Original research, August 2026
How Georgia cities take park reservations
We surveyed the public websites of all 536 municipalities in Georgia to answer one question: when a resident wants to rent a pavilion, a ball field, or a community room from their city, what actually happens? The answer divides almost entirely along population, and the line is sharper than we expected.
What we found
- Of the 124 cities under 1,000 residents whose websites we read, not one showed any recreation or reservation platform in use. Between 1,000 and 5,000 residents, 8 of 138 did.
- Above 50,000 residents, 11 of 19 cities did. Adoption rises with every population band without exception.
- 392 of Georgia’s 536 cities have fewer than 5,000 residents. That is 73 percent of the state’s municipalities and about 506,000 residents.
- 129 cities have no website at all listed in the Georgia Municipal Association directory, and 114 of those are cities under 1,000 residents.
- Among the 141 cities where we could determine the process with confidence, 20 offer a true online reservation system. 45 take reservations by telephone only.
Recreation software adoption by population
A city is counted here only when its own website links to, embeds, or loads assets from a known recreation or reservation platform. The denominator is the number of cities in the band whose website we were able to read, because a city we could not reach is a city we cannot make a claim about.
| Population | Cities | No website | Sites read | With software | Share |
|---|---|---|---|---|---|
| Under 1,000 | 238 | 114 | 124 | 0 | 0% |
| 1,000 to 5,000 | 154 | 15 | 138 | 8 | 6% |
| 5,000 to 15,000 | 65 | 0 | 63 | 13 | 21% |
| 15,000 to 50,000 | 58 | 0 | 58 | 22 | 38% |
| 50,000 and over | 19 | 0 | 19 | 11 | 58% |
| All bands | 534 | 129 | 402 | 54 | 13% |
Two cities, Mableton and Mulberry, are both very recently incorporated and carry no population figure in the source data, so they sit outside the bands. Three further cities had a website that could not be read. Neither group showed recreation software.
The pattern is not that small cities looked at recreation software and declined it. It is that the products are packaged and priced for departments with program catalogs, league play, memberships, and a front desk, and sold with implementation fees and multi-year terms. A city of 2,400 with four pavilions has never been a viable customer for that shape of product, so it was never really sold one.
How a reservation actually happens
These 141 cities are the ones where the process was stated clearly enough on the city’s own pages to record it with confidence. Cities whose pages were ambiguous were excluded rather than guessed at, which is the largest single limit on this study and is described in the methodology below.
| Published reservation channel | Cities | Share |
|---|---|---|
| Several channels at once, for example call the office or download a form | 58 | 41% |
| Telephone only | 45 | 32% |
| A true online reservation system | 20 | 14% |
| A printable PDF form | 5 | 4% |
| No reservation information published at all | 5 | 4% |
| Email only | 4 | 3% |
| In person only | 2 | 1% |
| Could not be determined | 2 | 1% |
Fourteen percent of these cities can take a resident from availability to a confirmed booking on the web. Everything else routes through a person: a phone call during office hours, a form printed and returned, an email that waits for someone to open it. Separately, among the 62 cities where we could determine it, 29 accept any kind of online payment for a facility.
The most common single answer, 58 cities, is several channels at once: a page that says to call the office, and also links a PDF, and also gives an email address. That is usually what a manual process looks like after a few years of well-meaning additions to the website.
Who the incumbents are
Across all 404 cities we read, 54 showed a recognized recreation or reservation platform. Every one of those detections rests on a booking domain or an embedded asset belonging to the vendor, not on a content management system footprint, which is a distinction that materially changes this table.
| Platform | Cities |
|---|---|
| CivicRec, part of CivicPlus | 27 |
| RecDesk | 14 |
| ACTIVE Network, ActiveNet | 5 |
| MyRec | 4 |
| Vermont Systems, RecTrac | 1 |
| PerfectMind | 1 |
| Amilia | 1 |
| CommunityPass | 1 |
| Total | 54 |
Two platforms hold three quarters of the installed base in Georgia. The concentration matters less than the size of the field it sits in: 54 cities out of 404means the entire competitive market for recreation software in this state is about one city in eight, and it is almost entirely the larger ones.
What cities actually rent
Across 139 cities that list their facilities, we counted 886 of them, a median of five per city. The mix is a useful corrective to how recreation software is usually demonstrated. This is not a market of rec centers and program catalogs. It is a market of pavilions.
- Pavilion or shelter76 cities
- Playground60 cities
- Community center28 cities
- General park space28 cities
- Basketball court27 cities
- Baseball field23 cities
- Tennis court23 cities
- Walking trail22 cities
- Dog park18 cities
- Gymnasium16 cities
Methodology
The roster of 536 municipalities comes from the Georgia Municipal Association member directory. For each city we resolved its official website from its GMA profile page, then read the parks, recreation, facilities, and rentals sections of that site with an automated crawler that follows only links on the city’s own domain, obeys robots directives, and rate-limits itself per host. 404 cities were read this way.
Recreation platforms were identified deterministically, by matching links, embedded frames, form targets, and script sources against the known booking domains of eight named vendors. No language model was involved in naming a vendor. Detections carry a confidence value, and only detections at or above 0.70 are counted here.
The reservation process itself was extracted from the collected page text by a language model constrained to cite a source page for every claim it made. Claims citing a page we had not actually fetched were discarded, and any extraction below the confidence threshold was marked for review and excluded from the process figures. Every figure on this page traces back to a specific page on a specific city’s website.
Limits
Each of these pushes the results in a knowable direction, and we would rather state them than let a reader assume the survey is tighter than it is.
- 1Pages were read as delivered in HTML. Content that appears only after JavaScript runs, including some staff directories and a few embedded booking widgets, was not seen. This biases the vendor count downward.
- 2Recreation programs run by a county, an authority, or a nonprofit on the city’s behalf were not counted as city software. A resident in one of those cities may well book online, through an organization that is not the city.
- 3Vendor detection recognizes eight named platforms by their booking domains and embedded assets. A city using a platform outside that list, or a general-purpose scheduling tool, would be recorded as having none.
- 4Reservation process was extracted by a language model from the text of the city’s own pages, and every extraction below a confidence threshold was excluded rather than counted. 157 cities were excluded on that basis, which is why the process figures rest on 141 cities and not on all 404.
- 5A city that publishes nothing about reservations may still rent facilities. Absence of published information is recorded as exactly that, and not as absence of a program.
- 6Everything here is a point-in-time snapshot of public web pages in August 2026. Cities change vendors and redesign sites.
A note on named cities
We hold a per-city record, and we are not publishing it. How a city runs its parks office is that city’s business, and no clerk should read a vendor’s description of their department on the open web. If you work for a Georgia city and want to see what we recorded about yours, including the source pages behind it, emailparks@parkspilot.com and we will send it to you, correct it if it is wrong, or remove it if you prefer.
Study conducted August 2026 by Kahnputers LLC, the maker of ParksPilot. We build reservation software for exactly the cities this study finds underserved, which is the reason the study exists and a fact you should weigh when reading it. The figures and the method are published in full so that anyone can check them. Corrections are welcome at parks@parkspilot.com.
If your city is in the 86 percent
Taking reservations on the phone is not a failure. It is what was available. See what the alternative looks like, in a complete working demo city, before anyone calls you.