Operators routinely pour money into Google Ads, Meta Ads, programmatic placements, retargeting, and influencer drops, looking for the channel that finally produces direct bookings. Each campaign launches with hope, runs for six weeks, generates clicks but few conversions, and ends with a familiar conclusion: this channel does not work for STR. The conclusion is wrong. The channel works. The channel is doing exactly what a channel is built to do, which is deliver attention to your site at a known price. What the channel cannot do is convert the attention. Conversion is a property of the operation behind the click, not the channel that delivered it. When a campaign produces clicks and no conversions, the failure is downstream. The site is slow. The booking flow has friction. The price story is inconsistent. The follow-up sequence is missing. The retargeting pixel is misconfigured. The CTAs lead to a generic homepage instead of a property-specific landing page. The funnel does not exist as a funnel. It exists as a series of disconnected pages that collectively fail to move the visitor toward a booking. This is the operator's paradox: more advertising spend exposes the broken backend to more people. The clicks get more expensive because the operator increases bids trying to force conversions. The conversion rate stays flat because the conversion problem was never advertising. The cost per acquired guest climbs until the operator pulls the budget and concludes that the channel failed. The honest sequence is the reverse. Fix the backend first. Then run ads against a backend that can convert. The backend has six components that have to work together. The site has to be fast enough to compete with platform load times on mobile. The booking flow has to commit to one price story from search card to checkout receipt. The calendar has to reflect real-time availability with no manual sync. The payment relationship has to use a real processor with visible trust signals. The follow-up sequence has to respond to inquiries within minutes, automatically. The retargeting infrastructure has to track the visitor across sessions and re-engage with the specific property they viewed. When all six work, an ad campaign delivers a guest to a system that can move them to a booking. The same dollar of ad spend produces a different result, not because the channel changed, but because what sits behind the channel changed. Operators who skip this sequence learn the lesson the expensive way. They burn a quarter of ad budget on a backend that cannot convert. They blame the channel. They try the next channel. They burn another quarter. By the end of the year they have a confident opinion about which platforms do not work, and they have not addressed the actual problem. The operators who skip the expensive lesson run a diagnostic first. They measure their site speed honestly. They walk through their own booking flow as a guest and notice the friction. They check their inquiry response time and find it is hours, not minutes. They acknowledge that the backend is broken. Then they fix it. Then they advertise. The order matters because ads cost money the moment they run, and the conversion math is unforgiving. A campaign with a low conversion rate produces a high cost per booking that no campaign optimization can fix. The same campaign with a healthy conversion rate produces a sustainable cost per booking on the first day. The difference is not the campaign. The difference is the conversion rate, which is a property of the backend. The Scorecard at /scorecard surfaces the backend leaks that destroy advertising economics. Website, payment, follow-up, pricing, and calendar all show up as scored zones. Operators who run the Scorecard before launching a campaign know which leak to close first. Operators who skip it find out later, by checking their cost per booking and being unable to explain it. You can advertise into a broken system. The market does not stop you. The market just refuses to convert.