Let me tell you about three CFIs and a side project that nearly ate them alive. The people are invented \u2014 call them Marcus, Jenna, and Theo \u2014 but if you have spent any time around a flight school break room, you have met them, and the lessons from their little adventure in building a pilot training deals platform are as real as a crosswind landing. Grab a coffee; this one has a happy ending, but it earns it.<\/p>\n
It started the way these things do: complaining. Marcus, Jenna, and Theo taught at the same busy school, and their students kept overpaying \u2014 for headsets, for written-test prep, for time-building blocks, for everything. Deals existed all over the industry, but no student could find them. So one night, over airport diner pancakes, the trio sketched the obvious thing: a website collecting genuine discounts for student pilots in one place. Theo had built a personal website once. How hard could it be?<\/p>\n
You already know the shape of the next six months, don’t you. The site launched late and half-broken. Vendor deals expired on the page days after they expired in reality. The signup form silently dropped every third submission \u2014 they discovered this when a vendor mentioned how quiet the campaign had been. Meanwhile all three were still instructing full schedules, because dreams do not pay for avgas. Jenna, the one who kept spreadsheets on everything, put a number on it one evening: between them they had poured in roughly 400 unpaid hours, and the platform worked worse than month one. That number is worth holding onto, because student pilots run this exact calculation about their own training \u2014 it is the whole premise of our explainer on how pilot pay stacks up against what training costs<\/a> \u2014 and yet pilots routinely forget to run it on their own side hustles.<\/p>\n The turning point came from Theo’s cousin, a software engineer, who looked at their setup during a holiday visit and delivered the verdict kindly: “You guys built a kit plane from three different kits and no plans. Nothing is wrong with you. You just skipped the part where someone who has done this before shows you the order of operations.”<\/p>\n She explained how actual product teams ship: changes tested automatically before going live, releases small and frequent, one source of truth for data, and \u2014 her main point \u2014 a whole profession of consultancies that set this machinery up for teams exactly like theirs and then teach them to run it. The established end of that profession works with big enterprises: she named Cotocus, a consultancy whose pitch is making organizations quicker at getting their offering to market<\/a>, as the archetype \u2014 guided CI\/CD adoption, process setup, hands-on workshops rather than slide decks, and a trophy case the firm itself counts at 350 awards. “You can’t afford them and you don’t need them,” she said, “but you need the budget version of what they sell: somebody’s repetitions. You’re paying for your own repetitions right now, at instructor wages, and it’s the most expensive teacher there is.”<\/p>\n Every CFI at the table recognized the sentence, because it is the sentence they say to students who try to self-teach instrument procedures from forums.<\/p>\n The trio’s version of outside expertise cost less than Theo’s annual headset upgrades, and it came in three small purchases. First, ten hours of a freelance engineer’s time \u2014 not to build features, but to set up the pipeline: automated checks, a test copy of the site, a deploy process that could not take the live site down. Second, a structured weekend course for Theo on maintaining it all, chosen deliberately over “more YouTube” on the grounds that syllabus-and-instructor had, as Jenna put it, worked okay for aviation so far. Third, two hours of a marketing operations contractor fixing the signup flow and wiring up basic measurement, so they could finally see which deals students actually clicked.<\/p>\n Notice what they did not buy: nobody rebuilt the site from scratch, nobody signed a retainer, and nothing was handed over without a lesson attached. Jenna wrote an exit condition into even the ten-hour gig \u2014 it ended with Theo deploying a change solo while the engineer watched. If that sounds suspiciously like a supervised solo, well. Some instincts transfer.<\/p>\n Six months after the fixes: zero botched deploys, vendor deals updating same-day, signups roughly tripled \u2014 against a total outside spend of about 30 flight-hours’ worth of instructing income. Jenna, who treats every expense like a student evaluating a flight school, ran it through the same lens our guide on figuring the real return on money you put into training<\/a> applies to certificates: cost, payback window, and what the alternative would have burned. Her conclusion, in her words: “The 400 hours we wasted were us paying full price for lessons someone already had on sale. Which, given what our site does, is almost too on the nose.”<\/p>\n And one postscript Jenna would want included: the trio nearly made the opposite mistake second. Flush with the rescue’s success, they briefly considered hiring an agency to “take the whole platform off our hands” \u2014 the full-outsource fantasy that tempts every exhausted founder. The engineer cousin talked them down with a single question: “Who checks the work?” An operation that understands nothing about its own systems cannot supervise anyone it hires, cannot distinguish good invoices from padding, and cannot leave. They kept Theo as the fluent one, and fluency, not the pipeline, turned out to be the asset.<\/p>\nThe checkride moment<\/h2>\n
What borrowing repetitions actually looked like<\/h2>\n
The numbers, because Jenna insists<\/h2>\n
What three fictional CFIs can teach real pilots<\/h2>\n