How to build a B2B sales pipeline from scratch (founder's playbook)

How do you build a B2B sales pipeline from scratch?

Short answer: you do not build a pipeline. You build a system that produces a pipeline. The pipeline is the output. The system is the thing you are designing — and it has five components: an ICP, a targeting layer, a channel mix, a qualification framework, and a measurement loop.

When founders ask how do I build a sales pipeline, what they usually mean is how do I get more qualified conversations on the calendar this month. The two questions have very different answers. The first is structural and lasts. The second is tactical and degrades the moment you stop running it. This post is about the first one.

TL;DR — the five components

Total realistic timeline from zero to a working pipeline: 8–12 weeks. Anyone selling you a 30-day version is selling you a list, not a pipeline.

Component 1 — Define the ICP (Ideal Customer Profile)

The ICP is not a vague description of "who we want to sell to." It is a precise filter that produces a finite, addressable list of companies.

A proper ICP has three layers:

Firmographic — industry, employee count, revenue band, geography, business model.

Technographic — what they use, what they do not use, what they recently bought.

Behavioural — hiring signals, funding signals, leadership changes, expansion signals.

Example, for a fictional B2B consultancy targeting UK and UAE-based IT services firms:

Companies in the IT services / managed services industry, headquartered in the UK or UAE, with 30–250 employees, £3M–£25M revenue, founded 4–15 years ago, currently using HubSpot or Pipedrive, who have hired at least one senior salesperson in the last 12 months and posted a "Head of Growth" or "Sales Director" job in the last 6 months.

That ICP returns roughly 400–600 companies on Apollo with verified contact data. That is the Total Addressable Market for this quarter's pipeline. It is small, finite, and approachable.

A typical founder ICP looks more like:

"Mid-sized B2B companies that need help with sales."

That returns 80,000 companies and produces no pipeline because it produces no list.

The test for a usable ICP: can you generate a list of 100 named companies that pass the filter? If yes, the ICP is concrete enough. If no, narrow it.

Component 2 — Build the targeting layer

Once the ICP exists, the targeting layer turns it into a list of named accounts and the named contacts inside them.

The minimum-viable targeting stack:

The output of this layer is a working spreadsheet (or, better, a CRM-loaded list) with five columns at minimum:

• Account name
• Industry + employee count
• Primary contact: name, title, email, LinkedIn URL
• Secondary contact: same
• "Why now" trigger: the specific signal that makes this account relevant this quarter

That "why now" column is the difference between a pipeline that converts and a pipeline that bounces. A list of 500 ICP-fit companies with no "why now" is a cold list. The same 500 companies with a documented trigger ("just hired Head of Sales," "raised Series A 8 weeks ago," "expanding into UAE office") is a warm-ish list that converts 3–5× better.

Component 3 — Pick the channel mix

There is no universal answer for which channels produce pipeline. There is only the right answer for a given ICP, deal size, and team capacity.

The realistic options:

The pipeline-from-scratch decision is almost always between three channels initially:

• Cold email + LinkedIn outbound — fastest to install, cheapest to run, most predictable for £500K–£5M ARR B2B service firms. This is the default.
• Warm intros — necessary but unscalable. Use it for the first 5–10 conversations to validate the ICP, then layer outbound on top.
• Inbound content — start it in week 1 but do not expect pipeline from it for 6+ months. It is a back-burner investment, not a primary channel.

Founders fail this stage when they try to run six channels at once with no team. Pick two. Run them well. Add the third in 6 months.

Component 4 — Build the qualification framework

The qualification framework is what stops the pipeline from becoming a list of vague maybes.

Pick one framework and use it consistently. The serious options for B2B:

• MEDDPICC — Metrics, Economic Buyer, Decision Criteria, Decision Process, Paper Process, Identify Pain, Champion, Competition. Heavy, complete, suited to enterprise sales with 6+ month cycles.
• SPICED — Situation, Pain, Impact, Critical Event, Decision. Lighter, suited to mid-market and shorter cycles.
• BANT — Budget, Authority, Need, Timing. Old, simple, often dismissed but perfectly adequate for founder-led sales at small deal sizes.

For most £500K–£5M ARR B2B service firms, SPICED is the sweet spot. MEDDPICC is overbuilt; BANT is underbuilt.

Whatever you pick, the framework should turn into stage exit criteria in your CRM. Example:

If a deal is sitting in "Demo" and the rep cannot point to a critical event, the deal is not in Demo. It is in Discovery wearing a costume. The framework forces honesty.

Component 5 — The measurement loop

A pipeline you cannot measure is not a pipeline. It is hope.

The seven metrics that matter for a pipeline-from-scratch:

Once you have these numbers stabilising — typically after 3–4 months of consistent activity — the pipeline is no longer "from scratch." It is a system you can scale by increasing the inputs.

The math — how much outbound produces how much pipeline

If your conversion rates are at the bottom of the typical range:

• 100 cold emails → 1.5 meetings booked → 0.45 opportunities → 0.07 deals closed
• £30K deal size → £2,100 in revenue per 100 emails sent

That sounds terrible until you scale it:

This is the math behind the question "do we have enough pipeline?" It is also the math behind "how many SDRs do I need?" — each SDR realistically handles 75–125 quality cold sends per day with proper personalisation, so building a £1M ARR pipeline by outbound alone means 4–6 SDRs at steady state.

A 90-day rollout

This is faster than typical "agency timelines" because it does not pretend to need 6 months. It is slower than what most founders hope for because pipeline math does not care about hope.

Common failure modes

The "I sent 50 emails and got nothing" panic. 50 emails produces, statistically, 0.75 meetings. The sample size is wrong. The fix is volume, not copy iteration.

The "I will write each email by hand" trap. This works at 10 sends/day and breaks at 50. The right move is templates with deep personalisation in the first sentence and last line only — the middle of the email is templated.

The "we tried outbound, it does not work for our industry" verdict at week 6. Outbound takes 12–16 weeks to produce a representative result. Verdicts before that are noise.

The "pipeline is just a list" misunderstanding. A list is not a pipeline. A pipeline is a list + a stage + an exit criterion + a forecast value. Most early pipelines in HubSpot are actually lists with stage labels stuck on top — they do not produce useful forecasts.

The "everyone is a potential customer" ICP. Always fatal. Always.

The "we will hire when we have pipeline" loop. You cannot generate pipeline if there is nobody to do the work. The founder runs the first 90 days personally or hires a fractional operator. There is no third path.

For UAE & KSA teams

A pipeline built into the GCC market needs structural adjustments that founders coming from UK/US/EU contexts almost always miss the first time.

Warm intros do more of the work. In KSA particularly, the first 30% of pipeline often comes through introductions, not cold outreach. Build the targeting layer with a "who introduces us to this account?" column from day one. Treat your existing UAE/KSA professional network like a CRM field, not an afterthought.

Multi-threaded by default. GCC B2B deals routinely have 6–8 stakeholders, and the formal decision-maker is frequently not the person you negotiate with. Plan to identify and engage two contacts per account from the start of outreach — a single-thread approach is a slow no.

Longer cycles, more touches. A typical UK B2B service deal closes in 60–90 days from first touch. A typical UAE deal closes in 100–140 days. A typical KSA enterprise deal closes in 150–220 days. Build the pipeline math with these cycle lengths or the forecast will be permanently off.

Ramadan compresses the year. Effective selling weeks in the GCC: roughly 38 per year, not 48. Ramadan is the obvious one (4 weeks of compressed productivity), but also Eid (1–2 weeks), Hajj for KSA-facing deals (variable), and summer (mid-June through August, when senior decision-makers travel). Plan pipeline ramps around these explicitly.

Local channels matter disproportionately. WhatsApp Business is a legitimate sales channel for follow-ups in the UAE/KSA. LinkedIn is strong in both markets — Saudi LinkedIn engagement is higher per-capita than most assume. X (Twitter) is still relevant for KSA government-adjacent business. Cold email works, but the channel mix should over-index on LinkedIn and warm channels compared to a UK/EU baseline.

Currency in dashboards. A pipeline serving UAE/KSA buyers and a UK/EU back-office accounting system needs CRM dashboards that handle AED/SAR/GBP cleanly. This sounds trivial; it routinely costs founders 2–3 weeks of reporting work to retrofit later.

What MAVEN does about it

Building a B2B pipeline from scratch is the spine of our Sales Process Program — the 90-day install includes ICP definition, list building, mailbox infrastructure, sequence library, CRM configuration, qualification framework, and live coaching. By the end of the engagement, the team is running the pipeline themselves and the founder is no longer the bottleneck.

If you are earlier than that — the ICP is unclear, the first 10 conversations have not happened yet — the Apollo Quick-Start is the standalone option: 4 weeks to a verified ICP list, a configured prospecting layer, and a first sequence live.

The Sales OS Blueprint and the Cold Email Playbook cover the architecture and the copy in more depth — both are free to read.

If you would rather talk it through, book a virtual coffee. 30 minutes, no slides, we look at your current state and tell you whether to start with outbound, partnerships, or content.

Frequently asked

How long until a brand-new pipeline produces closed revenue?

For a B2B service firm with a £10–30K average deal size and a 90-day cycle, the first closed deal from a freshly-built outbound pipeline typically lands in week 10–16. Faster than that means a warm intro accelerated it; slower than that means something in the system is broken.

Can I build a pipeline without a CRM?

For the first 50 conversations, a spreadsheet works. By conversation 50, you need a CRM. By conversation 100, a spreadsheet has cost you more in lost context than the CRM would have cost in subscription fees.

Cold email or LinkedIn — which one first?

Cold email if the ICP is broad and the contact data is reliable. LinkedIn if the ICP is narrow and high-trust (e.g. CEOs of competitor firms, partners at top-25 consultancies). For most B2B service firms, the answer is "both, with cold email producing the volume and LinkedIn producing the higher-quality 20%."

How many sequences do I need at the start?

Four. Cold (3–5 emails), follow-up after a meeting that went quiet, breakup (single-email last-shot), and re-engagement (90+ days after a closed-lost). More than four is premature optimisation. Fewer than four leaves response signals on the table.

Should the founder be sending the outbound personally?

Yes for the first 30–60 days. Founder-sent outbound has materially higher response rates than rep-sent outbound at the start, and the founder learns what works in a way no team report can convey. Hand off to a hired or contracted SDR around month 3.

What is a realistic close rate from a cold pipeline?

For B2B service firms: 1.5–3% cold-to-meeting, 25–35% qualified-opp-to-close. So roughly 0.4–0.9% cold-touch-to-closed-deal. Anything dramatically above this is either an exceptional brand or unreliable data.

---

Post 6 of 10 in our outbound + sales OS series.

Related reading

• What is a sales operating system?
• What is MEDDPICC?
• How to forecast B2B sales accurately
• How to build a B2B sales pipeline from scratch
• Healthy pipeline coverage ratio
• Sales discovery questions that work