Short answer: Approve the ideal customer and a sending policy in week 1, and start mailbox warm-up on day 1. A new managed mailbox sends nothing until it is warm, which takes at least 14 days and usually about 3 weeks [[1]](#sources). Usually the small send lands in week 4, and the replies then decide whether to scale or stop. Nothing sends outside a policy a person approved [[2]](#sources).
Who this is for
This plan is for a founder, a first sales hire or a small revenue team running a first outbound test. You may have a few closed deals, or none. You want a month that ends in a decision, not a month that ends in a pile of sent email.
It assumes one person is accountable for approvals. In Outlevo, policies set who can be contacted, on which channels, how often and with what budget [[2]](#sources). Someone has to own those answers.
It does not tell you which markets to enter. Outlevo asks which countries you sell into and never proposes markets [[3]](#sources). You bring that answer to week 1.
If you are at an early-stage SaaS company, read outbound for early-stage B2B SaaS alongside this plan. If you are weighing approval-first outbound against fully autonomous tools, the SDR comparison covers that choice.
Steps
The month has four stages. Each one ends with a go/no-go check, listed in the next section. Do not start a stage until the previous check passes.
Week 1: ICP and policy sign-off
Week 1 produces two documents: an ideal customer profile (ICP) a person has approved, and a sending policy a person has approved. Mailbox work also starts this week, covered under week 2.
Start from evidence, not opinion. If you have past deals, import them. A CSV import of up to 200,000 rows can bring in companies and won or lost deals for the ICP and win/loss analysis, and importing uses no credits [[4]](#sources). You can also import past deals from your CRM, and the win/loss analysis proposes ICP hypotheses [[5]](#sources). If you use HubSpot, Outlevo reads companies, deals and owners [[6]](#sources). Nothing is written back until an owner or admin approves the field mapping [[6]](#sources).
Read the confidence label. Win/loss analysis counts companies, not deals, and labels small samples. Under 10 won companies is very low confidence [[3]](#sources). If you are under that line, treat the ICP as a bet you are testing, not a finding. That is a fair way to run a pilot, as long as you say so in the review.
No deal history? Start with the assessment. The free ICP assessment reads up to 5 public pages of a website. It returns facts labelled as verified observations or AI inferences, and up to 5 experimental segment hypotheses [[7]](#sources). It also asks which countries you sell into [[8]](#sources).
Approve one segment. AI-proposed ICP segments are labelled experimental until a person approves them [[3]](#sources). For a 30-day pilot, approve one segment. Several segments in one small send will make the replies hard to read.
Write the policy. The policy is where a person decides who can be contacted, on which channels, how often and with what budget [[2]](#sources). Fill in each of those four fields on paper first. Then confirm that every external action is checked against the approved policy before it runs and recorded in a ledger [[2]](#sources). You will use that ledger in week 4.
Pick one channel. LinkedIn is available, but a one-channel pilot is easier to read. If you are tempted to add it, read the terms first. LinkedIn's User Agreement, checked 2026-10-22, bars users from this: "Use bots or other unauthorized automated methods to access the Services, add or download contacts, send or redirect messages" [[9]](#sources). That decision belongs to you and your own account. Outlevo's LinkedIn step is a connection invitation with a note of at most 300 characters, then one message after acceptance [[10]](#sources). Automatic sending comes with Self-Serve and Managed, or with a LinkedIn seat on Builder. Otherwise the invitations and messages are drafted for you to send by hand [[10]](#sources).
Set the credit budget. Check the pricing page and decide how many credits the pilot may use. Builder is USD 29 per month with 320 credits a month [[11]](#sources). Growth is USD 149 per month with 3000 credits a month [[11]](#sources). Self-Serve is USD 199 per month with 1600 credits a month [[11]](#sources). Some per-action costs matter for a pilot: contact discovery costs 4 credits, email verification costs 1 credit, qualification costs 3 credits and deep research costs 30 credits [[11]](#sources). Sending, follow-ups and the mailboxes your plan includes use no credits [[12]](#sources).
Week 2: Domain and mailbox readiness
Mailboxes take the longest to get ready, so begin as soon as you have chosen domains and sender names. Do not wait for week 2 to start.
Managed mailboxes. Outlevo buys sending domains, creates the mailboxes, sets up DNS and warms them up, so you never touch your main domain [[1]](#sources). You choose the domains Outlevo proposes and the sender names [[1]](#sources). A new mailbox sends nothing until it is warm. That takes at least 14 days and usually about 3 weeks [[1]](#sources). The Self-Serve plan includes 2 domains and 4 mailboxes [[1]](#sources). If setup starts on day 1, a mailbox is warm on day 15 at the earliest, so the earliest send is in week 3. At the usual three weeks, it is warm around day 21, so plan the send for week 4 and shrink the review window to match.
Your own mailboxes. If you already send through Smartlead, Outlevo builds each campaign there and sends from your mailboxes [[12]](#sources). Before a campaign sends, every mailbox must be connected and warmed up [[12]](#sources). Its domain must also have SPF, DKIM and DMARC checked in the last 7 days [[12]](#sources).
Know what the authentication standards say. You do not need to write DNS records by hand if Outlevo does it, but you should recognize the terms when a check fails.
- SPF has a lookup limit. RFC 7208, checked 2026-10-22, says: "SPF implementations MUST limit the total number of those terms to 10 during SPF evaluation" [[13]](#sources).
- DMARC is a policy layer. RFC 7489, checked 2026-10-22, describes how "a mail-originating organization can express domain-level policies and preferences for message validation, disposition, and reporting" [[14]](#sources).
- Google's sender guidelines, checked 2026-10-22, say: "email senders who send more than 5,000 messages per day to Gmail accounts must meet the requirements in this section. Set up SPF and DKIM email authentication for your domain." [[15]](#sources) A small pilot is not near that line, but authentication is still checked before Smartlead sends [[12]](#sources).
Rehearse without sending. Test keys run against a sandbox with fake companies and contacts, and are not charged [[16]](#sources). The sandbox holds 100 fake companies, so you can build a workflow without sending anything or spending credits [[17]](#sources). If your team wants replies routed into other tools, rehearse that here. Webhook subscriptions cost 1 credit per delivered event, while failed attempts and test pings are free [[18]](#sources). The developer hub has the details.
Check the legal basics. The FTC's CAN-SPAM guide, checked 2026-10-22, says: "The law makes no exception for business-to-business email." [[19]](#sources) A B2B pilot is covered. Read the guide before week 3 and confirm that your email copy and unsubscribe handling meet it. Smartlead reports unsubscribes back to Outlevo by webhook [[12]](#sources), which you will see in week 4.
Week 3: A small send
Week 3 is the first real send. Keep it small.
Size it so you can read every reply. Outlevo's pages do not set a pilot size for you, so you choose one. For scale, one vendor, agentsdr.ai, defaults each mailbox to 30 emails a day (checked 2026-11-09) [[20]](#sources). Treat that as a reference point, not a target. The rule for your pilot is simple: the batch must be small enough that you can read every reply yourself.
Check the leads before they go. A Readiness Card needs at least three observations with a source and a date [[21]](#sources). Facts read from a website are kept only when their quote is found on the page, and guesses are labelled as hypotheses, never presented as findings [[21]](#sources). Open a handful of cards from your batch. Look for dated observations you could defend to the prospect. Remove any lead where you cannot.
Approve the copy. Read the sequence, every step of it, before it is scheduled. If you are new to the term, see sequence. Outlevo writes and schedules the emails, and Smartlead sends them from the mailboxes in your account [[12]](#sources).
Know the push mechanics. The campaign, its sequence and the leads are pushed to Smartlead in batches of 100 [[12]](#sources). Events then come back by webhook: sent, reply, bounce, unsubscribe and category changes [[12]](#sources). That is your review data.
Set up alerts. Telegram can send hot leads, approvals, problems that need you and credit alerts to your phone [[22]](#sources). Every alert also appears in the app [[22]](#sources).
Decide how replies are handled. Interested replies can be answered with your Calendly link [[23]](#sources) or your Cal.com link [[24]](#sources). On autopilot the answer goes at once. Otherwise it waits in Replies for you to send [[23]](#sources). For a pilot, keep a person in that loop so you see what the first interested replies look like before anything goes out under your name.
Find the kill switch before you send. Kill switches pause a campaign, a channel or the whole workspace at once [[2]](#sources). Know where each one is before the first email goes out.
Week 4: Review replies and decide
Week 4 is for reading, not for sending more. If your mailboxes finished warming late, this review may be short. Decide it still counts, and write down that the sample is thin.
Pull the events. Group what came back into sent, reply, bounce, unsubscribe and category changes [[12]](#sources). Then read every reply. Count each person once. A reply rate is only useful if you know what it divides by. Smartlead's State of Cold Email 2026, on more than 850 million emails sent from January to June 2026, puts the median cold email reply rate at 0.74% per contact [[25]](#sources). That is replies divided by people contacted, so a person who replies once counts once however many emails they were sent [[25]](#sources). The top quarter of senders reach 1.43% and the top tenth 2.63% [[25]](#sources). Use these as context only. At rates like these, a small send produces very few replies, so reading each reply tells you more than any percentage.
Sort the replies by what they say. Ask these questions of each one:
- Is the person in the segment you approved?
- Is the interest real, or a polite no?
- Did the message reach the right role?
Check the bookings. Each booking with one of your contacts counts as a meeting booked and is written to your CRM when one is connected [[24]](#sources). If your CRM is Salesforce, the connection needs the Outlevo package in your org, which Outlevo installs with you, and the integration is in pilot status [[26]](#sources). A rep's Accept or Reject on the Account is read by the 15-minute sync [[26]](#sources). Details are on the Salesforce integration page.
Audit the ledger. Every external action is recorded in a ledger [[2]](#sources). Open it and match what went out against the policy you approved in week 1. Any mismatch is a stop signal, even if the replies look good.
Write the decision. One page, three options: scale, adjust and rerun, or stop. The signals are in the next section.
What to check
Set your own numbers before the send. Outlevo's pages do not give bounce, complaint or positive-reply cut-offs for a pilot, so write down the ones you will accept in week 1 and hold yourself to them in week 4. Moving the goalposts after you see the results defeats the point of a pilot.
Go/no-go checks by stage
End of week 1: go to mailbox work only if all of these are true.
- One ICP segment is approved by a named person.
- The policy states who can be contacted, on which channels, how often and with what budget.
- You have told Outlevo which countries you sell into [[3]](#sources).
- If you imported deals, you know whether the sample is under 10 won companies.
- A credit budget is written down.
End of week 2: go to the send only if all of these are true.
- Every mailbox is connected and warmed up [[12]](#sources).
- Each sending domain has SPF, DKIM and DMARC checked in the last 7 days [[12]](#sources).
- Your email copy and unsubscribe handling are checked against CAN-SPAM [[19]](#sources).
- You have run any reply routing in the sandbox [[17]](#sources).
- You know where the kill switches are [[2]](#sources).
End of week 3: go to review only if all of these are true.
- Every lead in the batch has a Readiness Card with dated observations [[21]](#sources).
- Sent, reply, bounce and unsubscribe events are arriving [[12]](#sources).
- Interested replies are being handled by a person, or by a setting you chose on purpose [[23]](#sources).
- Nothing has run outside the approved policy [[2]](#sources).
End of week 4: decide. See below.
Signals that justify scaling
- Replies come from people inside the approved segment.
- At least some replies show real interest, and bookings are counted as meetings.
- Bounces and unsubscribes are being recorded and acted on.
- The ledger shows only actions the policy allowed.
- You can explain the result without relying on a percentage from a tiny sample.
If you scale, change one thing at a time: more contacts in the same segment, or a second segment, but not both.
Signals that justify stopping or rerunning
- Replies come mostly from people outside the approved segment. Revisit the ICP, not the copy.
- Warm-up or authentication checks fail again. Fix them before sending more.
- The ledger shows an action that does not match the policy. Pause with a kill switch and find out why.
- Readiness Cards were thin and you sent anyway. Fix the evidence step and rerun.
- You cannot read the result because the sample was too small. Rerun with the same segment, not a new one.
A rerun is a legitimate outcome. A pilot that tells you the segment is wrong has done its job.
Next step
Start with the thing that has a clock on it: mailboxes. Decide your sending domains and sender names this week, and review the managed mailboxes integration or the Smartlead integration, whichever matches how you send.
While that runs, work out the ICP. If you have no deal history, run the free ICP assessment on your site and pick one segment to approve. If you do have deal history, import it from a CSV or your CRM and check the confidence label before you trust the result.
Then write the policy and the go/no-go thresholds on one page, with a person's name next to each approval. That page is the pilot. Everything else is execution.
Sources
- Managed mailboxes integration
- How Outlevo works
- ICP from evidence
- CSV and Excel integration
- Release 2026-10-19
- HubSpot integration
- Free ICP assessment
- Release 2026-10-19
- LinkedIn User Agreement (section 8.2, Don’ts)
- LinkedIn integration
- Outlevo pricing
- Smartlead integration
- RFC 7208: Sender Policy Framework (SPF)
- RFC 7489: Domain-based Message Authentication, Reporting, and Conformance (DMARC)
- Google: Email sender guidelines
- Developer hub
- REST API and SDK integration
- Webhooks integration
- FTC: CAN-SPAM Act compliance guide for business
- agentsdr.ai
- Readiness Cards
- Telegram integration
- Calendly integration
- Cal.com integration
- Smartlead, The State of Cold Email 2026: reply rate per contact (top quartile; January to June 2026, more than 850 million emails)
- Salesforce integration