A working one-person go-to-market stack runs roughly $50 to $150 a month in 2026 if you buy in the right order, and four figures if you buy the way the internet suggests. Six layers matter: data, sending infrastructure, sequencing, a place to put replies, a calendar link and basic analytics. Three of them you can skip entirely until you have paying customers, and the most common mistake is buying the CRM first.
Search for a go-to-market stack and you get a list written for a company with a revenue operations hire. Twelve tools, four figures a month, and an implicit assumption that someone's job is keeping them talking to each other.
Here is the version for one person, with prices, and with an explicit note on what to skip. Figures are published or third-party reported for September 2026.
The six layers, and what each is actually for
Data. Who exists, where they work, how to reach them. Sending infrastructure. Domains and mailboxes that are not the one your company runs on. Sequencing. Scheduling messages and follow-ups, with rules for stopping. A place to put replies. Somewhere conversations live so week two is not week one again. A calendar link. So booking does not take six emails. Analytics. What is working.
Six layers, and three of them you should not buy yet.
What each one costs
| Layer | Realistic monthly cost | Notes |
|---|---|---|
| Sending domains and mailboxes | $5 per address, plus ~$12 a year per domain | The line nobody budgets. Two or three addresses to start |
| Data | $0 to $79 | Apollo free is reported at ~900 credits a year; paid tiers reported $49 to $119 per user per month annual |
| Sequencing and sending | $30 to $100 | Instantly, Smartlead and similar sit at the low end at one-person volume |
| Replies | $0 | Your inbox. Genuinely |
| Calendar link | $0 to $15 | Free tiers are fine for a long time |
| Analytics | $0 | Your sending tool already reports what matters |
| An agent instead of the middle three | $39/mo (2,000 credits) after launch; $9 in beta | Replaces data, sending and sequencing rather than sitting alongside them |
Assembled at the low end, a working stack is roughly $50 to $150 a month. The same list bought the way most articles recommend comes to four figures, and the difference is almost entirely layers you do not need yet.
The order to buy in
Not by importance. By when the work actually happens.
First: sending infrastructure. Before data, before anything. Domains need about three weeks of warm-up before they can carry real volume, so this is the only item with a lead time. Buying it last means waiting three weeks with everything else already paid for.
Second: a way to build a list. Free tiers are enough to test a segment. You do not need a database subscription to find out whether a hundred people in one niche will reply.
Third: sequencing. Once you know the list is real and the domains are warm.
Later, and only when it hurts: the CRM, enrichment, analytics. Each of these organises something you do not have yet.
What to skip, specifically
The CRM. This is the most common first purchase and the least useful one. A CRM organises a pipeline; you do not have a pipeline, you have a hypothesis. Fifty conversations fit in a spreadsheet, and the discipline of keeping it current is the actual skill. You will know when to buy one: it is when the spreadsheet starts costing you more than it saves, and that moment is unmistakable.
The enrichment layer. Tools like Clay are genuinely powerful and are build-your-own systems. They reward someone who already knows what they want to build. On day one you do not, and the time spent learning it is time not spent finding out whether anyone wants the product.
Analytics beyond what your sender reports. Reply rate, bounce rate, complaint rate. That is the whole dashboard at this stage, and your sending tool already has all three.
The line item that is always missing
Your time. Running this stack properly is about a day a week, and most of that day is research and admin rather than the part that feels like selling.
For a technical founder that day comes out of the only work nobody else can do. It is the largest cost in the budget and the only one that never appears in it, which is why stacks get compared on invoice price and then quietly abandoned in week four.
That is the honest argument for an agent rather than a stack: not that it is cheaper on the invoice, though at $39 a month against a $50 to $150 assembled stack it can be, but that it removes the day. Operater's Sales agent covers the middle three layers, data, sending infrastructure and sequencing, and buys and warms its own domains as part of doing so.
What it does not do: it is not a CRM, it does not replace your calendar link, and it will not tell you your product is wrong. Those remain yours, and a tool claiming to replace all six layers is describing six things it does adequately.
The ninety-day version
Two sending domains with three addresses, warmed. A free data tier. One sending tool at the small end. A spreadsheet. A free calendar link. Roughly $50 a month, and enough to reach a few hundred of the right people and find out whether any of them care.
If they do, you will know exactly which layer is hurting and can buy that one. If they do not, you have spent $150 over three months learning something worth considerably more than that, and you have not spent it on a CRM.
Key takeaways
- Buy in the order the work actually happens: sending infrastructure and data before a CRM, because a CRM with nothing in it is a spreadsheet you pay for.
- Sending addresses and domains are the one line item nobody budgets and everybody needs. Roughly $5 a month per address plus about $12 a year per domain.
- The free tiers in this category are real but small. Apollo's is reported at around 900 credits a year, which is a test rather than a programme.
- Skip the CRM, the analytics platform and the enrichment layer until you have replies to manage. A spreadsheet holds your first fifty conversations fine.
- Your own time is the largest line in the budget and the only one that never appears in it.