Cold outbound for about $220 a month: the whole build

A complete cold outbound operation, from registered domains to deliverability monitoring, costs $221.10 a month at the advertised rates of the tools we track. It cost $160 when we first priced it, and one line moved: MailStrike replaced flat mailbox bands with graduated per-mailbox rates in September 2026, taking six warmed mailboxes from $45 to $106. Not the sending tool alone: the whole build. Three secondary domains, six mailboxes, AI-driven warming, a list source, verification, a sequencer, and continuous reputation monitoring.

The order matters as much as the tools. Most failed outbound programs die from sequencing errors, not tool choices: sending from the main domain, skipping the warmup period, mailing an unverified list. So this playbook walks the build in the order you should actually do it, with the budget assembled at the end.

Step 1: buy secondary domains, never touch your main one

Cold email carries risk to whatever domain sends it. A spam complaint spike or a blacklist listing on your primary domain damages every email your company sends, including invoices and password resets. So the first purchase is domains you can afford to burn: two or three variants of your brand (getacme.com, tryacme.com) that redirect to your real site.

Mailforge registers domains at roughly $14 a year for a .com, and Maildoso at a flat $12. Three domains cost about $42 a year, or $3.50 a month, and they are the cheapest insurance in this entire build. Set up SPF, DKIM, and DMARC on each before anything sends; the infrastructure vendors below do the DNS work for you, which is half the reason to use them.

Step 2: provision mailboxes, about $3 each

Six mailboxes across three domains, two per domain, is the right shape for a first build. The math that gets you there: each mailbox sends 20 to 30 cold emails a day once warm, so six mailboxes support 120 to 180 sends a day without any single mailbox looking like a blaster, and two per domain keeps any one domain's reputation from carrying too much of the program.

Mailforge sells mailbox capacity as slots at $3 per slot per month on annual billing with a ten-slot minimum, so the entry configuration is about $30 a month, and the four slots this build leaves empty on day one are replacement capacity for when a mailbox burns. Slots are capacity rather than named mailboxes, which means retiring a burned mailbox and creating a replacement costs nothing extra, a genuinely better model than per-mailbox billing for real outbound. The trade-off is that this is shared cold email infrastructure rather than real Google or Microsoft accounts; Primeforge, its sibling, provisions actual Google Workspace and Microsoft 365 mailboxes at $4.50 a slot for teams that want provider-native sending. Zapmail is the third option at $39 a month for ten Google mailboxes with pre-warming bundled in.

Step 3: warm for two to four weeks before sending anything

A new mailbox that starts sending cold email on day one gets filtered almost immediately. Mailbox providers score senders on history, and a fresh domain has none. Warming builds that history: the mailboxes exchange real-looking conversations with established inboxes, accumulating opens, replies, and thread depth before any prospect sees an email.

This build uses MailStrike for the warming layer, and the reason is architectural. Warming tools have been through two generations: fixed-volume bots trading identical messages, which providers learned to fingerprint years ago, and randomized shared networks, harder to spot but still recognizably mechanical. MailStrike is built as the third: every connected mailbox gets its own AI Agent, with a persona carrying its own reply rate, dwell time, active hours, and even a chance of typos, and the threads are written by an LLM at send time from your industry and company description rather than drawn from a template pool. A standard ramp reaches full volume by day 21, after which warming continues at a maintenance level of roughly 15 to 25 emails a day.

Pricing is per mailbox on graduated rates: $18 each for the first five and $16 for the next five, so this build's six mailboxes come to $106 a month on annual billing, with one deliverability test per mailbox per month included. That is the single biggest line in this budget, and it is worth stating plainly that it used to be $45. If it will not fit, TrulyInbox at $22 a month with unlimited connected accounts is the substitution to make, at the cost of the agent architecture described above. Mailboxes connect over OAuth for Google Workspace and Microsoft 365, or over SMTP for custom providers, on every tier. One honest caveat from our dossier: agent threads are English-only, so programs sending mainly in other languages should check that before committing.

Budget the calendar time, not just the money. The two to four week ramp is not optional, and every week you shave off it comes back as spam-folder placement later.

Step 4: build the list

While the mailboxes warm, build the list. GetProspect Starter is $34 a month on annual billing for 1,000 valid email addresses, and the pricing is honest in a way that matters: credits are consumed only for valid, verified emails, so a lookup that returns nothing costs nothing, and every feature including the database, the LinkedIn extension, and the API is available on every plan. A thousand net-new verified contacts a month is well matched to a 120-to-180-a-day sending operation once you account for multi-step sequences.

Apollo.io's free plan is the supplement, with limited credits and basic filters that cover looking up the contacts your primary source lacks. Spend the warming weeks here: a tightly filtered list of the right 1,000 people will outperform a loose 5,000 every time, and it costs less to verify and send.

Step 5: verify everything, even the pre-verified

GetProspect verifies what it returns, but lists age and catch-alls lie, and deliverability punishes overconfidence. Run everything through a dedicated verifier before it enters a campaign. MillionVerifier charges $39 for 10,000 credits, $0.0039 per address, with no subscription and no expiry, and you are charged only for good and bad results; risky results, including unresolved catch-alls, cost nothing. At this build's volume a 10,000 block lasts about four months, so it lands in the budget at roughly $10 a month.

The operating rule: anything the verifier calls risky stays out of the campaign. Bounce spikes are how mailbox providers learn to distrust a domain, and the entire point of the preceding four steps is not to spend that reputation on addresses a $0.0039 check would have caught.

Step 6: the sending tool

Instantly Growth is $47 a month, or $37.60 on annual billing, and its pricing model fits this architecture: sending accounts are never metered, so all six mailboxes connect at no extra cost, and the plan meters uploaded contacts and monthly email volume instead. Rotation across mailboxes, sequencing, and reply detection are the core of the product. Note that contact caps count uploads rather than sends, so keep the list hygiene from step 5; a bloated CSV consumes the allowance whether or not it is contacted.

Smartlead Base at $39, or $32.50 on annual billing, is the closest substitute, with unlimited mailboxes and warm-up on all tiers and a stored-contacts meter that experienced users manage by archiving. Either works. What you are buying in both cases is the orchestration layer, not the deliverability layer, and the distinction is the subject of the next step.

Step 7: monitor, test, and prime before every send

Most teams treat deliverability as something to investigate after a campaign fails. The correct posture is the reverse, and it is why the monitoring layer is a standing line item rather than an emergency purchase.

MailStrike again, because the same subscription that warms the mailboxes also watches them. Each mailbox carries an Inbox Reputation Score refreshed daily with a 14-day trend, and the dashboard clears a mailbox to send only once score, warmup time, and domain age all pass their thresholds, then tells you how much it can safely send. Flashman, the built-in assistant, writes a daily summary per mailbox covering unusual trends and what to do about them. For a build this size, that converts the classic silent failure, three weeks of spam-folder sending nobody noticed, into something you read over coffee.

Two of its capabilities deserve specific attention because they change how you launch campaigns rather than just how you monitor them.

The first is the built-in deliverability test, included at one per mailbox per month, so this build gets six. You compose the message in the app and MailStrike sends it from your own connected account, through the same authenticated path your real campaigns use, to a seed pool spanning Gmail, Microsoft, and custom-domain providers. Results arrive in about three minutes: inbox, spam, or missing per provider, with SPF, DKIM, and DMARC pass and fail results attached. Because the test sends from your account, the result reflects your actual reputation, not a testing relay's.

The second is the Seedlist, a network of thousands of trusted Gmail and Outlook inboxes sold as an add-on priced by seed volume. The idea inverts the usual testing sequence: instead of spot-checking a message after something breaks, you pre-send the actual campaign into the seedlist before launch. Two things come back. Per-provider delivery feedback shows exactly where that content lands while there is still time to fix it, and the seed inboxes open, read, reply to, and rescue the message from spam, so the exact content accumulates a body of positive engagement before the first prospect receives it. This is the general principle behind seed testing done right: placement tests run into or before a send, so the content is already trusted when the real campaign goes out, not audited after it flopped. The Seedlist is quoted per customer rather than self-serve, so it sits outside this budget; at this build's volume the included tests cover the same job at spot-check scale.

Worth being clear about the boundary: MailStrike sends no campaigns. It warms, monitors, tests, and primes the mailboxes and content that Instantly then sends. The two tools are adjacent layers, not substitutes.

The budget, honestly totaled

Line	Tool	Monthly cost
3 secondary domains	Mailforge registration, $14 a year each	$3.50
10 mailbox slots (6 in use)	Mailforge, annual billing, ten-slot minimum	$30.00
Warming and monitoring	MailStrike, 6 mailboxes, annual billing	$106.00
List building	GetProspect Starter, annual billing	$34.00
Verification	MillionVerifier, $39 per 10,000 credits, four months	$10.00
Sending	Instantly Growth, annual billing	$37.60
Total		$221.10

The honest caveats. Four of the lines assume annual billing at the advertised rate, so a strictly month-to-month version runs closer to $255 (MailStrike moves from $106 to $132.50, and Instantly from $37.60 to $47). The verification line is amortized, so the first month's cash outlay is $29 higher than the table shows. And four of the ten mailbox slots sit unused until a mailbox needs replacing, which is deliberate slack rather than waste. What the table does not hide: there are no per-seat charges anywhere in it, so a second person joining the operation costs nothing until the list volume grows.

The short version

Buy three burnable domains, put six cheap mailboxes on them, two per domain, and let MailStrike's agents warm them through the 21-day ramp while GetProspect builds the list. Verify everything through MillionVerifier, send through Instantly at 20 to 30 emails per mailbox per day, and read the daily reputation score so failure is an alert, not an autopsy. Test into or before the send, never only after. Total: $221.10 a month at advertised annual rates.