Proof of Concept in Sales: How to Run One That Closes

Priya's proof of concept went well: six weeks on the customer's site and an operations lead who kept saying "the team loves it." Then she asked for the order. "Finance will want a business case," the lead said, "and a couple of other options." The deal slipped two quarters, then died. (Priya is a composite, not a real person.)

Nothing failed in the trial. What failed was the agreement before it: nobody had written down what "working" meant, who would judge it, or what would happen the day it passed.

That is the whole game in proof of concept sales: a POC closes deals when it tests a few measurable criteria the buyer chose, inside a fixed scope and date, with the next step agreed before day one. This guide gives you a test for when to say yes, a success-criteria template with a filled-in example, and questions for the meetings that decide it.

Running the POC kickoff on site? Try ParrotNotes free and keep the buyer's exact words about what "success" means.

What is a proof of concept in sales?

A proof of concept (POC) in sales is a short, scoped test in which the buyer checks that your product does a specific job in their environment before they commit to buy. A demo shows what the product can do. A POC shows it doing the buyer's job, on their data, material or site.

The term comes from engineering. NASA's technology readiness levels note that "often during TRL 3, a proof-of-concept model is constructed," well before a full working prototype. In sales it means the same: prove the idea on a small, controlled scale.

People use these words loosely, so agree on the term with your buyer early. What matters is the agreement behind it, not the label.

When to agree to a POC (and when to say no)

A POC costs both sides time and weeks of the sales cycle. Agree to one only when it will settle a real question between the buyer and a decision. Use these five gates:

1. There is a named question. "Will it read our labels at line speed?" is a question. "We'd like to try it out" is not.
2. Discovery is done. You know the problem, the cost of the problem and who is involved. A POC never replaces discovery; it tests what discovery found. (If you're not there yet, start with the MEDDIC sales process.)
3. The buyer will write down success criteria. If they won't define success, they can't recognize it either. Treat the refusal as a stall.
4. Someone with budget authority knows about it. The economic buyer doesn't have to attend, but they should know the test is happening and what it's for.
5. The next step is agreed if it passes. A purchase decision, a pilot, a meeting with procurement. Something concrete, with a date.

If two or more gates fail, offer something smaller: a reference call, a demo on their own sample (see our guide to the in-person product demonstration) or a visit to an existing customer. These answer most "will it work?" questions in days.

The proof of concept success criteria template

MEDDICC describes Decision Criteria as "a set of principles, guidelines and requirements that an organization uses to make a decision," and Metrics as "the quantifiable measures of value that your solution can provide" (see the MEDDPICC methodology). POC success criteria are where the two meet: a few decision criteria, each turned into a number you can test.

Good criteria follow four rules:

• Three to five, no more. Each extra criterion is another way to fail.
• Pass or fail. "Improves accuracy" can't fail. "99.5% of labels read on the first pass" can.
• A baseline. Measure today's number first, or you'll argue about it at the end.
• The buyer's words. "No more manual re-keying" goes in the template as said.

Copy this template into a shared document and fill it in with your champion before the POC starts:

PROOF OF CONCEPT AGREEMENT                    Version / date:

Buyer:                      Seller:
Champion (name, role):      Our POC owner:
Economic buyer informed? Y/N   (name, date)

The question this POC answers (one sentence, buyer's words):

SUCCESS CRITERIA (3 to 5)
#  Criterion | Baseline today | Pass threshold | How measured | Signs off
1
2
3

SCOPE
In scope:
Out of scope (goes to phase 2):
Data, material or site access needed, from whom, by when:

TIMELINE
Kickoff:            Midpoint check:            Exit meeting:
Hard end date:

IF ALL CRITERIA PASS, the buyer will:
(the next step, owner and date, agreed now)

IF ANY CRITERION FAILS:
(what happens: fix and retest once, or stop)

Signed off by (buyer):                 (seller):

The two lines at the bottom do most of the work. When the buyer writes "If all criteria pass, we'll make a purchase decision by March 14," they've decided in advance, and the exit meeting confirms it.

Worked example: a filled-in POC template

Daniel sells print-and-apply labelers. A food packer wants proof his machine can keep up on their busiest line before replacing three older units. (Daniel and the packer are a composite; the numbers are illustrative.)

The question: "Can it label case packs on Line 2 at full speed
without the re-keying we do today?"

Champion: Maria, operations manager
Economic buyer informed: Yes, plant director, Feb 3

#  Criterion                 | Baseline  | Pass          | How measured        | Signs off
1  First-pass label read rate | 96.8%     | 99.5% or more | Line scanner log    | Quality lead
2  Cases per minute labeled   | 28        | 32 or more    | Line counter, 3 shifts | Maria
3  Manual re-keying per shift | ~40 cases | 5 or fewer    | Shift lead tally    | Shift leads
4  Changeover between SKUs    | 22 min    | 10 min or less | Timed, 3 changeovers | Maintenance lead

In scope: Line 2, two SKUs, day and night shift
Out of scope: ERP integration (phase 2), Lines 1 and 3
Access: line time and scanner logs from Maria by Feb 7

Kickoff Feb 10 / Midpoint Feb 17 / Exit meeting Feb 24 / Hard end Feb 25

If all pass: plant director reviews the quote for three units,
decision by March 14. Owner: Maria.
If any fails: one fix and a two-day retest, then stop.

Notice what's missing: no "ease of use" and no open-ended "evaluate the product." Those belong in conversation, not in the pass/fail list. And the ERP integration IT asked about has a home, phase 2, so it can't quietly stretch two weeks into two months.

How to run a sales POC in four meetings

A POC is decided in a handful of conversations, usually on the buyer's site. Here's what to ask in each.

1. Kickoff: lock the agreement

Walk through the template with everyone who will judge the result.

• "If we hit all four numbers, what stops you from moving ahead?"
• "Who else will look at these results, and what will they want to see?"
• "Is a criterion missing, or one you'd quietly ignore?"

Write the answers down word for word. "If it passes, I'll take it to the director myself" is a commitment you'll want to quote later. (To test whether your contact is really a champion, see champion in sales.)

2. Midpoint check: fix problems while there's time

Review the numbers so far with the champion, so a failing criterion surfaces now, not at the exit meeting.

• "Which numbers worry you?"
• "Has anyone new joined the evaluation?"
• "Is the decision date still March 14?"

If a criterion is slipping, agree the fix in writing. If the decision date moved, find out why.

3. Exit meeting: read the results against the agreement

Put the template on the table and go criterion by criterion. Huthwaite, the firm behind SPIN Selling, describes this stage of a sale as obtaining commitment: securing "next steps that advance the sales process towards a successful conclusion for both parties."

• "Here are the results for each criterion. Do you agree with each one?"
• "You said a pass meant the quote review by March 14. Is that still the plan?"
• "What could stop the director signing off?"

Want the buyer's kickoff promise in their own words when the exit meeting comes? Download ParrotNotes free, record the kickoff on your phone with everyone's consent, and ask the recording what they said would happen if it passed.

4. Commercial step: move straight to the agreed next step

Send the results summary within 24 hours, with the agreed next step in the first line. Add the outcome to your mutual action plan so procurement, legal and security steps start right away.

Back to Daniel. At his exit meeting three criteria passed and changeover came in at 11 minutes, one over. Because the template said "one fix and a two-day retest," nobody reopened the evaluation. The retest came in at 9 minutes, and the quote reached the plant director on schedule.

Mistakes that turn a POC into free consulting

• No written success criteria. The buyer judges on feel, and feelings change when finance asks questions.
• Open-ended scope. Every "can it also do this?" becomes a new criterion. Send extras to phase 2.
• No end date. A POC that runs "until we've had a good look" never ends.
• The wrong people in the room. If the engineer who judges the result skips the kickoff, expect new criteria at the exit. Our guide to selling to engineers covers what they look for.
• The rep disappears. Keep the champion, the midpoint check and the decision date in your hands, even when a sales engineer runs the technical work.

Capture every POC meeting with ParrotNotes

The template holds the numbers. The meetings around it hold the promises and conditions that turn a passed POC into a signed deal.

ParrotNotes records in-person meetings on the phone you already carry, with no meeting bot in the room. Ask for consent first (rules vary; see our guide to one-party consent states). Recording runs in the background with the screen locked, even on a plant floor with no signal. Transcription runs once you are back online. You get a transcript, a summary and action items, and you can ask the recording "What did Maria say would happen if it passed?" The MEDDIC framework sorts what you heard into metrics, decision criteria and the decision process.

The Free plan includes 100 minutes of recording a month, a summary on every recording and 5 AI-powered recordings, where frameworks and chat work in full. Pro ($19.99/month, or $14.99/month billed annually) adds 3,000 minutes a month, recordings up to 3 hours, every AI feature on every recording and AI search across your notes, so every meeting of a six-week POC is one search away.

Run the POC you'd want to close

Proof of concept sales comes down to agreeing the end before the start. Say yes only when there is a named question, finished discovery and a buyer willing to write down success. Keep to three to five pass-or-fail criteria with baselines, a fixed scope and a hard end date. Ask at kickoff what happens if it passes, and let the exit meeting confirm that decision.

Copy the template into your next POC kickoff and fill it in with your champion. Download ParrotNotes free so the buyer's words from every POC meeting are there when you ask for the decision.