Founders decide to run outbound themselves to save the money, and they are right that it is cheaper. They are usually wrong about which bill they are paying.
The cash side of cold email is small, knowable and boring. At the scale most small B2B companies need, the lines I can price from a public source come to a few hundred a month. The part that decides whether DIY works is a fixed block of about twenty to thirty hours to set it up, then roughly eleven to eighteen hours every week, forever, from somebody who is currently doing something else.
Here is the whole thing, line by line, priced where a price exists and counted in hours where it does not. If you are weighing this against a hire or a contract, this post prices the column you were planning to call free.
The cash column, with sources
Take a setup sized for a small company: fifteen domains and thirty mailboxes. That is inside the range I build to, and the reasoning behind those counts is in the sending setup this all sits on.
Mailboxes. Google's own pricing page lists Workspace Starter at 6.80 euros per user per month at list price, with discounts for the first months and a cheaper annual commitment (checked September 2026, pricing shown in euros on the page served to me). Thirty mailboxes at list is about 204 euros a month.
Domains. The one line everybody already knows the price of, and the smallest number on this page. Fifteen secondary domains is a rounding error against the mailboxes.
A data source. LinkedIn prices Sales Navigator Core at $119.99 monthly, with Advanced at $159.99, as of September 2026. One seat covers one person building lists.
A sending tool and verification credits. Both vary enormously by vendor and by volume, and I will not invent a figure I cannot point at a public page for. Price your two shortlisted options yourself and add them to this line.
So: roughly two hundred euros of mailboxes, about a hundred and twenty dollars of data, plus two subscriptions you will price in ten minutes. That is the entire cash story, and it is why the "just do it yourself" advice sounds so reasonable in a blog post.
The setup block is twenty to thirty hours and two weeks of waiting
None of that money does anything until somebody sits down and builds it.
- Buying fifteen domains, setting up thirty mailboxes, naming them so they do not look like a farm: 4 to 6 hours.
- SPF, DKIM and DMARC records on every domain. The first time, with the documentation open and one record that will not validate: 3 to 4 hours.
- Warm up every account for at least ten to fourteen days before it sends anything real. Almost no hands-on time, and it is still two weeks of the calendar you cannot compress.
- Building and verifying the first list properly: 6 to 10 hours, and more if the market needs research before it needs scraping.
- Writing the sequence, plus the variants you will test: 5 to 8 hours if you have written cold email before, considerably more if you have not.
- Connecting everything, setting sending limits, testing that a message actually arrives: 3 to 4 hours.
Call the whole block twenty to thirty hours of work spread across two to three weeks of waiting.
Then it becomes a weekly bill that never stops arriving
This is where doing it yourself quietly falls over, and skill is rarely what causes it.
- List building and verification, every week, because lists exhaust: 3 to 5 hours.
- Copy iteration. Reading what got answered, changing one variable, testing again: 1 to 2 hours.
- Inbox management. Every reply answered the same day, the positive ones and the "not now" ones and the annoyed ones: 5 to 8 hours across thirty mailboxes.
- Watching deliverability. Bounces, spam complaints, one mailbox going quiet for no obvious reason: about an hour.
- Booking, rescheduling and keeping the CRM honest: 1 to 2 hours.
Eleven to eighteen hours a week. Up to two working days, every week, in a business where you are also the person selling, delivering and making payroll.
And the moment you decide to double the volume, you repeat most of the setup block. More mailboxes, more warm-up, more calendar. Volume is the last step of this build rather than the first, and the reason is right here in the hours: scaling a system multiplies the maintenance you are already behind on.
The hours are not free, and pretending otherwise is the error
Any work should be paid. That is a principle I apply to other people and the one founders refuse to apply to themselves.
I am deliberately not going to multiply your hours by a rate. You know what an hour of yours is worth, and you know which part of your week those hours would come out of. My only claim is that "free" is the wrong word for a resource you cannot restock, and that the weekly hours are where the real price of the DIY column sits.
I have sat in enough month-four handover calls to know exactly which hours get skipped first. Not the sending, which is automated. The replies, the list rebuild, and the weekly deliverability check. The campaign does not stop. It just stops working, slowly, while the calendar tells you it is still running.
What the market charges to take those hours away
A hire and an agency retainer both cost real money, and the price ranges you will find online come from sellers with a bias. Get two real quotes, put your hour count next to them, and you have an actual decision rather than a feeling.
When doing it yourself is the right answer
Often, honestly.
Do it yourself when you have the hours and can protect them, when your market is small enough that list building is research rather than volume, and when you want to learn what good looks like before you pay anyone to do it. A founder who has run their own sequences for a quarter is a far better buyer afterwards, and negotiates like one.
Do not do it yourself when those hours come directly out of delivery or out of the sales calls the campaign is meant to produce. That trade is the one that shows up months later as a campaign that technically ran the whole time and booked almost nothing.
The output side is worth remembering while you count. One campaign we ran for an IT company in Germany turned 29 days of booked calls into 278,000 euros of new business. Building and running that whole system is the work I do with my team, which is exactly why I would rather you saw the hour count before you decide anything.
Open a calendar. Block the low end, eleven hours, next week and see which ones you could genuinely give. That is the whole analysis.