When to send it
Send this to people who cancelled two to six months ago. Sooner than that, the reasons they left still hold and the email feels pushy. Much later, many have forgotten why they subscribed at all. Paste the link into an email from your usual email tool, sent to former subscribers who agreed to hear from you, and add ?name= so the greeting uses their first name. If you want to match answers back to your list, add a customer reference to the link as ?ref= too.
Make the subject line about the question, not the discount. Something like "Can we ask why you left?" is honest, and the coupon at the end still comes as a pleasant surprise.
Why these questions
The first question asks why they stopped. The options differ from a cancellation survey's because time has passed: "my needs changed" and "the boxes got repetitive" often only become clear in hindsight. Anyone who mentions delivery or service problems is asked what went wrong, since a failure like that is worth hearing about in detail even months later.
Everyone is then asked what they've been using instead. That answer tells you who you're really competing with, whether a rival box, the supermarket, or nothing at all.
The checkbox question asks what would make them consider subscribing again: a lower price or smaller plan, more say over what's in the box, less frequent deliveries, better delivery, or new products. More than one answer is allowed, because a return usually depends on a couple of things changing.
The final question asks how likely they are to restart in the next few months. Those who say "not likely" go to a separate end screen that thanks them without pressing further.
The coupon
30% off the first month back is generous on purpose. A former subscriber already knows your product, and a meaningful offer is often what tips someone who was half-tempted anyway. Use unique codes, since a discount this size spreads fast on deal sites. Both end screens show the code, so people who doubt they'll return can still change their minds.
What to do with the answers
Set the reasons against what you've changed since those subscribers left. If many quit over repetitive boxes and you've since widened the range, say so in your next email to them. Those who answered "very likely" deserve a personal note if you can match them through the reference. For people who have only just cancelled, the subscription cancellation reasons survey catches the reason at the time; for one-off buyers who stopped ordering, use lapsed customer win-back.