet forced the execution.

The numbers by category

Here is the breakdown as cleanly as I can state it. Two hundred emails sent. One hundred and fourteen opens tracked. Twenty two replies. Eleven that turned into actual conversations. Three that converted to paid work.

The eleven percent overall reply rate masks a significant split. Emails sent in the first half of the campaign, when I was still iterating on approach, averaged eight percent reply rate. Emails sent after I had locked in the observation-based structure averaged nineteen percent. That means the last hundred emails I sent produced sixteen of my twenty two total replies. The first hundred produced six.

The conversion rate from reply to paid was twenty seven percent. This number surprised me because I expected it to be lower. Most of the people who replied were genuinely interested in what I was offering. They were not wasting my time. The problem was not the conversion rate. The problem was that the reply rate was low enough that the conversion rate did not matter much in raw numbers. Three paid projects out of two hundred emails reads like a failure. Three paid projects from eleven qualified conversations reads like a normal sales funnel.

The one metric I did not track well but should have: time spent. I kept notes on how long each email took to write in weeks one and two, before I built templates. Average of eighteen minutes per email. After I built templates and started batch writing, average of six to eight minutes per email. The time savings were real. The quality difference was not as large as I expected. Batch writing with a template produced emails that performed nearly as well as the individually crafted ones from weeks one and two.

What I would do differently

Two things I got wrong that mattered. The first: I should have started with the observation-based subject lines instead of getting there by accident in week two. The data difference between the two approaches was large enough that I left roughly six to eight replies on the table by opening with generic subject lines for the first third of the campaign.

The second: I sent too many emails to people who were not good fits. My list building got tighter from week two onward, and the reply quality improved noticeably. The first forty or so emails went to a mixed list that included some targets where my offer was only tangentially relevant. Those emails produced one conversation. The emails that went to tightly targeted, obviously relevant prospects produced conversations at nearly four times that rate.

What I would not change: the spreadsheet. Tracking everything was tedious and I almost skipped it. But having the data meant I could see the pattern that converted a five percent reply rate into nineteen, and that pattern was invisible without the numbers. Cold email advice is full of confident prescriptions about what works. Most of it is written without anyone actually measuring. The measurement is what teaches you.

If you are going to run a cold email campaign, pick one specific variable to test at a time. Hold everything else constant. Track it. After forty or fifty emails, check the data. Adjust one thing. Test again. A spreadsheet is not the exciting part of outreach. But the people who skip the tedious part are the people who conclude that cold email does not work after one month of sending without any measurement. The campaign that is tracked is the campaign that improves.