181% of plan in July.

Forest Springs Resort is a 50-key boutique resort in Eureka Springs, Arkansas. It is owned by Jacobo Hernandez, who built YieldWizard. It is the first hotel the Wizard worked in, every night, on real money. This page is what happened, how, and what has not worked yet.

The short version

Results include a full repositioning of the resort. The Wizard did not renovate rooms. It is one part of the story, and we say which part below.

Where it started

The property was a tired roadside inn. Under the prior owner it ran at about 29% occupancy over its June 2023 to June 2024 year.

Jacobo bought it in fall 2025, repositioned it as a boutique resort, and reopened in May 2026.

On day one the operation ran on eight separate logins, with no single view of the money. The resort was nearly invisible on Google.

What kept slipping through

In the first two months, three kinds of money kept leaking. None of them showed up in a normal monthly report.

1. Unpaid folios. Guests checked out with balances and nobody chased them. The first billing scan, run July 3, 2026 against all 430 June checkouts, found $3,000+ in unpaid guest balances on 16 folios, most from guests who had already left.
2. Weekends priced like weekdays. Busy Saturdays sat at ordinary rates while they filled.
3. Channels nobody was grading. One booking channel cancelled far more often than direct bookings, and nobody measured it per channel.

What changed, and when

The method

The Wizard runs the same loop every night. It is the loop on every YieldWizard page, and at Forest Springs it was built one leak at a time.

1. Watch. Read every rate, folio, payout and channel in Cloudbeds and the connected systems, every night.
2. Find. Flag what nobody chased: an unpaid folio, a virtual card nobody charged, commission on a cancelled stay, a Saturday priced too low.
3. Decide. Price each night against the hotel's own pace, comp set and events, not a market average.
4. Act. Write rates and minimum stays to Cloudbeds inside the owner's floor and ceiling, or hand the task to the right person on the team.
5. Verify. Read every change back from Cloudbeds and log the before and after.

Two rules came straight out of Forest Springs and now apply to every hotel:

• A number the Wizard cannot check is "possibly owed," never $0. A report that reads "nothing owed" when it simply could not read a folio hides money.
• Only confirmed balances get chased. A folio where money already went back to the guest, in full or in part, is never chased.

of plan, July revenue

181%

of plan, July revenue

July occupancy on rooms open for sale, from 29% under the prior owner

79%

July occupancy on rooms open for sale, from 29% under the prior owner

of room revenue booked direct, up from 29%

34%

of room revenue booked direct, up from 29%

pages indexed by Google, up from 1

105

pages indexed by Google, up from 1

in unpaid guest balances found on the first billing scan

$3,000+

in unpaid guest balances found on the first billing scan

• 181% of plan, July 2026 revenue. From closed books, checked against Cloudbeds. "Plan" is the resort's 2026 launch ramp.
• 29% to 79% occupancy. 29% is the prior owner's June 2023 to June 2024 year. 79% is July 2026 on the 44 rooms open for sale; on all 50 keys, July was 69.6%.
• 29% to 34% of room revenue booked direct, June to July 2026, and 34% again in August. Every direct booking is a booking with no OTA commission.
• 105 pages indexed by Google, up from 1.
• $3,000+ in unpaid guest balances found on the first billing scan, July 3, 2026, checked against Cloudbeds. We say "found," not "collected": some balances turned out to be OTA ledger timing, and the recovered amount is not verified.

Revenue from July 2026 closed books, checked against Cloudbeds. Occupancy on rooms open for sale; prior owner figure from its June 2023 to June 2024 P&L. Unpaid balance figure from the first billing scan of June 2026 folios, July 3, 2026, checked against Cloudbeds. Results include a full repositioning of the resort.

What didn't work, and what's still hard

We would rather tell you this than have you find it.

• Rate is under target. July grew on volume, not price. More rooms sold, at rates below where we want them.
• Midweek is still soft. Weekends fill. Sunday to Wednesday nights still run far below them.
• A floor can become the price. For a stretch, our weekday minimum rate was set higher than the pricing engine wanted, so every quiet weekday posted the same number to the dollar. It looked like the pricing had stopped working. The cause was our own setting.
• Being found on Google Hotels is slow. In our August 31 review the resort did not appear in the Google Hotels top 20 for "hotels in Eureka Springs" in any week measured. Pages indexed went up fast; ranking for the money search has not caught up.
• Group rooms can fool pricing. Rooms held for an event can look like normal demand and push rates up. We are changing how group holds feed pricing so they never read as walk-in demand.

What it means for another independent hotel

You do not need a 50-key resort or a renovation to use what we learned.

1. Look for money before you look for rate. The first scan at Forest Springs found $3,000+ in a single month of checkouts. Most small hotels have not run that check once. Here is how to run it in Cloudbeds.
2. Grade every night on your own pace. A Saturday filling faster than normal is priced too low, whatever the market average says.
3. Know which midweek nights are the market. If your comp set has empty rooms too, a rate cut just gives money away.
4. Count direct share by revenue. It is the number that tells you what OTAs cost you.
5. Ask every report how it got its number. Half of what our first scans flagged was noise. The owner who checks the method keeps the money.

The Forest Springs playbook only works if it works at a hotel that is not ours. That is what the Wizard is for: the same nightly checks, on your Cloudbeds data.

How each part works: Find the money · Price every night · Owner reports · Get found · For Cloudbeds hotels