Pilots, POCs, and the Pilot Trap
A free pilot is not a sales strategy — it is a way to give away professional services while a prospect gets comfortable saying no.
🎯 By the end you'll understand
The critical distinctions between a POC, a pilot, and a free trial — and why conflating them costs founders deals. The anatomy of a pilot that actually converts: paid, time-limited to 60–90 days, with mutually agreed success criteria defined before it starts, and a written conversion clause. How to use the Pilot Structure Designer to build a one-page pilot term sheet and identify structural risk. And the failure mode that kills more late-stage deals than almost anything else: the pilot that never ends.
KEY INSIGHTA pilot is a commitment device, not a free sample
Every month your pilot runs without a defined end date is a month the prospect is learning how to use your product without buying it — and discovering reasons not to.
Here is a situation that plays out in hundreds of B2B startups every year. The founder closes a "pilot" with a compelling enterprise prospect. The pilot starts. It goes reasonably well. The prospect asks to extend it "just one more month" to test an edge case. Then another month. Then there is a leadership change, a budget freeze, or the champion leaves. The deal dies — not because the product failed, but because the founder never established the structural conditions that force a decision.
A real pilot is a time-limited, paid, production deployment with mutually agreed success criteria and a written conversion clause. If any of those four conditions is missing, you do not have a pilot — you have a free trial in an enterprise disguise. The distinction is not semantic. It changes who has leverage at the end of the period, whether the prospect has skin in the game, and whether "success" is measurable or negotiable.
CORE CONCEPTThree pre-sale engagement types — and what each is actually for
Most founders use the words POC, pilot, and free trial interchangeably. They are not. Each serves a distinct purpose, involves different resources from the vendor, and carries different conversion implications. Misapplying the label — running a pilot when you should be running a POC, or giving away a pilot when the right answer is "buy or don't buy" — is a structural error that compounds through the rest of the deal.
| Engagement type | Purpose | Who pays | Duration | Output | Right for Forge? |
|---|---|---|---|---|---|
| POC Proof of Concept |
Prove technical feasibility in the prospect's environment — does your product work at all with their data, their infrastructure? | Usually vendor-absorbed (1–3 days eng time) | 1–2 weeks | Binary: technically feasible or not. Moves to pilot or to kill. | Yes — when a new Forge customer has non-standard sensors or edge infrastructure. Short, cheap, binary. |
| Pilot | Validate business value at limited scope in production — does it actually solve their problem, do users adopt it, is ROI real? | Prospect pays (50–100% of first-year ACV) | 60–90 days hard stop | Measured success vs. defined criteria → conversion decision | Yes — the standard pre-close engagement for Forge's $40–60K ACV deals. Must be structured correctly. |
| Free trial | Let prospects self-serve to reach an "aha moment" — appropriate for PLG SaaS with low ACV and no implementation complexity | Vendor absorbs | 7–30 days | Conversion or churn | Not appropriate. Forge ACV is $40–60K with hardware and integration. Free trials signal low value and attract tire-kickers. |
The key principle: the appropriate pre-sale structure depends on your product's complexity, implementation cost, and ACV — not on what the prospect asks for. A prospect who asks for a "free pilot" is not making a business request; they are making a negotiating move. The founder who grants it has already lost positioning.
HOW IT WORKSThe four non-negotiable elements of a pilot that converts
A pilot structure is not complicated — but every founder finds ways to skip one of the four required elements under prospect pressure. Each omission creates a specific failure mode that plays out at the end of the pilot period.
Paid — minimum 50% of first-year ACV, ideally 100%
Payment is not primarily about revenue — it is about commitment and signal quality. A prospect who pays $25K to run a pilot has made a budget decision, gotten internal approval, and signaled genuine intent. A prospect running a free pilot faces no switching cost at the end: saying no costs them nothing beyond an awkward email. The floor for a pilot payment is roughly $2,000 for any deal. Below that, the payment is too small to create genuine commitment and you are effectively running a free trial with extra steps. For Forge's ACV of $40–60K, the pilot price should be $20–40K — approximately equal to the first year of subscription.
Time-limited to 60–90 days with a hard end date
The pilot end date must be in writing before the pilot starts. Not "approximately 60 days from kickoff" — a specific calendar date. "Our pilot period concludes on September 15. On that date we will present our readout and ask for your conversion decision." This matters because the prospect's motivation to drive adoption internally is much higher when they know the clock is running. Without a hard date, the evaluation stretches indefinitely and the founder loses all leverage to drive a decision. 90 days is the practical maximum — beyond that, the evaluation becomes part of the operational background noise and urgency evaporates completely.
Mutually agreed, verifiable success criteria — defined before the pilot starts
This is the element founders skip most often, because nailing down success criteria requires difficult conversations about what exactly the prospect needs to see. Push through the discomfort. The success criteria must be: (a) specific and numeric — "reduce false escape rate by at least 20% vs. baseline" not "improve inspection quality"; (b) independently verifiable by the customer — they must be able to pull the data themselves without depending on your interpretation; and (c) documented in writing before Day 1. If you cannot agree on criteria before the pilot starts, you will have an unresolvable argument about whether the pilot succeeded at the end. The prospect will always find the criteria that your product did not meet.
A written conversion clause with a specific trigger and price
The pilot agreement must state: "If the success criteria in Section 3 are met, the Customer agrees to convert to a full subscription at $[X]/year, effective [date]." Without this clause, the prospect can acknowledge that the pilot succeeded and still restart the procurement process from scratch — requesting new proposals, involving new stakeholders, re-running their vendor comparison. With the clause, success equals conversion; you are not selling again after a successful pilot. The conversion price must also be specified in advance — do not leave room for post-pilot renegotiation. If you negotiate price before the pilot and they know the conversion number going in, you close the loophole where a successful pilot gets met with a request to discount.
The single most important conversation in a pilot setup is: "Before we begin, let's agree in writing on what success looks like. If we hit those numbers, I'll ask you to convert on [date]. What would we need to see for that to be an easy yes on your side?" If the prospect cannot answer this, they are not ready for a pilot — they are still exploring. Do not start a pilot with an exploring customer. Send them to a POC or schedule another discovery call to confirm the business case first.
SEE ITPilot Structure Designer — build your pilot term sheet
Run through the worst-case configuration: no success criteria, no conversion clause, 180-day duration, zero price. The widget shows you what that term sheet actually says — and why a prospect who accepts those terms is not your next customer, they are your most expensive distraction. Every flag the widget raises is a lever the prospect can pull at the end of the pilot to delay or avoid conversion. Remove the levers before the pilot starts.
FORGEStructuring Forge's pilot for conversion
Forge has 8 customers and $320K ARR. Every deal closed personally by the founder. The pattern is clear: Director of Quality as champion, VP Operations or COO as economic buyer, 60–90 day decision timeline triggered by a quality escape or audit failure. At this stage, the pilot structure is not just a deal tactic — it is a variable in the unit economics of growth.
Duration: 60 days hard stop. Start date and end date specified in the agreement before signature. No extensions without a signed amendment (rare, and only if new use cases justify it).
Price: $22,500 — approximately 50% of the $40–45K ACV for a single-line deployment. Credited toward the first year on conversion.
Success criteria (agreed before Day 1): (1) Forge detects ≥95% of defect instances that reached the manual inspector during the pilot period, validated against the customer's existing QA log; (2) False positive rate ≤2% of all flagged events; (3) Director of Quality is able to pull daily summary reports independently without Forge assistance by Day 30.
Conversion clause: "If the above criteria are met as of the pilot end date, Customer agrees to execute a 12-month subscription at $42,000/year, of which the $22,500 paid under this pilot agreement will be credited, resulting in a balance of $19,500 due within 10 business days of the pilot readout meeting."
Notice what this structure does. It forces the Director of Quality to be an active participant — criterion 3 requires the customer's own team to adopt the product, not just observe it. It makes the conversion decision binary at the readout meeting: the numbers either hit or they did not. And it eliminates price renegotiation by specifying both the pilot credit and the Year 1 ACV in advance.
All 8 of Forge's current customers went through structured pilots. The $320K ARR represents 7 converted and 1 currently live. Pilot-to-close rate has been 7/8 (87.5%) — high, because the success criteria were always hit. The one no-convert was a prospect where the champion left mid-pilot and the economic buyer had no context; Forge now requires a kickoff meeting with both champion and economic buyer before Day 1. The documented playbook captures this requirement.
WEAKNESSThe pilot that never ends — anatomy of a slow-motion deal death
The most common failure mode in pilot-led B2B sales is not a hard no at the end of the pilot — it is a soft yes that never converts. Here is the exact sequence:
Month 1: The pilot starts with vague success criteria
The founder was so eager to get the pilot started that the success criteria conversation was deferred. "We'll define exactly what success looks like once we see the data." The prospect agreed because vague criteria are comfortable — they cannot fail against what was never specified.
Month 2: The "extension" request
"The results are looking really promising, but we haven't had enough data from Line 3. Can we extend through end of month?" The founder says yes — it seems reasonable, and saying no feels like you're rushing a decision. The prospect has now established the precedent that an extension is available. The urgency that drove the evaluation in Month 1 has dissipated.
Month 3: Scope creep and new stakeholders
"Our VP of Operations wants to see how it works on our packaging line before we make a final decision." A new stakeholder who was never part of the original evaluation enters the process. They have none of the context the champion built over two months. The product now needs to prove itself again to someone who did not buy into the original criteria.
Month 4–5: Budget freeze / quarter close / champion departure
Any of these exogenous events, which would not have killed a deal with a hard end date, now kills this one. The prospect has been "evaluating" for five months with no skin in the game. They were never truly committed. The word "no" never had to be said — the deal just evaporated.
When a prospect asks to extend a pilot, the correct response is: "I want to make sure we get to a decision that works for you. Let's look at the data we have. Have we hit the success criteria we agreed on?" If yes: "Then we're in a position to convert now — let's get the paperwork moving." If no, but you are close: "Let's schedule the readout for Day 60 and present what we have. If you need one more data point, let's agree right now on what that specific data point is, and that you'll give me a conversion decision within 10 business days of seeing it." Never extend without a written amendment that specifies exactly what additional data is being gathered and a firm new decision deadline.
WEAKNESSWhere this bites
For some founders — particularly those with a technical background who find direct sales uncomfortable — a pilot is a way to postpone the closing conversation indefinitely. "If the product is good enough, they'll buy it." This is not how enterprise procurement works. The decision to convert is a separate event from the pilot performing well, and it requires its own selling motion: a readout meeting with the economic buyer, a presentation of ROI versus the criteria, and a clear ask. Founders who run great pilots and then wait for the prospect to call them with a purchase order are leaving most of their conversions on the table. A pilot sets up the closing conversation — it does not replace it.
If your product produces the measurement that determines whether the success criteria were met, you have a credibility problem at the readout meeting. "Our system detected 97% of defects" — says who? Your system. The prospect has no independent verification, and any friction in the relationship will cause them to question the numbers. Success criteria must be measurable using the customer's own systems or data sources, or against objective records (their QA log, their production data, their audit reports). Criteria that require trusting your instrumentation invite arguments at exactly the moment you need clean closure.
A POC is a technical feasibility test. If you have already proven technical feasibility for your ICP, running a POC every time is a waste of your resources and a delay of the conversion timeline. Conversely, starting a pilot before you know the product will work in the customer's environment is setting yourself up for a pilot that fails for technical reasons unrelated to business value. The sequencing question — POC first or straight to pilot — should be resolved based on infrastructure standardization. For Forge's 8 existing customers, all running similar sensor infrastructure, the answer is straight to pilot. For a customer with a genuinely novel production line configuration, run a 2-week POC first, absorb the cost, then move to a paid pilot.
KEY INSIGHTHow this connects
- The Discovery Call — Learning While Selling (4.4) — the discovery call is where you establish the business case that justifies a pilot. If you cannot articulate the cost of the customer's current problem in numbers during discovery, you cannot write credible success criteria for a pilot. The two are linked: discovery sets the baseline, pilot measures against it.
- The Buying Committee — Understanding Everyone Who Can Kill Your Deal (4.5) — the buying committee is the reason a conversion clause and pre-specified price matter so much. Without them, the economic buyer who was not involved in the pilot can restart the procurement process. Map the committee before the pilot starts and get the economic buyer into the kickoff meeting.
- Handling Objections — The Real Ones vs. the Smoke (4.7) — "we need more time to evaluate" is an objection, not a request. Lesson 4.7 covers the full set of enterprise objections and the responses that move deals forward without conceding structural leverage.
The pilot structure is where the entire GTM motion crystallizes. Everything learned in discovery (4.4) about the business case feeds the success criteria. Everything known about the buying committee (4.5) determines who to get in the room at the kickoff and the readout. And the objections that will come during the pilot extension conversation (4.7) are handled with frameworks, not improvised capitulation. Get the structure right before Day 1, and the closing conversation becomes a readout — not a sale.
GLOSSARYKey terms introduced
- POC (Proof of Concept)
- A short (1–2 week), typically vendor-absorbed technical test to establish whether the product works at all in the prospect's environment. Binary output: technically feasible or not. Precedes a pilot when the customer's infrastructure is non-standard.
- Pilot
- A time-limited (60–90 day), paid, production deployment with mutually agreed success criteria and a written conversion clause. Distinct from both POC and free trial. The appropriate pre-sale structure for B2B products with mid-to-high ACV and real implementation complexity.
- Success criteria
- Specific, numeric, independently verifiable targets agreed in writing before the pilot begins. Determine whether the pilot is declared a success. Must be measurable using the customer's own data sources or systems — not solely the vendor's instrumentation.
- Conversion clause
- A provision in the pilot agreement specifying that if defined success criteria are met, the customer agrees to convert to a full subscription at a stated price within a stated timeframe. Eliminates the need to re-sell after a successful pilot and closes price renegotiation as a post-pilot option.
- The pilot trap
- The failure mode in which a pilot runs indefinitely without a decision because it lacks one or more structural elements (hard end date, success criteria, conversion clause, or financial commitment). The prospect remains in evaluation mode without urgency; the deal dies to entropy rather than a stated objection.
- Pilot credit
- The standard commercial practice of applying the pilot fee paid toward the first year of a full subscription on conversion. Incentivizes early commitment, removes the perceived financial penalty of paying for the pilot, and is revenue-neutral over the customer lifetime.
- Baseline data lock
- The process of establishing and documenting the pre-pilot performance of the customer's current approach (e.g., defect escape rate with manual inspection) before Forge's system is deployed. Required to make success criteria measurable — without a baseline, "20% improvement" is unmeasurable.
ACTIVE RECALLCheck your understanding
Answer in your own words first, then reveal.
1A prospect says "we'd love to do a free pilot before committing to a purchase." What do you say, and why?›
You do not agree to a free pilot. The correct response is something like: "I want to make sure this is worth your team's time and ours — a real pilot means deploying into production and getting data that answers a real question. To do that well, we need your team's commitment, and we signal our commitment by treating it as a real engagement, not a free sample. Our standard pilot is 60 days at $22,500, credited toward your first year if you convert. That structure means we're both invested in making it work. Let me send you the pilot agreement and we can review the success criteria together."
Why: A free pilot gives the prospect zero skin in the game, signals that your product is worth zero until proven, and eliminates the most powerful commitment device available before close. Framing the paid pilot as mutual investment — not vendor rigidity — is the right positioning.
2What makes a success criterion "independently verifiable by the customer" and why does it matter?›
A criterion is independently verifiable if the customer can measure it using their own existing data systems without depending on the vendor's interpretation or instrumentation. For Forge: "Forge detects ≥95% of defects" is NOT independently verifiable if Forge's system is the only source of that number. "≥95% of defects that appear in the customer's QA inspection log for the same period are also flagged by Forge within 4 hours" IS independently verifiable because the customer's QA log is their own record.
Why it matters: At the readout meeting, if the numbers are close to the threshold, any vendor-provided metric will be questioned. If the customer believes the numbers, the conversion decision is easy. If they have to trust you, there is friction. Eliminate the friction in advance by designing criteria the customer can confirm themselves.
3A Forge pilot reaches Day 60. The results show 93% defect detection vs. the 95% criterion. The champion wants to extend by 30 days. How do you handle this?›
First, resist the instinct to immediately agree to the extension. Instead: "I appreciate that — the numbers are really close and I understand why you'd want more data. Let me ask: is the 93% number materially different from 95% in practice for your quality program, or is this about the contractual threshold we set? If it's the former, let's talk about whether the business case is still strong — I think it is. If it's the threshold, I want to understand what specific additional scenario you need to see. Let's define that right now, in writing, with a firm 15-day window and a conversion decision at the end."
Do NOT agree to a general 30-day extension. Instead, write a specific amendment: "Customer and Vendor agree to a 15-day extension to test Forge on Line 3 (previously untested) with the criterion that Forge detects ≥95% of defects logged by QC on Line 3 during this period. If met, Customer will execute the full subscription agreement within 5 business days of the extension end date." Specific scope, specific timeline, specific decision trigger.
4Why should the economic buyer attend the pilot kickoff meeting, even if they are not involved in the day-to-day evaluation?›
Two reasons. First, to establish shared ownership of the success criteria. When the economic buyer is in the room on Day 1 and hears the criteria agreed to out loud, they are a party to the evaluation — not someone receiving a summary 60 days later. This dramatically reduces the risk of the post-pilot "I need to revisit this with leadership" response, because leadership was already there.
Second, to identify the conversion process in advance. Ask the economic buyer at kickoff: "When the pilot concludes and the data looks as we're planning for, what does the contracting process look like from your side? Who signs, and what's the typical timeline?" This surfaces procurement delays (legal review, procurement portal requirements, signatory approvals) before they become end-of-pilot surprises. If conversion takes 45 days on their end, you want to know that on Day 1, not Day 60.