Sales Process Validation SOP

A sales process validation SOP checks whether the way you sell actually turns interested prospects into signed customers, and shows you exactly where deals stall. Setup takes two days and the test runs for six to ten weeks of live deals. Founder-led sales feels like talent and luck until you write the stages down. Once you do, you see that most lost deals die at the same step, and that step can be fixed.

What is the sales process validation SOP and when should you use it?

A sales process is the series of stages a deal moves through, from first contact to signed contract, with a clear rule for when a deal moves to the next stage. Validating it means running real opportunities through those stages, measuring conversion between each pair, and fixing the weakest link.

This SOP is for B2B sales with calls, demos or proposals. Self-serve products with a signup button need the customer onboarding validation SOP instead. Finding leads is the job of customer acquisition channel validation; this SOP starts when a lead says yes to a first call.

Use it:

• After your first 3 to 5 customers, when you want to know if you can repeat those wins.
• When deals stall and you can't say why.
• Before you hire a salesperson, who needs a written process to follow.
• When the sales cycle threatens your runway and you need to shorten it.

What do you need before you start?

• A list of stages you think your buyers follow, even if it is a guess.
• A free CRM such as HubSpot CRM or Pipedrive's trial, set up with those stages.
• Call recording with consent, through Fathom, Gong or your video tool.
• A lead flow of at least 20 opportunities over 6 to 10 weeks.
• Your pricing, ideally tested with pricing sensitivity testing.

How do you run it, step by step?

1. Write the stages and exit rules

A simple B2B process has five stages: discovery call, demo, proposal, negotiation, closed (won or lost). Give each stage an exit rule that a buyer action triggers, such as "buyer agreed to a demo with their manager present". Seller activity never moves a deal forward on its own.

2. Choose a qualification framework

Pick one and use it on every discovery call. BANT (Budget, Authority, Need, Timeline) came out of IBM and works for simple sales. MEDDIC (Metrics, Economic buyer, Decision criteria, Decision process, Identify pain, Champion) was developed at PTC in the 1990s and fits deals where several people sign off. Turn the letters into required CRM fields.

3. Set up the pipeline and the rules for logging

Every opportunity gets a stage, an expected value, a next step with a date and a source. No next step means the deal is dead, and you mark it lost instead of letting it rot in the pipeline.

4. Run at least 20 opportunities through it

Sell as you normally would, and log every move the same day. Twenty is a floor: fewer deals and one lucky win changes your win rate by 10 points.

5. Record calls and tag objections

Record discovery calls and demos with permission. After each one, tag objections in a short list: price, timing, missing feature, wrong person, existing tool. Neil Rackham's SPIN Selling, built on research into 35,000 sales calls, found that in larger sales, the most successful sellers asked more questions about the buyer's problem and its consequences, and spent less time listing features.

6. Find the leak

After 20 opportunities, build a stage conversion table. The stage with the biggest drop relative to your plan is your leak. Read the call notes and objection tags for the deals lost there.

7. Fix one thing and run 10 more deals

Change one thing at the leaking stage: a required meeting with the economic buyer before any proposal, a paid pilot instead of a free trial, a shorter demo built on the buyer's own data. Run at least 10 new opportunities and compare.

8. Write the one-page sales playbook

Stages, exit rules, qualification questions, the top five objections with your best answers, and the numbers you expect at each stage. This page is what you hand to your first sales hire.

Template you can copy

PIPELINE STAGES AND EXIT RULES
Stage          | Exit rule (buyer action)                     | Target conversion
Discovery call | Buyer confirms problem + agrees to demo      | 60%
Demo           | Buyer asks for proposal, economic buyer named| 60%
Proposal       | Buyer reviews it with economic buyer         | 50%
Negotiation    | Verbal yes + contract sent                   | 80%
Closed won     | Contract signed, first invoice paid          |

OPPORTUNITY LOG (one row per deal)
Company | Contact role | Source | Date created | Stage | Deal value |
Next step + date | M | E | D | D | I | C (MEDDIC fields, Y/N) |
Objections (price / timing / feature / wrong person / tool) |
Date closed | Won/Lost | Reason

STAGE CONVERSION TABLE
Stage | Deals entered | Deals moved on | Conversion % | Target % | Gap

Worked example: compliance software for small logistics firms

This case is illustrative. A founder sells compliance tracking software to logistics companies with 10 to 50 trucks at €4,800 a year. She runs 24 opportunities over 8 weeks.

Her stage table: 24 discovery calls, 15 demos (63%), 11 proposals (73%), 3 closed won (27% of proposals). Average cycle: 71 days. The leak is clear: proposals stall. Call notes show that in 6 of the 8 lost proposals, the finance director, the real economic buyer, never joined a call.

She changes one rule: no proposal until the economic buyer has been on a call. Over the next 12 opportunities, fewer deals reach the proposal stage (7), but 4 of those 7 close, and the average cycle drops to 52 days. She writes the rule into her playbook as the exit rule for the demo stage.

What should you measure, and when do you stop?

The process is validated when win rate and cycle length hold for two months in a row with at least 30 opportunities logged. If the leak moves to a new stage each month, you are still learning, and that's fine; keep fixing one stage at a time.

Which mistakes should you avoid?

• Moving deals on seller activity. "I sent the proposal" is your action. The deal moves when the buyer acts.
• Changing three things at once. You won't know which change worked.
• Skipping the economic buyer. A happy user with no budget authority produces a lost deal 4 weeks later.
• Keeping zombie deals. No next step, no deal. Close them as lost and record why.
• Writing the playbook from memory. Write it from the CRM data and call recordings.

How can AI speed this up?

Feed call transcripts from Fathom or Gong into an AI assistant and ask it to fill in the MEDDIC fields, list objections with exact quotes and flag calls where the economic buyer was never named. Ask it to draft better answers to your top five objections. The human check: read the actual calls behind any pattern before you act on it, because AI summaries miss tone, hesitation and the silence after you mention price.

Where does this SOP fit in your validation plan?

Sales process validation follows your first paying customers and comes before hiring:

1. Know the buyer's problem from customer interviews and set the price with pricing sensitivity testing.
2. Find leads with customer acquisition channel validation.
3. Run this SOP to prove you can close them.
4. Check that each deal earns money with unit economics validation.

See all 10 Go-to-Market Validation SOPs. For the short spoken pitch at the start of every discovery call, try the free elevator pitch generator. To practise sales conversations in a safe setting first, the Fe/male Switch startup game puts first-time founders through real tasks with feedback.

FAQ

How many deals do I need to validate a sales process?

At least 20 opportunities to find the leak, and 30 or more before you call the process validated. Fewer deals and one lucky win distorts every number.

Should I use BANT or MEDDIC?

BANT for simple deals with one buyer and a short cycle. MEDDIC for deals above a few thousand a year where several people sign off.

When should a founder hire the first salesperson?

When you have a written playbook, a win rate that held for two months, and more qualified leads than you can handle yourself. A salesperson repeats a process; they can't invent one for you.

What is a good sales cycle length for a startup?

One that fits your runway. If your cycle is 90 days and you have 6 months of cash, you get about two tries per prospect before the money runs out.