A sales process is not a CRM. It’s not a pipeline with stages. It’s not a list of steps that someone wrote down once and nobody reads. A sales process is the repeatable, documented system by which your business takes someone from first contact to signed contract, reliably, without the founder being personally required for every deal above a certain size.
Most B2B businesses at the £500k–£3M stage don’t have this. They have a founder who sells brilliantly and some support staff who help. That’s not a process; it’s a person. The distinction matters enormously when you want to scale, hire, or step back.
This post is about building the process from scratch: what it actually involves, in what order, and what to expect. It connects to Signs Your Sales Process Is Founder-Dependent if you want to understand the symptoms first, and to Pipeline Stage Definitions with Exit Criteria for the specific pipeline architecture piece.
Start with how you actually sell, not how you think you should
The most common mistake in building a sales process is designing it from scratch rather than extracting it from what’s already working.
If your founder closes 80% of qualified deals at a 30% margin, they’re doing something right. The goal isn’t to replace their approach with a generic framework; it’s to understand what they do that works, articulate it clearly enough for other people to replicate, and build the supporting infrastructure around it.
Spend time on this before you build anything. Interview the founder. Sit in on calls. Review won deals and understand why they closed. Review lost deals and understand what went wrong. Map the actual journey a deal takes from first contact to close. This is your raw material.
What you’re looking for is the pattern: what questions get asked in discovery, what objections come up and how they get handled, what makes the difference between a deal that moves forward and one that stalls. That pattern is the foundation of the process.
Define your ideal customer profile properly
A sales process only works on the right deals. Before you build the process, you need to be clear about who it’s designed to work with.
This isn’t a marketing exercise. It’s a sales infrastructure decision. The qualification criteria in your process (the conditions that determine whether a deal is worth pursuing) need to be grounded in a real understanding of which customers close, which ones close quickly, which ones go on to renew and expand, and which ones churn or cause problems.
If you’ve been selling for more than a year, you have data on this even if you haven’t organised it. Go through your won deals from the last twelve months. What do the best ones have in common (size, sector, growth stage, decision-maker seniority, use case)? What do the worst-to-close or worst-to-retain ones have in common?
The ICP you articulate should be specific enough to disqualify. If your ICP is “B2B SaaS companies,” it’s not an ICP. If it’s “B2B SaaS businesses between 20–150 employees, post-revenue, founder still in a sales role, dealing with inconsistent pipeline,” you can build qualification criteria around that.
Build the discovery framework
Discovery is where most deals are won or lost. It’s also where the most institutional knowledge lives in the founder’s head, and where it’s most critical to get the documentation right.
A discovery framework is a structured set of questions (not a script, but a guide) that ensures every discovery conversation covers the things that matter. At minimum, it should address:
The problem. What are they actually trying to solve? Not what they said when they asked for a call, but the underlying commercial problem that’s causing them to look for help now. Why now?
The cost. What is the problem costing them, in revenue, in time, in opportunity? This isn’t a pressure tactic; it’s the foundation of your value proposition. If the problem isn’t costing them much, you have a qualification problem.
The decision process. Who decides? Who else is involved? How have they made decisions like this before? What does their procurement or legal process look like? A deal that can’t close because you’ve only ever spoken to the wrong person isn’t a good deal.
The timeline. Is there a real deadline, or is this exploratory? What happens if they don’t solve this in the next six months?
The alternatives. Are they speaking to other providers? What would “doing nothing” look like? What would they do if you couldn’t help?
Write this down. Make it available to every person in the sales team. Build it into your CRM so reps can complete it against each deal record. The goal is that any rep running discovery is asking the same substantive questions, even if they phrase them differently.
Define qualification criteria
Qualification is the decision about whether a deal belongs in your pipeline. It’s not a judgement about whether the prospect is a good person; it’s whether this deal, at this time, with this prospect, is worth the resource of pursuing.
Most businesses qualify too loosely. Deals get into the pipeline because the prospect seemed interested, or because the company is the right size, or because the founder had a good call. That’s not qualification; it’s wishful thinking.
Proper qualification requires objective criteria that can be checked. If you use a framework like MEDDIC, BANT, or a variation of either, the criteria need to be specific to your deals, not generic. What does “confirmed budget” mean for your typical deal? What level of decision-maker seniority is sufficient? What does “confirmed timeline” look like in practice?
How to Qualify Beyond BANT covers qualification frameworks in depth, including why BANT falls short for complex B2B sales and what to use instead.
Build the CRM to reflect the process
Once you have the process (discovery framework, qualification criteria, stage definitions), configure your CRM to reflect it. Not the other way around.
This means: stage names that match your actual pipeline stages with documented exit criteria, properties that capture the qualification information you need, a deal structure that makes the state of each deal visible at a glance.
The test: could a new rep who’s never spoken to any of these prospects understand the state of every live deal from the CRM alone? If not, the CRM isn’t reflecting the process properly.
Establish the review cadence
A process is only a process if it’s being used and enforced consistently. The mechanism for that is the rhythm of pipeline reviews.
Weekly pipeline reviews with every rep. Not status updates; actual deal inspection. For each deal: what’s happened since last week, what’s the next step, what’s the close date based on the exit criteria, what help does the rep need? If a deal has been in the same stage for two weeks without a defined next step, something’s wrong.
Monthly win/loss reviews. What closed? What didn’t? What were the patterns? This is how the process improves over time, not theoretically, but based on what’s actually happening in your pipeline.
What “done” looks like
A finished sales process (one that’s actually working) has specific characteristics.
Reps can run discovery without the founder. Qualified deals are moving through the pipeline on the strength of the process, not the founder’s personal involvement. The forecast is produced from the CRM and is accurate within a reasonable margin. New reps ramp in three to four months rather than twelve. And the founder can take a week off without deals stalling.
It typically takes three to six months to build this properly and embed it. The building is faster; the embedding is slower because it requires habit change. But the output (a revenue function that doesn’t depend on one person) is worth every month of the effort.
For a real-world example of what this build looks like end-to-end, the Series A Revenue Rebuild case study covers the full sequence from diagnostic to functioning revenue infrastructure.
If you recognise that your business is at the founder-dependent stage and want to understand what building this looks like in practice, Discovery Week is where that conversation starts.