The monitoring programs have both a ratio and a minimum event count, and the count is what saves you. Visa's excessive threshold moved down to 1.5% for most regions in April, but it only bites once you're above a monthly minimum in the thousands of combined fraud and dispute events. At three events a month you are nowhere near that floor - your ratio looks scary and is statistically meaningless at this volume.
What will hurt you first is your own processor's risk team, which has its own thresholds and doesn't publish them.
Release it quietly now, days before the announcement. Let it be live and unmentioned so you can install from the store yourself, check the listing renders properly, and confirm purchases work against production. Then do the launch as a marketing event on your date. A launch is an announcement, not a deploy.
Just went through this and the fix that worked for me was embarrassingly small. I removed the link, removed the logo from my signature, removed the tracking, and made the email look exactly like an email I would send a colleague. Four lines, plain text, one question mark. Replies came back immediately.
The other thing I changed: the second email in the sequence is the one with the calendar link now, and only if they replied. Nobody has complained about a link in a conversation they started.
I moved everything to the first meal about two years ago, dropped the morning slot entirely, and the only thing I noticed was that I stopped forgetting doses. That is not nothing: adherence is the effect size that actually matters for anything slow acting.
Before touching the flow, confirm the mail is actually arriving. Auth stacks have surprisingly small default sending limits that nobody reads about until this exact moment - the built-in mailer on some hosted auth services is a couple of messages an hour and will only send to addresses on your own team, and even after you plug in your own SMTP there is typically a default cap in the tens of messages per hour that you have to raise manually in the settings. If you ever had a day with a spike, mail was silently dropped.
Go look at your sending provider's accepted, delivered, bounced and complained counts for those 340 addresses. If delivered is well under 340, this is a plumbing problem and no amount of button redesign fixes it.
Going to disagree with the framing rather than the numbers. If money were the objective you would be doing overtime, not sewing, and per hour it is a terrible trade. What you are actually buying is a wardrobe with no compromises in it, and that is not available at any price on the high street for a lot of people. I stopped doing the maths when I noticed I wear my handmade trousers four times as often as anything I bought.
Squeezing the air out is the right instinct but a soft bottle with no air will still weep at the threads because the cabin drops to roughly 2400m of pressure and everything inside wants out. Two things fixed it for me: switch to bottles with a proper flip cap and a silicone gasket rather than a disc cap, and stop filling past about 80%. The leaks I get now are basically zero, and when there is one it's because I cross-threaded the cap in a hurry.
Did my first one in March and the one pager is genuinely all I use now, so here is what is in mine in case it saves you an evening.
Who runs it and how many people have production access. Where data lives, which country, which provider. What is encrypted and where. Backup frequency and the last time I tested a restore, with the date. Subprocessors listed in a table with what each one receives. Retention and what happens on cancellation. How to report a security problem and how fast I promise to reply.
It is a page and a half. Two prospects have now accepted it in place of most of their spreadsheet, and one asked for exactly three follow up questions.
Two things that get around the limit without any argument at the gate. Sunscreen sticks are treated as solids in most places, so a couple of those for faces and ears do not touch your liquids allowance at all. And if you are checking any bag at all, even a small one on the way home, full sized bottles are fine in the hold.
Mild disagreement with everyone recommending the workshop. If your budget is one thing, I would spend it on a made to measure block or a pattern drafted to your measurements instead, and then use that block to check every commercial pattern before you cut. It solves the same problem permanently rather than one garment at a time, and around here it costs about the same as the six week course.
Cheap experiment before your ten weekends start: go find where their customers complain in public and read six months of it. Support forums, app store reviews, the replies under their launch posts, their Discord if they have one. Sort mentally into "wish it did X" and "it broke and nobody helped me". Two hours of that will tell you more than any framework.
Second thing, and this is the one people skip: try to buy from them badly. Ask a pre-sales question by email as a small customer and see how long it takes to get an answer. If a $19 product takes six days to reply to a buying question, you have found out how much they care about the low end.
Polyester dries faster, sure, but 'a bit' undersells it. One humid day in a poly tee and you can't wear it again without washing, which puts you right back at the same laundry problem with worse company on the bus.
Wringing is your problem. Twisting merino leaves water trapped in the yarn and stretches the shoulders at the same time. Lay the shirt flat on a dry bath towel, roll the towel up like a swiss roll with the shirt inside, then stand on it with your full weight and walk down the roll heel to toe for about 30 seconds. The towel pulls out far more water than your hands can, and the shirt comes out damp rather than wet. After that it needs airflow, not heat.
What actually solved this for us was clothing. A long sleeved swim top and a wide hat means the sunscreen job shrinks to face, hands, feet and the back of the neck, which is thirty seconds of lotion and no fight. Two summers in and I would not go back, and it costs less than the sprays did over a season.
Also worth cleaning the coil before you do anything else. Grit and old detergent residue in a coil makes the slider ride badly. Old toothbrush, warm water, dry it fully, then a rub of candle wax along the teeth.
The honest constraint here isn't volume, it's the warm layer. A synthetic or down midlayer that packs to the size of a grapefruit fits either setup. A fleece does not, and that alone can push a 20L over. Buy the packable insulation first and then the bag question gets easier.
Pack in flat layers rather than by category. Clothes go in as a single compressed slab against the back panel, shoes sole to sole in the bottom corner, and anything lumpy gets buried in the middle so the outer walls stay flat. A pack full of round-ish stuffed cubes will always dome out even at the right volume.