How to Build a B2B Outbound Prospecting System, Step by Step
The short answer
To build a B2B outbound prospecting system, write a weighted ideal customer profile from your best customers, tier accounts by fit and recent signals, keep verified contact data with a basis for each country, send from a separate domain with SPF, DKIM and DMARC, run a written cadence per tier, review numbers every Monday and hand off every held meeting.
Want the tools, not just the method? The The Outbound Prospecting Playbook kit turns this guide into ready-to-use templates, spreadsheets and printables, with an instant download.
- Start with five checkable, weighted ICP criteria, not a bought list
- Tier 1 is strong fit plus a live signal; it gets fully personal work
- Never send cold email from your main company domain
- Consent rules follow the recipient's country; follow the stricter rule when two apply
- One true reason, one small ask, under 120 words
- Change one thing a week and measure it
What are the steps to build an outbound prospecting system?
Outbound prospecting means reaching business buyers who have not asked to hear from you, by email, phone, LinkedIn, video and events, and turning some of those conversations into meetings. It works best as a system: written down, measured and repeatable by someone other than you. These eight steps are the order that avoids the expensive mistakes.
- 1. Score your current practice, so the biggest gaps set your order of work.
- 2. Write an ideal customer profile (ICP) of five checkable, weighted criteria, and a persona page for each buying role.
- 3. Tier accounts by fit and by recent public signals, so the best-fit accounts with a reason to act now get your most personal work.
- 4. Build lists from named sources, verify every email, record a basis or consent for each contact's country and screen against a suppression list.
- 5. Set up a separate sending domain with SPF, DKIM and DMARC, warm up new mailboxes and cap daily volume.
- 6. Write short emails, call openers, voicemails and objection answers built on one true reason and one small ask.
- 7. Run a written cadence per tier across email, phone, LinkedIn and video, stopping on any reply, opt-out or bounce.
- 8. Type the numbers every Monday, review for 30 minutes, change one thing, and hand off every held meeting within a day.
How do you define an ideal customer profile for outbound?
List your five best customers (fast to buy, happy, still paying) and three that went badly. Look for what the best share that the worst do not, using facts you can check from outside before you contact anyone: industry, size, number of sites, systems used, region. 'Innovative culture' fails that test; '20 or more inbound trucks a day per site' passes it.
Turn the pattern into five criteria, give each a weight so the weights add to 100, and describe what scores 3, 1 and 0. Every account then gets a fit score from 0 to 100. Check the ICP against your customers: most of your best should score high and none of your bad fits should. Criteria describe companies and roles, never people's protected characteristics.
How should you prioritise accounts and use trigger signals?
Effort should follow fit and timing. A signal is a public, recent change that makes your problem more urgent: a new site, a leadership hire, a revealing job advert, a contract win, an expansion, a technology change, new owners. Log each with its date and source, and let a signal count only while it is fresh (60 days is a reasonable starting window).
A simple rule works: Tier 1 needs a high fit score and a live signal; Tier 2 and 3 follow lower fit thresholds; everything else is 'not now'. Tier 1 gets fully personal research and messages, Tier 2 a personal first line, Tier 3 templates. In the worked example below, a founder's 120 accounts split into 20, 46 and 41 across the three tiers, with 13 parked.
Is cold email legal? The rules by country
The rules follow where the recipient is. In the United States, the FTC's CAN-SPAM guidance says the law makes no exception for business-to-business email, but it allows a first message without consent if you use honest headers and subject lines, say it is an advertisement, include a valid postal address and honour opt-outs within 10 business days.
Canada, Australia and New Zealand use consent models. Canada's CRTC explains that implied consent can arise from a conspicuously published address, with no statement refusing messages, where the message relates to the person's role. Australia's Spam Act has a similar published-address rule, and ACMA says unsubscribe requests must be honoured within five working days.
In the United Kingdom, the ICO explains that PECR lets you email corporate subscribers such as companies and LLPs without prior consent, as long as you identify yourself and offer an opt-out, while sole traders are treated as individuals; UK GDPR still applies to named people. In the EU, GDPR applies to named business contacts, with email rules for businesses set nationally: Germany's Unfair Competition Act requires prior express consent for email advertising, business recipients included. Check the current position with your adviser; these points were correct at November 2026.
| Country | First cold email to a work address | Opt-out timing |
|---|---|---|
| United States | Allowed if CAN-SPAM rules are met | Within 10 business days |
| Canada | Needs express or implied consent | Within 10 business days |
| United Kingdom | Companies: allowed under PECR; UK GDPR for named people | Promptly; keep a suppression list |
| Germany | Prior express consent | Not applicable to cold email |
| Australia | Needs consent, express or inferred | Within 5 business days |
| New Zealand | Needs consent | See DIA guidance |
What do SPF, DKIM and DMARC have to do with prospecting?
Inbox providers decide whether your email reaches the inbox. SPF lists the services allowed to send for your domain, DKIM signs each message, and DMARC tells receivers what to do when checks fail and sends you reports. Gmail's sender guidelines require all senders to use SPF or DKIM, and senders of more than 5,000 messages a day to use SPF, DKIM and DMARC with alignment, one-click unsubscribe and a spam rate below 0.30%, ideally below 0.10%. Yahoo has similar bulk sender rules and asks for unsubscribes within two days; Microsoft's Outlook.com began rejecting non-compliant high-volume mail on 5 May 2025.
Most outbound programmes never reach 5,000 a day, but the same practices protect everyone. Send cold email from a separate domain, never the one that sends your invoices; publish all three records before the first send; warm up new mailboxes for about two weeks; cap sends per mailbox per day; verify every list; and stop sending when your bounce rate passes the line you set.
Reading helps; measuring tells you what to work on. These AI-graded assessments on AssessAll pair with this topic:
- Prospecting Discipline (AssessAll)
- Value Articulation: Building the Buyer's Business Case (AssessAll)
- Negotiation Fundamentals: Concessions, Trades and the Walk-Away (AssessAll)
How do you write cold emails and calls that get answered?
A cold email earns a reply when it is relevant, brief and easy to answer. Use one true reason this person should care now, one small ask such as a 15-minute look or the name of the right person, and keep first touches under 120 words. Cut every sentence about yourself. Subject lines must match the content. Follow-ups should add something new each time and the last one should close the loop politely.
On the phone, open with permission ('this is a cold call: can I have 30 seconds to say why I rang?'), give the same reason, ask two questions about their situation, handle an objection once and respect the answer. Voicemails should be under 25 seconds and point to your email. 'Take me off your list' is an instruction, not an objection: stop and suppress the same day.
AI assistants can summarise a company's public pages and draft messages quickly. A person must check every fact against its source and read every message before it goes; invented personalisation damages trust faster than a generic email.
What does a worked example look like?
Here is one illustrative quarter for a fictional 12-person software company in Melbourne, run by its founder for 8 hours a week after two weeks of set-up. All figures are a composite, not a benchmark or a promise, and they reconcile with each other.
Three things stand out. First, the 20 Tier 1 accounts produced 9 of the 17 meetings: effort concentrated on fit plus timing paid back more than volume. Second, week 3 had a 7.5% bounce rate from one old, unverified data file; pausing for two days, re-verifying and removing 22 risky addresses brought the quarter back to about 1%. Third, the phone booked 5 meetings that email alone would have missed. The founder's own business case had estimated 6 to 18 meetings for the quarter; the actual 17 sat inside that range, which is why he used it to decide on a second quarter before hiring rather than as a forecast. Your numbers will differ by market, offer and list, so treat these as a worked calculation to copy, not a target.
| Measure | Calculation | Result |
|---|---|---|
| Prospects started | 20 x 4 + 46 x 3 + 41 x 2 contacts | 300 |
| Bounce rate | 15 bounces / 1,390 emails | 1.08% |
| Replies per prospect | 34 / 300 | 11.3% |
| Positive share | 14 / 34 | 41% |
| Meetings per 100 prospects | 17 / 300 x 100 | 5.7 |
| Show rate | 14 held / 17 booked | 82% |
| Acceptance rate | 8 accepted / 14 held | 57% |
| Meetings per 10 hours | 17 / 92 x 10 | 1.8 |
| Cost for the quarter | time at the founder's own hour value, tools, set-up and a risk allowance | A$16,470 |
What are the most common outbound mistakes?
Most outbound failures are avoidable. These are the ones that cost the most.
- Sending cold email from the main company domain, so a reputation problem reaches invoices and support mail.
- Importing a bought list without verifying it or recording where it came from.
- Using one consent rule for every country on the list.
- Personalising with flattery instead of a true, work-related reason.
- Asking for three things in one email.
- Automating LinkedIn actions against the platform's terms.
- Deleting people who opted out instead of keeping them on a suppression list.
- Sending AI drafts that nobody has read.
- Changing five things at once, so nothing can be learned.
- Hiring an SDR before the method is written and tested.
How do you measure and improve outbound every week?
Keep a short list of numbers: prospects started, bounce and complaint rates, replies per prospect and the positive share, connect rate, meetings per 100 prospects, show rate, acceptance rate, and opt-outs actioned late, which must be zero. Set your own targets from your first eight weeks rather than borrowing benchmarks from another market.
Every Monday, type the week's row, read the buyer's own words in replies and objections, choose one change and decide how you will know if it worked. Score your practice again each quarter. If you want an outside measure of your prospecting habits, AssessAll's Prospecting Discipline assessment is one option, and Jobulary's guide to the WOOP method helps turn the biggest gap into a plan. The Outbound Prospecting Playbook kit from Bodhih Training includes the workbook, review form and dated rules sheet used throughout this guide.

Run the whole system from one kit
The Outbound Prospecting Playbook gives you the workbook, scripts, forms, consent check and dated rules sheet behind this guide, with a worked example to copy.
Sources
- US Federal Trade Commission: CAN-SPAM Act, a compliance guide for business
- CRTC: Frequently asked questions about Canada's Anti-Spam Legislation
- UK Information Commissioner's Office: Business-to-business marketing
- EUR-Lex: Regulation (EU) 2016/679 (GDPR)
- Australian Communications and Media Authority: Avoid sending spam
- NZ Department of Internal Affairs: Commercial electronic messaging in New Zealand
- German Federal Ministry of Justice: Unfair Competition Act, section 7
- Google Gmail Help: Email sender guidelines
- Yahoo Sender Hub: Sender best practices
- Microsoft Outlook blog: New requirements for high-volume senders
More from the Bodhih family
Questions people ask next
How many touches should a B2B cadence have?
Enough to give each prospect a fair, varied chance without pestering them. A common pattern is 6 to 12 touches over two to three weeks across email, phone and LinkedIn, with more and more personal touches for your best-fit accounts. Stop on any reply, opt-out or bounce.
Do I need consent to send B2B cold emails in the UK?
Not to companies and other corporate subscribers under PECR, provided you identify yourself and offer an opt-out. Sole traders and some partnerships need consent or the soft opt-in, and UK GDPR applies to named people's data, so you need a lawful basis such as legitimate interests and privacy information.
Can I email someone whose address is on their company website?
In Canada and Australia, consent may be inferred from a conspicuously published work address when the message relates to the person's role and there is no statement refusing messages. In the US, CAN-SPAM rules apply instead. Being public does not by itself mean consent everywhere, so record where and when you found it.
Should I use my main domain for cold email?
No. Use a separate sending domain with SPF, DKIM and DMARC, so any reputation problem from cold email cannot affect invoices, support or password emails from your main domain.
What bounce rate is too high?
Set your own red line; many teams treat anything above a few per cent in a week as a signal to stop and clean the list. Inbox providers publish spam complaint limits rather than bounce limits: Gmail recommends keeping spam rates below 0.10%.
Can AI write my cold emails?
AI can draft and summarise, but a person should check every fact against its source, keep the reason true and specific, and read every message before it is sent. Never paste personal data into a tool your policy does not approve.
When should a founder hire a first SDR?
When there is a written, tested method and a baseline of numbers for the new hire to follow. Compare the cost and likely pipeline of hiring now against running another quarter yourself, using your own estimates.