# The Echo OS pilot: the first $250 is on us > Approved businesses get $250 in credit toward the work their twin does. Here is what the credit covers, what it does not, and why every application is read by hand. Canonical page: https://myorbit.ai/notes/the-echo-os-pilot --- ## The offer, first I'm running a pilot for Echo OS. The terms, before the reasoning: - **$250 in credit** for approved businesses, toward the work your business twin does - Apply up to **four times, three months apart** — $1,000 across the whole program - Credit **expires 90 days** after it's applied - It pays for **the work, never the subscription** - Every application is **read by a person** — applying is not the same as being approved The full terms live on the [pilot page](/pilot). Everything below is the reasoning behind them — because I think how a company structures a giveaway tells you more about it than the giveaway does. ## Why credit and not a free trial A free trial is a countdown. It measures how long you have, which means the clock is the product's opinion of you: *hurry up and decide.* Two weeks in, you have learned whether you can install something, not whether it can run your business. Credit measures something else — how much work you actually asked for. A business that puts real load on its twin will spend $250 faster than a business that pokes at it. That is the right sorting. I would rather find out what you use it for than how quickly you can be pressured. There is a second reason, less noble and worth admitting: every answer a twin writes, every call it takes, costs real money the moment it happens. Credit is the honest unit because it is the unit the work is actually billed in. A trial pretends the cost is zero until a date, then pretends it is enormous the day after. Credit just tells the truth continuously. ## What the credit covers The credit pays for **the work your twin does** — the answers it writes, the calls it takes, the tasks it runs. When your twin does something, the cost comes out of the credit first, before anything you have paid for. Here is the part most announcements would bury, so it gets its own paragraph: **the credit does not pay for a subscription.** Plans are billed separately, to your card. If you are on a paid plan during the pilot, you are still paying for the plan. The credit is for the work, not the seat. I'm being blunt about that because the alternative is worse. If the line read "credit toward your plan and usage" and you discovered at your first invoice that only half was true, I'd have bought a little enthusiasm today and spent all my credibility to get it. That is a bad trade at any price, and a catastrophic one for a company asking businesses to hand it their operations. ## What expires, and when Credit expires **90 days** after it is applied. Whatever is unused on day 91 is written off — not refunded, not rolled over. I could have quietly omitted that. Expiry is the clause every promotional page is tempted to leave for the fine print, because it is the one that sounds least generous. But an expiry you discover later is not a term, it is an ambush. Ninety days is also the honest shape of the thing: this is a pilot, and a pilot with no clock is just an indefinite discount pretending to be an experiment. The three-month gap between applications is deliberately the same rhythm. Use a pool, see what happened, come back and tell me. ## Why a person reads every application There is no promo code here. No coupon field, no voucher, nothing to paste at checkout. You email, a person reads it, and if it is approved the credit is placed on your account by hand. That is slower than a code, and I chose it anyway. A code is a broadcast — it goes to whoever forwards it fastest, which is rarely whoever needs it most. Reading applications means learning what businesses actually want to put on a twin, which at this stage is worth more than the efficiency being given up. It also means being able to say no, kindly and specifically, instead of letting a form say nothing at all. So: **applying is not the same as being approved.** You get a reply either way. ## What I have not decided yet There is no published list of qualifying criteria, because there isn't an honest one yet. I could invent a checklist that sounds rigorous — years in business, headcount, a category list — and it would be theatre, reverse-engineered to look official. What is true is that eligibility is confirmed when your application is read. Tell me what your business does and what you want the twin to run, and you'll hear where you stand. When there are criteria worth publishing, they'll be published on the [pilot page](/pilot) rather than left to be guessed at. ## The bet Echo OS is an [operating system for a business](/echo-os) — a twin that talks, remembers, and acts, with decisions coming to you and work going to agents. That is a large claim, and the correct response to a large claim is *show me*. The pilot is me paying for the first stretch of showing you. Not a countdown, not a locked demo, not a sales call with a slide about ROI. Two hundred and fifty dollars of your twin doing real work, on us, and then an honest conversation about whether it earned its place. If that sounds like the right trade, [the pilot page](/pilot) has the terms and the way in.