How to choose feeder markets for hotel advertising

To choose feeder markets for hotel advertising, start with where your guests already come from, measured as room nights by guest origin market. Rank those markets by size and by how deeply you already penetrate them, then test advertising in a few of them against matched cities where you don’t advertise. The cities where spend creates extra room nights become next season’s feeder markets.

Most hotel and resort groups pick feeder cities by habit. The list looks like last year’s list, which looked like the year before. Sometimes that’s right. Often it means a city that sends you no new guests keeps its budget because nobody checked.

What a feeder market is, and why habit picks them badly

A feeder market is a city or region that sends a meaningful share of a property’s guests. For a beach resort it might be three drive-market metros and two cities with nonstop flights. For a downtown hotel it might be the whole region within a tank of gas.

Habit fails in two ways. It keeps funding cities that would book anyway, because they show up big in the booking report and nobody separates loyal demand from advertised demand. And it never tests the cities that could grow, because they look small in last year’s numbers. The fix is to measure both how much a city sends and how much more it sends when you advertise there.

Step 1: Map room nights by guest origin

Export reservations from your booking engine or PMS (SynXis, Opera, Cloudbeds and the rest all hold an origin field), with guest ZIP or city, booking date, arrival date and room nights. Aggregate to metro level. In the US, the Nielsen DMA is the natural unit because it’s how TV and CTV are bought.

• Keep both dates. Booking date tells you when ads worked. Arrival date tells you when the revenue lands.
• Include OTA reservations where origin is known. Some OTA bookings arrive without a usable address. Track how many, and keep the share steady across the period you compare.
• Exclude groups and contracts. Wedding blocks and corporate contracts don’t respond to consumer media and make small cities look huge.
• Counts only. You need room nights by origin market and date, not guest names, emails or loyalty numbers.

Step 2: Score each market on size and penetration

Room nights alone favour the cities you already own. Divide by households (or adults, or households above an income line that fits your rate) to get penetration. Then plot every candidate on two axes: how big the market is, and how deeply you’re already in it.

• Large and well penetrated (Cities A and B). Your core. Advertising may still add stays, or it may just reach people who were coming anyway. Only a test tells you which.
• Large with low penetration (Cities C and D). The headroom. Worth a test, especially if air access is good.
• Small and saturated (City E). Loyal, close and already booking. Extra media usually buys very little here.
• Small and far (City F). Skip unless something changes, such as a new route.

Step 3: Plan around the booking window

Hotel advertising works on a delay. Someone sees your resort on a smart TV in early March and books a week in July. Two different clocks matter: when they book, and when they stay.

An average hides a long tail. Summer leisure trips to a resort often book months out, while drive-market weekenders may book a week ahead. Look at your own lead-time distribution by origin before you plan.

Two practical consequences. First, time flights to the booking window for the stay period you want to fill, not to the stay period itself: a campaign for a slow September has to run in July and August. Second, measure by booking date. If you read by arrival date the week the flight ends, most of what it produced hasn’t checked in yet.

Step 4: Test CTV in feeder cities against matched holdouts

Connected TV suits feeder-market tests because it’s bought by geography and can be switched on in one city and not the next. Paid social works the same way. The design is simple:

1. Pair similar cities. Match on access (drive time or flights), size, penetration and how their bookings moved over the past year.
2. Advertise in one of each pair. Pick at random where you can. Hand-picked test cities tend to be the ones someone is already excited about.
3. Run six to eight weeks. Long enough for frequency to build in each city.
4. Read for the flight plus a few weeks. By booking date, covering at least your typical booking window after launch.
5. Count total room nights from each city. Direct plus OTA. Otherwise a shift from Expedia to your own site looks like growth.

The same logic applies to any location business running CTV; the guide to measuring CTV ads for local stores covers the channel-specific details, such as household frequency caps and spillover between neighbouring DMAs.

Worked example: two feeder cities, very different answers (illustrative)

The resort runs $20,000 of CTV in City A and $20,000 in City C over eight weeks, and reads bookings for 12 weeks from launch. City B is held back as A’s match and City D as C’s, alongside the resort’s other steady-spend origin markets. The resort’s average daily rate is $240.

1. City A. 1,250 − 1,040 = 210 extra room nights. $20,000 ÷ 210 ≈ $95 each. At $240 a night that’s 210 × $240 = $50,400 of room revenue for $20,000 of media.
2. City C. Direct bookings rose by 120 room nights, which a direct-only report would celebrate. But OTA room nights from City C fell by 80. Net: 40 extra room nights, $500 each, or $9,600 of revenue for $20,000.
3. The decision. Scale CTV in City A and test it next in City B. In City C, the ads mostly moved guests who were coming anyway from Expedia to your site. That has some value (lower commission), but it isn’t new demand. Try a different channel or offer there before spending more.

Notice that the quadrant chart pointed at City C as the bigger opportunity. That is exactly why you test. Headroom says where growth could be. Only a holdout says whether your advertising can get it.

Drive markets, fly markets and what changes the answer

Drive markets tend to book closer in and respond faster, so they’re easier to test and quicker to read. Fly markets depend on things outside your control: a new nonstop route can turn a dormant city into a feeder overnight, and a cancelled one can empty it. Log route changes alongside your media calendar, the same way a restaurant logs a menu launch.

Big local events in a feeder city (a holiday weekend, a school break that starts a week earlier than everyone else’s) move bookings too. If the test city and its match have different school calendars, choose a different match.

Common mistakes when choosing hotel feeder markets

• Ranking by room nights alone. It funds the cities you already own and ignores the ones with room to grow.
• Counting direct bookings only. A shift from OTA to direct looks like lift unless you count both.
• Reading by stay date too early. Most of a flight’s effect hasn’t arrived when the flight ends.
• Testing one city with no match. A single city’s bookings swing with events, routes and weather. Without a comparison, you can’t tell those from the ads.
• Letting brand search and metasearch take credit. They intercept guests who already chose you. Judge them on extra room nights too.
• Including groups and contracts. A single wedding block can make a small city look like a top feeder.
• Picking test cities by enthusiasm. Randomize within matched pairs where you can.

Questions people ask about hotel feeder markets

Feeder markets are a hypothesis until you test them. Map where guests come from, score the headroom, respect the booking window, and let the holdout cities tell you where your next season’s budget belongs.

Written by The Countlift team at First Byte in The Woodlands, Texas. Questions or corrections? Email hello@countlift.com.