Our entire marketing plan is one email

August 2, 2026 · The Retreat Register

The Retreat Register’s launch list works one way. You give an email address, and you get one email, sent when the register opens. There is no sequence behind it, no second message, and your address is never put on a shared line with anyone else’s.

That is an easy promise to make and a hard one to be worth. This page is the argument for why it is worth it, and the argument is made entirely out of things you can go and check.

The work is published instead of announced

Most directories run a mailing list because the mailing list is where the story lives. The product is quiet, so the emails have to be loud.

This register does the opposite. It keeps a build log, which is a dated list of what was done to the file on each day, in public, without being asked. It is not a newsletter and there is nothing in it about how the team is feeling. It is a record.

The log starts on July 7, 2026, the register’s first day, with the rule that still runs it: a venue is either good enough to be in the register, or it is not in it.¹ From there it runs day by day. If you want to know what happened to the file on a Tuesday in the middle of July, the answer is on that page with a date next to it.

What a record includes that a newsletter does not

A newsletter reports arrivals. A record has to report subtractions too, and the subtractions are the reason this page exists.

July 9, 2026. Somebody read sample rows by hand and found that messy public data had pushed a luxury fashion retailer and an online dictionary into the catalog as venues.¹ Both were cut. Every cached verdict was thrown out so that every candidate could be judged again under sharper rules, against a raw catalog standing near 1,850 candidates.

July 10, 2026. Ninety-five convention and expo facilities were moved off the public list onto a review shelf.¹ Real businesses, wrong category. Faith-based retreat centers were reviewed one at a time against the same bar as everyone else on the same day: seven promoted, two declined.

July 11, 2026. A person read every published row, every name and every website, and cut 274 entries that no automated filter had caught.¹ They were budget hotel chains and dozens of template-name fakes pointing at unrelated websites.

July 12, 2026. The register put 2,811 of its own past rejections back on trial against a rebuilt judge. It restored 606 of them, upheld 2,193 rejections, and set 12 aside as obvious junk.² That is an overturn rate of 21.6%, published by the party it makes look worst.

July 18, 2026. The register held 2,281 curated venues on July 17. The next day it held 2,125.³ That is 156 venues gone in a day, and the build-log entry for July 18 records the quality pass that removed them. Two separate pages say the file got smaller.

None of that is flattering and all of it is dated. A marketing plan built on announcements would have skipped the July 18 entry entirely, because nobody announces a bad week.

The number that argues on its own

Here is the finding that removes the need for a persuasive email.

Between July 12 and July 18, 2026 the register published its own size and its own phone-number share on seven separate dates, one page per date. Over that window the file grew from 1,374 venues to 2,281 and was then cut back to 2,125. The share of venues carrying a direct phone number across those seven readings was 90.44%, 90.59%, 90.59%, 90.62%, 90.66%, 90.73% and 90.76%.⁴

The whole spread is four tenths of a point. The file changed size by two thirds and the ratio would not move.

A number that holds still while the sample nearly doubles is describing the industry rather than the sample. It says that roughly nine in ten American retreat properties are reachable by direct phone, and that the one in ten that are not is a stable feature of the trade rather than a gap that better searching would close. That is a useful thing for a planner to know before the first afternoon of phone calls, and it did not need a subject line.

The same seven readings say something less comfortable about price. On July 12, 133 of 1,374 venues published a starting price anywhere on their own website, which is 9.7%. On July 17, with 2,281 venues on file, 188 did, which is 8.2%.⁵ The bigger the sample, the smaller the share of venues willing to publish a number. Pricing in this industry is quote-only by default, and the register measured how deeply that is true rather than complaining about it.

The map, and the state nobody would have guessed

The register published one full state table, dated July 12, 2026, and it has not been replaced since.⁷ On that date the twenty deepest states held 715 of the register’s 1,374 venues, which is 52.0% of everything. The remaining 659 venues were spread across the other thirty states and the District of Columbia, an average of twenty-one each.

Half the file in twenty states is worth sitting with if you are choosing a destination. It means that a planner searching one famous market is looking at a crowded slice, and a planner willing to look anywhere else is competing with far fewer people for the same weekend.

The register’s response to its own thin half was a rule dated July 11, 2026: every state was given a minimum coverage target sized to its market, and the worst-covered states were searched first rather than last.¹ What that produced is the best single argument on this page. Louisiana does not appear anywhere in the July 12 table, so it held twenty-one venues or fewer that day. Five days later it held 92.³

An email campaign would have called that a milestone. The log just records it and moves on.

One fix that an announcement would never mention

On July 12, 2026 the register changed how it treats hotel groups that run many properties from a single website. Until then those groups were tracked as one entry, so every property but one was invisible. From that date each property is tracked separately, and the log says exactly what the flaw had been costing: it had been hiding whole resort markets, including most of Hawaii’s.¹

That is a directory publishing the sentence “we were missing most of a state, and here is why.” It is the kind of admission a launch email is designed to keep out of the reader’s hands, and it is on a public page with a date on it.

What is not known, said plainly

The last exact total the register has published is 2,152 curated venues, on July 19, 2026.⁶ Pages after July 21 carry no figures at all, only the dataset note at the foot of each one. So there is no newer count to quote, and this page will not imply one.

There is also no published count of how many listings carry an owner’s own photographs rather than a stand-in image, because that has never been measured. The photograph rule is stated on the methodology page as a policy: a listing never shows a different property, and until an owner claims a listing and uploads photographs of their own, the venue is matched with a close stand-in image.⁶ A policy is not a proportion, and it is not written here as one.

Publishing the shape of your own gaps is the part of this that no competitor can borrow, because borrowing it would mean having a record to be honest about.

What you can use today without giving anyone an address

The markets page carries the coverage table as of July 12, 2026, and it is the only full state table the register has published. The methodology page carries the four questions every property is judged against. The build log carries the dated history above. None of them ask for anything.

If you are about to call a venue, the four things that actually determine your final number are date flexibility, the buyout threshold, food-and-beverage minimums, and service fees.⁵ Ask those four up front. That is not the nightly rate, and the nightly rate is not the number you will end up paying.

Updated August 10, 2026

The build log now holds forty dated entries, running from July 7 to August 9, 2026. Two days in that stretch carry no entry at all, July 23 and July 25, and both gaps are visible on the page because nothing was backfilled to cover them.¹

Three of those entries record quality passes that removed venues, dated July 18, July 22 and July 24, 2026.¹ Cuts are not a one-off event in this register’s history. They recur, which is why the sentence on other pages reads that venue cuts happen in ongoing quality passes.

The launch list has not been used once in that time. The count remains at one email, sent when the register opens.

Have you read: "Quote only" is a feature, not a failure

Based on The Retreat Register internal venue dataset as of August 2, 2026. How the register is built: the methodology.