How do you calculate the true cost of a free trial motion vs a pilot program?

What a free trial motion and a pilot program actually are

A free trial motion gives a prospect self-serve access to the product for a fixed window — commonly 7 to 30 days — with little or no human involvement. The buyer provisions themselves, learns the product on their own time, and either converts on a credit card or a light sales assist, or they lapse. The commercial terms are usually untouched: the prospect converts at list price or at whatever standard discount your rate card already allows. The motion is designed to scale horizontally, meaning the hundredth trial costs you almost the same marginal amount as the tenth.

A pilot program is a structured, time-boxed evaluation with named success criteria, usually a signed pilot agreement or an order form with a defined term, and almost always a human on your side owning the outcome. Pilots commonly run 30 to 90 days. They involve configuration work, sometimes a real integration, a kickoff meeting, a midpoint check, and an exit readout against the criteria that were agreed at the start. The pilot scales vertically — you go deeper on fewer accounts, and the marginal cost of the tenth concurrent pilot is roughly the same as the first, because each one consumes another block of a human's week.

The distinction matters for costing because the two motions load their costs in completely different places. A trial's cost is dominated by volume: infrastructure for provisioned environments, marketing spend to fill the top of the funnel, some support tickets, and a very large denominator of people who were never going to buy. A pilot's cost is dominated by labor and time: solutions engineering hours, customer success hours, project management, legal review of the pilot agreement, and the executive alignment meetings that nobody puts on a spreadsheet but everybody attends.

There is a third shape worth naming because teams often mislabel it: the paid pilot or paid proof of value. Here the buyer pays a reduced fee, sometimes creditable against the first-year contract, in exchange for the structured evaluation. It behaves like a pilot operationally and like a small deal financially. If your organization runs these, cost them as pilots but net out the pilot fee — and be honest that a creditable fee is not revenue, it is a discount with a delay. Similarly, "reverse trial" motions — where a new account gets full premium functionality for the first two weeks and then drops to a free tier — should be costed as trials with a longer infrastructure tail, since those accounts keep consuming storage and compute long after the evaluation ends.

The reason people ask how to calculate this rather than simply picking one is that the two motions are not substitutes across all segments. Self-serve trials tend to fail above a certain deal size and complexity, because the buyer cannot evaluate the product without connecting it to their own data, and connecting it to their own data requires someone from your side. Pilots tend to fail below a certain deal size, because the labor cost of running them exceeds the gross profit on the resulting contract. The calculation is really about finding where those two curves cross in your business.

Choosing the motion by segment, complexity, and payback math

The decision is not a matter of taste. It comes down to three questions, answered in order, and each one can eliminate a motion outright before you get to the cost math.

Question one: can a competent buyer reach a genuine "aha" moment without a human? Not a demo moment — a moment where they see their own data, or a realistic proxy of it, doing something useful. If the product needs a warehouse connection, an SSO handshake, a permissions model mapped to their org chart, or a data migration before anything meaningful happens, the trial will produce a wall of stalled accounts and a support queue. If a motivated buyer can get there in under an hour with documentation and in-app guidance, a trial is viable.

Question two: is the gross profit on the resulting contract large enough to absorb the labor? This is the hard gate for pilots. Take your expected first-year contract value, multiply by gross margin, and compare it to the fully loaded pilot cost divided by the pilot-to-paid conversion rate. If a pilot costs you $8,000 in labor and converts at 50%, every won customer carries $16,000 of pilot cost. On a $25,000 first-year deal at 80% margin, that is $20,000 of gross profit against $16,000 of pilot cost — technically positive, but it leaves nothing for the rest of your sales and marketing spend, and your payback period will be measured in years.

Question three: what does the buying committee actually require to sign? Some purchases are blocked not by the buyer's belief but by their procurement process. Regulated industries, security-reviewed categories, and anything touching customer PII often *require* a documented evaluation with sign-offs. In those cases the pilot cost is not optional overhead — it is the price of admission, and the comparison is pilot versus no deal, not pilot versus trial.

Running these in sequence prevents the most common error, which is picking the motion based on what the competition does or what the last company the VP worked at did. The right answer varies by segment inside the same company. It is entirely normal for a business to run trials for everything under $15,000 ACV, hybrid trial-plus-assist between $15,000 and $60,000, and structured pilots only above that — with the thresholds recalculated once a year as loaded labor costs and conversion rates move.

A note on the last box: motions are not permanent. The most disciplined revenue teams review motion assignment quarterly against actual cost per conversion, and they are willing to move a segment from pilot to trial once product maturity closes the setup gap. The reverse also happens — a segment that used to convert on self-serve stops doing so as you move upmarket, and the trial quietly becomes a very expensive way to generate unqualified support tickets.

Building the two cost models line by line

Both models resolve to the same denominator: cost per converted customer. Everything below feeds that number. Use loaded labor rates, meaning salary plus benefits plus payroll tax plus a share of overhead — typically 1.25 to 1.4 times base salary — not the raw hourly wage. If you skip that multiplier, every people-heavy motion looks cheaper than it is.

Free trial cost stack, per 100 trials started. Infrastructure and provisioning: for most SaaS products this lands somewhere between a few cents and a few dollars per trial account per day, depending on whether you are spinning up isolated environments or multi-tenant rows. If a trial account costs $1.50 a day to host and runs 14 days, that is $21 per trial, or $2,100 per hundred. Support and onboarding touch: even a "no-touch" trial generates tickets — budget 0.3 to 1.0 support contacts per trial at 15 to 25 minutes each; at a loaded support rate of $45 an hour, that is roughly $5 to $10 per trial. Marketing acquisition cost to generate the signup, pulled straight from your paid and organic blended CAC per trial start. Sales assist, if any: a fifteen-minute qualification call and two follow-up emails at a loaded AE rate of $80 to $120 an hour is $30 to $45 per trial. Product and engineering time spent maintaining the trial experience — sandbox resets, demo data seeding, expiry logic — amortized across trials started, which for a mature product is small but is never zero.

Then apply the conversion rate. Self-serve SaaS trial-to-paid rates vary enormously by whether you require a credit card up front, but the honest planning range for opt-in (no card) trials is low — often under 15% — while card-required trials convert far higher on a much smaller top of funnel. Whichever you run, the arithmetic is the same: total stack cost divided by conversions. If 100 trials cost you $6,000 all-in and 12 convert, your trial cost per customer is $500.

Pilot cost stack, per pilot run. Solutions engineering or implementation: this is the big line. A typical structured pilot consumes 5 to 15 hours a week of combined SE and CS time during the active period. At a loaded SE rate of $100 to $200 an hour and a CSM rate of $80 to $150, a 12-week pilot at the low end of hours still lands in the $5,000 to $15,000 range in internal labor before anything else. Account executive time: kickoff, weekly or biweekly checkpoints, the exit readout, and the contract conversation — 20 to 40 hours across a quarter. Legal and security review: pilot agreements, DPAs, and security questionnaires are real work, and if they route through outside counsel they are real invoices. Custom configuration or integration work that will not be reused for other accounts. Executive sponsor time on both the kickoff and readout, which is expensive per hour and easy to forget. Free or discounted product usage during the pilot term, valued at the price the customer would otherwise pay.

Then apply the pilot conversion rate, which is normally much higher than a trial's — pilots are self-selecting, since a buyer who signs a pilot agreement and assigns internal resources has already cleared several qualification hurdles.

The three costs almost everyone omits. First, forgone revenue during the evaluation window. If a $30,000 ACV deal spends an extra six weeks in evaluation instead of in contract, that is roughly $3,500 of revenue shifted out of the period — and if it happens across a dozen concurrent evaluations, the working-capital effect is material even though nothing was "lost." Second, the discount that pilots invite. Buyers who negotiate pilot terms very frequently negotiate pilot pricing alongside them, and the resulting rate tends to anchor the first renewal too. If you convert a pilot at 75% of list and never fully recover it, a three-year relationship on a $30,000 list ACV leaves roughly $22,500 on the table — which can exceed the entire labor cost of the pilot. Model a pricing penalty explicitly rather than pretending list price holds. Third, opportunity cost of the team's time. Every hour an SE spends on a pilot is an hour not spent on a deal already in late stage. Value it at the gross profit that hour would otherwise have produced, not at zero.

The learning-loop asymmetry, which is a cost even though it never appears on a P&L. A hundred concurrent trials produce a hundred independent, unassisted data points on where people get stuck — activation drop-off, the feature nobody finds, the step where sessions end. You get that signal in two weeks. Four pilots produce deep, articulate, high-context feedback from named humans, but you wait a quarter for it and the sample is tiny. Worse, pilots are prone to what is sometimes called success theater: your team over-invests to make the pilot look good, which papers over the exact friction that would have shown up unassisted. The hidden cost is a roadmap built for a product that only works with white-glove service — a model that does not scale, and one you often do not discover until you try to sell it without the gloves on.

Instrumenting it in the CRM and sequencing the rollout

A cost model that lives in a spreadsheet on someone's laptop degrades within a quarter. The RevOps job is to make the inputs fall out of systems you already run, so the model refreshes itself.

Fields and objects to add. On the Opportunity, add a picklist for evaluation motion — none, trial, trial with assist, pilot, paid pilot — set at the point the evaluation begins, not backfilled at close. Add date fields for evaluation start and evaluation end so cycle length is computed, not estimated. Add a numeric field for pilot fee, if any. On the Account or a related custom object, log support contacts during the evaluation window so the support-touch cost stops being a guess. Make the motion field required before an opportunity can move past the stage where evaluations begin — validation on save beats a cleanup project six months later.

Where the labor numbers come from. You do not need a timesheet system, but you do need something better than memory. The workable middle ground is a lightweight activity log: SEs and CSMs tag calendar events and logged activities to the opportunity, and you roll up hours monthly. Even at 70% capture accuracy this is dramatically better than the alternative, which is a VP asserting that pilots "take about a week of someone's time." Pair it with a quarterly calibration where you take five completed pilots, reconstruct the true hours by hand, and compare to the logged number. Apply the resulting correction factor to everything else.

Sequencing matters more than completeness. Do not attempt to instrument every cost line in the first pass. Start with the three that move the answer most — labor hours, conversion rate, and evaluation duration — and get those right on one segment for a full quarter. A model with three accurate inputs beats a model with fourteen inputs where nine are invented.

The review cadence. Once a quarter, put one table in front of the revenue leadership team: for each segment, cost per conversion by motion, average evaluation length, conversion rate, and realized price as a percentage of list. Four columns, one row per segment. The conversation should be about whether any segment has crossed a threshold, not about methodology — methodology gets argued once, at the start, and then frozen for at least two quarters so the numbers are comparable across periods.

Adjacent effects worth watching. Motion choice ripples downstream in ways that show up months later. Trial-sourced customers tend to arrive with lower initial spend and expand over time, which flattens the first-year number and makes net revenue retention the more honest scoreboard. Pilot-sourced customers arrive larger but with implementation expectations already set high, which loads the post-sale team. Upstream, marketing needs to know which motion a segment runs before it builds the campaign, because "start free" and "request an evaluation" are different landing pages, different lead scoring, and different SLAs on follow-up. And support needs a heads-up before a trial motion scales, because unassisted users generate a ticket profile — short, repetitive, documentation-shaped — that is different from the deep technical questions a pilot produces.

When leadership pushes back. Someone will argue that pilots "always close better" and that the labor is worth it. They may be right for their segment. The way to settle it is not debate but the same table: pilot cost per conversion versus trial cost per conversion, in that segment, over the last four quarters. If pilots win, keep them and stop apologizing for the SE hours. If they do not, the conversation shifts to what would have to change in the product for a trial to work — which is a far more productive conversation than arguing about motion religion.

Related questions

Should I require a credit card to start a free trial?

Card-required trials convert at a much higher rate but shrink the top of funnel sharply. The right choice depends on whether your bottleneck is volume or quality. Test it as a genuine experiment on a single segment, and compare cost per converted customer, not signup counts.

How long should a pilot program run?

Long enough for the buyer to hit their stated success criteria, and no longer. Most fall between 30 and 90 days. Anything past 90 days usually signals that the criteria were never concrete, and the extra weeks add labor cost without adding evidence.

Can you run both motions at the same time?

Yes, and most companies above a certain size do. Assign motion by segment with an explicit threshold — deal size, integration complexity, or industry — and enforce it with a required field in the CRM so the data stays clean enough to compare.

Does a paid pilot fee actually offset the cost?

Partially. A creditable fee reduces cash outlay but functions as a discount on the eventual contract, so net it against first-year revenue rather than treating it as new revenue. A non-creditable fee is genuine offset, but buyers resist it much harder.

What conversion rate makes a pilot worth running?

Divide fully loaded pilot cost by the gross profit on the resulting contract. If the pilot costs $8,000 and the deal yields $20,000 in first-year gross profit, you need a conversion rate above roughly 40% just to break even against that single line, before any other sales and marketing cost.

FAQ

What exactly is a free trial motion?

A free trial motion is a go-to-market approach where prospects get time-limited product access at no cost, typically 7 to 30 days, with minimal human involvement. The true cost includes infrastructure for provisioned accounts, support contacts, acquisition spend, and the large share of trialists who were never going to buy — all of which lands on the small number who convert.

How is a pilot program different from a free trial?

A pilot is a structured, time-boxed evaluation with agreed success criteria, usually a signed agreement, and a named owner on the vendor side running configuration, check-ins, and an exit readout. Pilots carry much higher upfront cost — custom setup, dedicated support, legal review — but generally convert at far higher rates because both parties have already committed real resources.

Which cost categories do teams most often forget?

Three: forgone revenue during the extended evaluation window, the pricing concession pilots tend to invite and that persists into renewal, and the opportunity cost of solutions engineering hours that would otherwise have gone to late-stage deals. Legal and security review time is a close fourth, especially in regulated categories where questionnaires consume days rather than hours.

How do I quantify the risk of non-conversion?

Take your historical trial-to-paid or pilot-to-paid rate and treat the complement as sunk cost distributed across the winners. If 12 of 100 trials convert, all 100 trials' costs are carried by those 12. Add the qualitative cost of team time spent on prospects who never intended to buy, valued at the gross profit that time could have produced elsewhere.

What loaded rate should I use for internal labor?

Base salary divided by roughly 2,000 working hours, then multiplied by 1.25 to 1.4 to cover benefits, payroll tax, and overhead. Using raw salary understates every people-heavy motion. If you cannot get finance to bless a rate, use a consistent placeholder and note it — consistency across motions matters more than precision in the absolute number.

How should I present the comparison to leadership?

One table, one row per segment: cost per converted customer by motion, average evaluation length, conversion rate, and realized price as a percentage of list. Freeze the methodology for at least two quarters so periods are comparable. Debate thresholds, not math — the math should be settled before the meeting starts.

Sources

• https://www.gartner.com/en/sales/topics/b2b-buying-journey — Gartner research on the B2B buying journey and evaluation stages
• https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights — McKinsey insights on B2B sales economics
• https://www.forrester.com/blogs/category/b2b-marketing/ — Forrester analysis on B2B marketing and evaluation motions
• https://www.saas-capital.com/research/ — SaaS Capital benchmarking research on SaaS metrics
• https://openviewpartners.com/product-led-growth/ — OpenView on product-led growth and self-serve trial motions
• https://www.paddle.com/resources — Paddle/ProfitWell resources on SaaS pricing and retention
• https://a16z.com/tag/enterprise/ — Andreessen Horowitz writing on enterprise SaaS go-to-market

Related on PULSE

• [What's the right way to handle a deal where the buyer wants a 6-month free pilot?](/knowledge/q237)
• [How do I structure an enterprise pilot that converts to a paid contract?](/knowledge/q133)
• [What's a good free trial conversion rate — and how do you actually lift it?](/knowledge/q10847)
• [How do you calculate true CAC payback period when you have multi-quarter sales cycles?](/knowledge/q414)
• [How do you calculate true pipeline coverage when usage-based deals have variable ACV?](/knowledge/q10457)