pilot (proof of concept)
Also called: pilot project, proof of concept, paid pilot
A small, time-limited trial of the product with one team or site, with success criteria agreed in advance and a decision date. A pilot without criteria and a date tends to drift on unpaid and prove nothing.
A pilot (or proof of concept) is a small, time-limited trial of a product with one team or site before a full purchase. Its purpose is to let the buyer decide with evidence rather than promises: the buying side can see whether people use the product and whether it moves the measure they care about, at a fraction of the full commitment.
A pilot decides something only if it is designed to. It needs one real group that looks like the future rollout (not hand-picked enthusiasts), a fixed length, success criteria agreed before the start, a named person who decides, a decision date, and a clear exit if the answer is no. Without criteria and a date, a pilot tends to drift on, unpaid, while everyone waits for someone else to call it.
Whether to charge is a real choice. A free pilot costs the seller its delivery costs and gives the buyer nothing at stake, so nobody watches the results closely. A paid pilot, often credited to the contract if the buyer signs, keeps both sides attentive and filters out buyers who were never going to commit. Criteria should be the buyer's own targets, written down, not an industry average nobody can source.
Example
Klasna for Teams designs a pilot for Veloria: one dispatch team of 12 people on its normal shifts, 2 lessons a week for 4 weeks, from 11 January to 5 February.
- Lessons: 12 × 2 × 4 = 96.
- Paid at the usual price: 96 × 800 = 76,800 UAH, credited to the contract if Veloria signs.
- Free, it would cost Klasna 96 × (430 tutor + 31 video and AI) = 96 × 461 = 44,256 UAH, with nothing at stake for Veloria.
Success criteria agreed before the start: (1) at least 3 in 4 scheduled lessons attended, i.e. at least 72 of 96; (2) progress on the same short level check in week one and week four; (3) the finance director decides on 10 February, and on a no Veloria owes nothing more. The rollout behind it is 300 lessons a month for a year, 2,880,000 UAH, so the pilot puts 76,800 / 2,880,000 = 2.67% of the annual value at stake to test the rest.
Common mistakes
- No criteria and no date. "Let's see how people feel" produces a trial that drifts on and proves nothing.
- Hand-picked participants. Enthusiasts flatter the numbers; use a group that looks like the full rollout.
- Free by default. With nothing at stake the buyer stops watching; a paid pilot credited to the contract keeps both sides attentive.
- Borrowed benchmarks. Agree the bar with the buyer and write it down instead of quoting an average you can't source.
- No exit. Say in advance what happens on a no; a clear way out is what lets the budget owner say yes to the trial.
Learn it in the course
- Reducing the buyer's risk · Proof each person on the buying side can use, and a pilot that ends in a clear yes or no.
- Case: map Veloria's buying committee · Prepare the first meeting with a logistics prospect: who is in the decision, what each person fears, which proof and pilot to bring, and what to ask first.
- Content a committee uses: one asset per role and stage · Why ten blog posts a month rarely move a company deal, which four assets each committee member actually uses, when to give them, and what never to write.
- The demo and a mutual plan to close · Show each person on the committee only what solves their problem, then agree a shared plan with steps, owners and dates on both sides, from the demo to the first lesson.
- A quote built from the buyer's economics · Start a proposal from what changes for the buyer, price a program instead of a lesson, offer a few clear options and terms, and fit it on one page.