See your product the way a directory will
Submitting to a directory hands over a URL and nothing else. What appears in the listing is whatever a scraper can pull off that page in one request, without running any JavaScript. If a tag is missing or an image won’t load, the listing is built from whatever is left — and nobody tells you.
This does that fetch and renders the card from what came back.
The three that break most often
A relative og:image. /og.png resolves fine in your browser, because your browser knows what page it is on. A scraper reading the tag out of context often does not, and shows nothing. The fix is an absolute URL.
An image behind a login. If the file sits somewhere that needs a cookie, it works for you every time you check and never for anyone else.
A page that needs JavaScript. The scraper gets an empty shell, and your listing gets whatever text it can scrape from it — often a nav link. The crawler check measures the same gap for AI assistants, which have it worse.
What this can and can’t tell you
Every directory renders its cards a little differently, so treat the preview as the shape rather than a pixel-exact copy. What is exact is the input: these are the tags and the image bytes a scraper receives from your page, read the same way twice.
Once the card looks right, list it on LetsLaunch — or work through the directories we checked, where each entry says whether its outbound link is really dofollow.