We spend more time on pricing than we do on any other single piece of property management. Owners sometimes find that surprising — they think the work is cleaning logistics, or guest communication, or maintenance coordination. Those things matter, but they’re tablestakes. The thing that actually decides whether a property nets $30,000 a year or $42,000 a year, with the same furniture, the same cleaning crew, and the same level of guest care, is pricing.
This post is twelve months of what we’ve learned running pricing across about 90 short-term rental properties in Dallas-Fort Worth and Houston from June 2025 through May 2026. Some of it confirmed what we already thought. Some of it surprised us. A couple of things we’ve stopped doing entirely because the data wasn’t there.
If you self-manage your properties and you set rates by intuition, by what the unit two doors down is charging, or by whatever your dynamic pricing tool defaulted to when you signed up, the math here might change how you think about it.
The baseline: what pricing actually does
Before getting into the experiments, here’s the most important sentence in this post: pricing is the largest single variable you control in short-term rental returns. Bigger than cleaning quality, bigger than photo quality, bigger than amenities, bigger than which platform you list on. Two identical properties operated by two identical owners can produce 15–25% different annual revenue based on pricing strategy alone.
That number isn’t a guess. We’ve seen it directly in our portfolio. We’ve onboarded properties where the previous host was charging $145/night when the right number was $185. We’ve onboarded properties where the previous host was charging $220/night when the right number was $190. The difference in annualized revenue between “right number” and “off by 15%” is usually larger than the entire management fee an owner pays.
This is why we obsess about pricing. Everything else either supports good pricing or rides on top of it.
The infrastructure: what we use and why
For owners curious about tooling, here’s our stack:
- PriceLabs is our base dynamic pricing engine. We’ve used Wheelhouse and Beyond historically; PriceLabs has, for our portfolio composition, been the most accurate at predicting demand inflection points. The difference between the three tools is closer than the marketing makes it look, but we found PriceLabs’s neighborhood-level demand modeling slightly better for the mid-density Dallas neighborhoods where most of our portfolio lives.
- Manual overrides on about 30% of our nights. The dynamic engine is good at the broad pattern but bad at the long-tail edges. We override around major event windows, around weather-related demand spikes, around obvious shoulder-season transitions, and around any night where the base-rate prediction looks visibly off compared to the actual booking pace.
- Internal weekly review every Monday, on every property. About 20 minutes per property of human attention per week. This is the unglamorous work that the spreadsheet-only operators skip and that we think drives most of our outperformance.
Total cost per property: about $35/month in software plus the labor. For a single-property owner doing this themselves, the software alone is workable; the weekly review is hard to sustain across more than 2–3 properties without a system.
What we changed in the last 12 months — and what the numbers showed
We treat pricing as a permanent set of small experiments. Here’s what we ran in the last 12 months and what the data said.
Experiment 1: Aggressive weeknight base-rate increases
Hypothesis: Most Dallas hosts under-price weeknights relative to weekend nights because weekend pricing is loud and visible while weeknight pricing is sleepy. We tested raising base weeknight rates 5%, 7%, and 9% across three different subsets of the portfolio.
Result: The 5% bucket showed conversion drop of 1.4%, ADR lift of 5%, and net RevPAN lift of 3.5%. The 7% bucket showed conversion drop of 3.1%, ADR lift of 7%, net RevPAN lift of 3.8%. The 9% bucket showed conversion drop of 6.2%, ADR lift of 9%, but net RevPAN lift only 2.7%.
The sweet spot was a 6–7% weeknight increase. Going higher hit conversion harder than the rate lift compensated.
Conclusion: We standardized on 6% as the default weeknight rate increase for new portfolio onboards. This single change, applied across the portfolio in late 2025, contributed about 1.5 percentage points to our overall RevPAN outperformance for 2026.
Experiment 2: Length-of-stay discount restructuring
Most dynamic pricing tools come with default length-of-stay discounts: typically a small discount for 3+ night stays and a larger one for 7+ night stays. The defaults are reasonable but not optimized for any specific market.
We tested four different length-of-stay configurations across the portfolio:
- Default (8% off for 3+ nights, 18% off for 7+): baseline
- No discount: removed length-of-stay discounts entirely
- Tight (5% off for 3+, 12% off for 7+): smaller discounts than default
- Wide (10% off for 4+, 25% off for 14+): different breakpoints, more aggressive at the long end
The no-discount configuration tanked. We lost about 12% on bookings of 3–6 nights because guests had real alternatives where the per-night-effective rate was lower. RevPAN dropped 4.1% on properties where we tested this.
The tight configuration was about flat with baseline.
The wide configuration was the winner. By shifting the discount breakpoint from 7 nights to 4 nights, we captured more of the “long weekend / Tuesday-through-Friday” stay pattern that has become more common in Dallas as remote workers extend trips. RevPAN lifted 2.3% on the test subset.
Conclusion: We rolled out the wide-discount configuration across the portfolio in Q1 2026.
Experiment 3: Same-day and short-notice pricing
Most pricing tools drop the rate aggressively on rooms still available within a few days. The assumption is that empty nights are worse than discount nights. This is usually right but not always.
We tested three short-notice configurations:
- Aggressive same-day discounting (default): 25% off for nights with <2 days lead
- Moderate (-15%): smaller short-notice discount
- Floor-only: no automatic discount, just a hard minimum-price floor
The aggressive default left meaningful money on the table. In our markets, last-minute bookings have less price elasticity than the standard model assumes. Many last-minute bookers are business travelers, displaced travelers (canceled flights, family emergencies), or event-driven bookers — all of whom are less price-sensitive than the “vacation deal hunter” archetype the discount is built for.
The moderate configuration showed flat RevPAN with slightly fewer empty nights. The floor-only configuration showed the highest RevPAN of the three, with a small uptick in empty nights.
Conclusion: We use the floor-only configuration with a minimum price set at 75–80% of the property’s average ADR. The slight increase in empty nights is more than offset by the revenue captured on the nights that do book.
Experiment 4: Event-driven premium pricing
Dallas has a stacked event calendar — major sporting events, concerts, conventions, festivals. PriceLabs and similar tools pick up some of these, but not all, and not always at the right magnitude.
We hand-built a 12-month event calendar of the 30 largest demand-driving events in the DFW metro and manually overrode pricing on those nights. The premium we applied varied from 15% (for second-tier conventions) to 70% (for the largest single events).
Net effect on portfolio RevPAN: roughly 2 percentage points of lift, just from manual event-driven premiums.
Conclusion: This is the highest-ROI manual work in our pricing process. The dynamic engine misses events it doesn’t have data on, and the magnitudes it does pick up are usually conservative.
Experiment 5: Cleaning fee restructuring
This one came up in our post about the Airbnb 15.5% fee earlier this month — short version, we tested cutting cleaning fees by 30% across a subset and found it didn’t move the needle enough to justify the per-booking margin hit. We rolled it back.
What we stopped doing
Two things we tried that we’ve now stopped:
Stopped: Trying to time the orderbook on weekends specifically
We spent a quarter trying to dynamically raise weekend prices in the final 2–3 days before the weekend if booking pace was strong. The math seemed right — late-stage demand is less price-sensitive — but the execution was messier than we expected. Sometimes we’d raise prices and then have a cancellation, and end up with a high-price empty night. The net RevPAN effect was approximately zero with extra operational complexity. We rolled it back.
Stopped: Property-specific manual ADR floor calibration weekly
We used to recalibrate each property’s minimum price floor every week based on the previous week’s booking velocity. It turned out to be too tight a feedback loop — we’d react to noise. We now recalibrate floors once a month, and aggregate-level (across submarket groups), not property-by-property. Total time saved per week: about 6 hours. RevPAN effect: imperceptible.
What we’d tell a single-property owner doing this themselves
If you have one property and you’re managing pricing yourself, here’s the simplest version of what we’ve learned that will get you 80% of the benefit with 10% of the time investment:
- Use a dynamic pricing tool, but don’t trust it blindly. PriceLabs, Wheelhouse, and Beyond all work. The tool gets you a reasonable baseline. Your job is the override.
- Raise weeknight base rates 6% relative to whatever the tool suggests. Hold for four weeks. If conversion drops less than 4%, hold the new rate.
- Build a manual event calendar. Find the 10 largest demand-driving events in your market in the next 12 months. Override prices on those nights with a premium of 20–40%.
- Set a hard floor at 75% of your average ADR. Don’t let the dynamic tool drop prices below it. Yes, you’ll get a few extra empty nights. The net RevPAN is better.
- Review your pricing once a week for 20 minutes. Set a recurring calendar event. Look at the next 30 nights. Override anything that visibly looks wrong.
That’s it. Five steps. If you do those five things consistently for six months, you will outperform a host who just installed PriceLabs and trusted it by 5–8% in net revenue.
What we’d tell an owner debating whether to outsource this
The single-property math is workable if you have the time and discipline. The five-step process above takes about an hour a week including the setup. Over a year, that’s 50 hours of work to capture maybe $3,000–$4,000 of additional revenue on a typical Dallas property — call it $70 an hour of effective return.
That’s a fine return if you enjoy the work, are organized about it, and you have time. Most owners we talk to overestimate their willingness to do this consistently. The first month is easy. Month four, when the calendar event hasn’t fired and the dashboard hasn’t been opened in three weeks, is when the system breaks down. Pricing’s value comes from consistency, not from heroic effort in week one.
Outsourcing only makes sense if the manager actually does the work. There are a lot of property managers in the market who collect a management fee and run pricing on autopilot — they have PriceLabs installed, they never look at it, they pocket their percentage. That’s worse than doing it yourself. Before signing with any manager, ask them: how many manual pricing overrides did you make on your portfolio last week?
If they can’t give you a number, they’re not doing pricing work.
How we built our own process
For owners curious about how we structure pricing operations internally:
- Monday weekly review of every property in the portfolio. We look at occupancy curve, ADR trend, conversion rate, and any anomalies. 20 minutes per property.
- Event override review the first Monday of every month. We update the rolling 90-day event calendar and apply premiums.
- Quarterly tooling review where we look at whether PriceLabs is still the right primary tool, whether our manual override frequency is too high or too low, and whether the rules need adjustment.
- Owner reporting monthly, showing what we changed, why, and what the effect was.
This process didn’t exist three years ago. We built it because we’d grown to a portfolio size where ad hoc pricing wasn’t working anymore, and we’d seen owners get burned by managers who never opened the dashboard. The single biggest investment we’ve made in our operation over the past three years has been in pricing rigor.
A closing observation
Pricing is the part of short-term rental management that’s most visible in the spreadsheet and least visible in the listing. The guest never sees it. The cleaning crew never sees it. The dashboard sees it, the deposit sees it, and over a year of decisions, the difference between a thoughtful pricing process and an autopilot one is the difference between a property that pays for itself and a property that just runs.
If you’d like a second opinion on your current pricing — whether you do it yourself or have a manager who does it for you — schedule a free 30-minute consultation. We’ll look at your last 90 days of rates and bookings and tell you specifically where we’d adjust. No pitch required.
HostStarter is a 12.5% flat-fee Airbnb management company. Active dynamic pricing is included for every property we manage. We do the weekly review, the manual overrides, and the monthly reporting — owners just see the deposits.