Three products land in five out of six answers, driven by a single domain
Three scheduling tools appeared in five of the six sets of recommendations we got from ChatGPT, Gemini, and Perplexity: Acuity Scheduling, Calendly, and Setmore. We asked each assistant six questions verbatim, from the straightforward “best appointment scheduling software” to “best appointment scheduling software for small teams,” “what appointment scheduling software should I use,” “best free appointment scheduling software,” “most affordable appointment scheduling software,” and “appointment scheduling software recommendations.” All six queries returned readable lists of named products. The assistants agreed so frequently on those three that the real surprise lies in what happened on the edges: one review site fed every answer, and a simple modifier like “for small teams” gutted the consensus.
emailtooltester.com surfaced as a source in all six answers, across all three assistants. No other domain came close to that coverage. lunacal.ai, reddit.com, and zapier.com each appeared in five of six replies, but emailtooltester.com was the constant. For someone selling appointment scheduling software, that means visibility on that one aggregator likely shapes AI-generated recommendations far more than any other single publication. The lens is narrow, even though the recommendations themselves look diverse on the surface.
Diversity among the product names is real, but it frays fast. Beyond the three leaders, SimplyBook.me, Square Appointments, and Zoho Bookings each got three mentions. Appointlet, Appointy, Lunacal, and YouCanBookMe each showed up twice. Then ten products earned a single mention: AnswerConnect, Appointment.com, Baluu, Bookedin, Booksy, Bookwhen, Cal.com, Koalendar, NeetoCal, Picktime, PracticeSuite (plus variant spellings that the assistants treated as separate entities). SimplyBook.me appeared three times, but its lowercase twin simplybook.me got one additional mention. Square Appointments, Square, and Square Booking were each counted individually. The tools made no visible effort to consolidate brand variants, which means a vendor’s name recognition can be penalized by its own inconsistent web presence, even within a small set of six prompts.
The small-team question scrambled the rankings completely. For “best appointment scheduling software for small teams,” the products named were Acuity Scheduling, Appointlet, Appointy, Appointment.com, Baluu, Bookedin, Bookwhen, and Booksy. Calendly and Setmore, which both ran five-for-six overall, did not appear in that answer. This is not a trivial shift. A buyer who qualifies their search with “small teams” gets a slate that excludes two of the most perennially recommended tools, and that buyer might never see them. The sources for that answer included emailtooltester.com, lunacal.ai, calendly.com, reddit.com, connecteam.com, youcanbook.me, ruby.com, and zapier.com. calendly.com showing up in the source list while Calendly itself did not make the product list suggests the assistant pulled comparison pages that mentioned Calendly in passing, or evaluated it and left it out, which we cannot determine from the output alone.
The free-software question introduced yet another cohort. “Best free appointment scheduling software” returned Calendly, Square Booking, Setmore, Picktime, Zoho Bookings, SimplyBook.me, and Koalendar. Picktime and Koalendar were one-off mentions in the overall data, surfacing only under this prompt. That is a small signal but a meaningful one: a tool with no presence on general lists can become the answer when the framing narrows to a price tier.
The sourcing strategy itself raises questions. lunacal.ai was one of the most cited domains and Lunacal is also a product that appeared in two answers. Vendor-owned domains including calendly.com, squareup.com, practicesuite.com, and acuityscheduling.com fed multiple responses. The assistants routinely treated vendor content as recommendation-worthy, blurring the line between independent review and self-promotion. For a user, the output feels like a curated list. For the vendors inside that loop, the path to being named runs through publishing comparison content on their own site and getting it indexed by the review sites the assistants already trust. For everyone else, the gap between zero mentions and even two mentions looks difficult to bridge without a presence on emailtooltester.com and reddit.com.
The leaderboard: five out of six, three out of six, and the long tail of one-offs
No product appeared in all six answers. The closest were Acuity Scheduling, Calendly, and Setmore, each of which showed up in five of the six sets of recommendations. That missing single answer is the gap between a close-to-default pick and an actual default. For a product manager, missing one query out of six across ChatGPT, Gemini, and Perplexity means the assistants nearly always volunteer these three names regardless of how the question is phrased. For someone shopping, that consistency says these are the tools you will see over and over.
SimplyBook.me, Square Appointments, and Zoho Bookings landed in three answers each. That is half of the queries, a solid showing but well below the top group. A buyer might still discover these tools through this category, but they would need to ask a few different questions or try a different assistant to run into them. Notably, Square Appointments and the shorter “Square” appeared as separate entries (Square in one answer, Square Booking in one more), and SimplyBook.me had a lowercase variant (“Simplybook.me”) logged once. If those are the same product under different renderings, the real count for Square’s scheduling tool might be four or even five, and SimplyBook.me might be four instead of three plus one. The data does not resolve that, but a vendor scanning this should know that name variants can split their tally.
Four products reached two mentions: Appointlet, Appointy, Lunacal, and YouCanBookMe. Lunacal is the outlier here. It is a newer name compared to Appointy or YouCanBookMe, yet it matched their count. Its own domain, lunacal.ai, also fed five of the six answers, which might explain the lift. A product appearing twice across six wide-framing questions is getting just enough traction to register, but not enough to feel like a safe recommendation. For a sales team, two mentions is a signal that the product is present in the sources these assistants pull from, but not yet sticky enough to break into the top three.
The rest of the field is a list of single-mention products: AnswerConnect, Appointment.com, Baluu, Bookedin, Booksy, Bookwhen, Cal.com, Koalendar, NeetoCal, Picktime, PracticeSuite, and YouCanBook.me (a dotted variant of YouCanBookMe). A single mention is the difference between being in the consideration set and being invisible. If a buyer asks exactly the question that triggered that mention, they find the product. One question different and it is gone. That does not mean the tool is weak. It means the assistances’ source pool, weighted heavily toward a handful of domains like emailtooltester.com (used in all six answers), lunacal.ai (five), reddit.com (five), and zapier.com (five), simply did not surface these names more often.
What is missing from these lists matters as much as what appears. Doodle, a well-known scheduling poll tool, did not get named once. OnceHub, ScheduleOnce, HubSpot Meeting Scheduler, and Travis CI’s scheduling-adjacent tools are absent. Even Microsoft Bookings, which ships with many business subscriptions, never appeared. A reader who expects those names might conclude the assistants have a blind spot, or that the specific queries we asked (“best appointment scheduling software,” “most affordable,” etc.) led the models toward a different set. Absence from our six questions is not absence from the category. A vendor who sells one of those missing tools should not take this as a verdict, only as a snapshot of what six prompts returned on the day we ran them.
The spread from five mentions to one is steep. The top three nearly swept the board. After that, the counts drop quickly to three, then two, then a long flat tail. That shape tells its own story: the assistants are converging on a small set of brands, and everything else is a peripheral mention that a user might see once and never again. If your product is in the tail, the assistants are not actively working against you. You simply are not part of the default answer yet.
The full list, counted
Every product serper+or:chatgpt,gemini,perplexity named across 6 answers, with how many of those answers named it. The share column is out of the answers we read, not out of the category.
| Product | Named in | Share | Read as |
|---|---|---|---|
| Acuity Scheduling | 5 of 6 | 83% | 5 answers |
| Calendly | 5 of 6 | 83% | 5 answers |
| Setmore | 5 of 6 | 83% | 5 answers |
| SimplyBook.me | 3 of 6 | 50% | 3 answers |
| Square Appointments | 3 of 6 | 50% | 3 answers |
| Zoho Bookings | 3 of 6 | 50% | 3 answers |
| Appointlet | 2 of 6 | 33% | 2 answers |
| Appointy | 2 of 6 | 33% | 2 answers |
| Lunacal | 2 of 6 | 33% | 2 answers |
| YouCanBookMe | 2 of 6 | 33% | 2 answers |
emailtooltester.com in every answer, and the odd power of a vendor’s domain
No single domain fed all six answers, but emailtooltester.com came closest, appearing as a source in every set of results we examined. The second tier, each showing up in five of six answers, included a vendor’s own site (lunacal.ai), a forum (reddit.com), and a blog from an integration platform (zapier.com). That top group tells you a lot about what it takes for a page to shape recommendations in this category: one dedicated review site plus a handful of community and vendor pages that consistently surfaced regardless of how the question was phrased.
emailtooltester.com was the only source that ChatGPT, Gemini, and Perplexity all used for every single query, from the broad “best appointment scheduling software” to the more specific “most affordable appointment scheduling software” and “appointment scheduling software recommendations.” A review site that tests and compares email and scheduling tools, emailtooltester.com effectively defined the skeleton of the recommendations. Its lists, rankings, and pricing breakdowns were pulled into answers even when the assistants also layered on other sources. For a vendor, that means your presence or absence on that one domain is the single biggest outside factor in whether you get named. For a buyer, it means that what the assistants present as an original synthesis often starts with one editor’s top picks.
What sat alongside it is stranger. lunacal.ai appeared in five answers, placing it ahead of almost every other vendor domain and even ahead of sites like calendly.com and squareup.com. Lunacal the product was only named in two of six answers. So Lunacal’s own pages were being used as source material for recommendations that frequently excluded Lunacal. The most plausible explanation is that lunacal.ai hosts comparison pages or listicles that the assistants’ retrieval systems pulled in when building answers for competitors. A vendor site can become a high-citation resource that surfaces other tools, a dynamic you rarely see with standard SEO but that makes sense when an assistant stitches together snippets from across the web.
Reddit.com fed five answers, which is not surprising for a category full of small-team buyers asking for real-world experiences. Threads from reddit.com were cited for the open-ended “what appointment scheduling software should I use” and the pricing-focused “most affordable appointment scheduling software,” among others. The assistants treated Reddit as ground truth, pulling in named products like Setmore, SimplyBook.me, and Square Appointments from comment threads. Quora.com appeared in only two answers, so Reddit clearly dominated the forum signal.
Zapier.com also fed five answers, acting as a connector between the review-site world and the how-to world. Zapier’s blog posts often list scheduling tools that integrate with other apps, and that angle seems to have made it sticky across multiple assistants. YouTube.com appeared in three answers, adding a video layer that surfaced an odd assortment: product demos, roundups, and maybe a spoiler for someone who makes comparison content on video. That mix of text and video sources is a reminder that an assistant’s answer can be built from a transcript of a YouTube video just as easily as from a written review.
Further down, vendor domains like calendly.com, squareup.com, and waitwellsoftware.com appeared in three answers each. These are the domains behind products that were named often, so the appearances likely reflect the assistants double-checking feature pages or pulling pricing from the official source. practicesuite.com, youcanbook.me, and quora.com each surfaced twice. A long tail of single-appearance vendor domains (acuityscheduling.com, picktime.com, setmore.com) shows that even when a product was named five times like Setmore, the assistant pulled directly from setmore.com only once, preferring third-party sources.
The pattern is not just “review sites win.” It is that one review site, fed into every answer, became the backbone, while vendor domains that publish comparison content (like lunacal.ai) got pulled in almost as often as Reddit and Zapier, whether or not they boosted their own product. That suggests a practical lever for a company not on the list: if your competitive comparison page is authoritative enough to be cited, your domain can enter the source rotation without your product being the subject of the answer, which might slowly shift the entire recommendation landscape.
The source-product link: why the top three appeared and Lunacal didn't
The names an assistant spits out are a reflection of the pages it pulls from, not an independent ranking of features. That explains why Acuity Scheduling, Calendly, and Setmore reached five of six answers while other tools topped out at three or fewer. The single domain that fed all six sets of results, emailtooltester.com, almost certainly carried those three products across the finish line. The same cannot be said for other high-frequency sources. Lunacal.ai appeared in five answers but Lunacal the product only surfaced twice. That gap is the clearest signal in the data that source presence does not automatically turn into product presence.
The top three’s consistency points to emailtooltester.com as the backbone recommendation engine here. The site’s reviews or comparison pages were surfaced by the assistants no matter how we phrased the question. In the raw per-query breakdowns, when emailtooltester.com was a source, all three products appeared together in most sets. The two queries where one of them dropped out (we can’t see which query lost which product from the aggregate alone) coincided with different source compositions. That suggests that missing emailtooltester.com on a query would have cost some of these products their spot. But since the domain was present every time, the risk never fully materialized. For Calendly, there was an additional lift: calendly.com itself fed three answers, likely including the product’s own name whenever that domain was pulled. Acuity Scheduling had its own domain appear only once, yet still made five answers, which supports the idea that third-party coverage, not a vendor’s own site, carried the weight.
The next group, products named three times, splits along obvious source lines. Square Appointments’ count matches squareup.com appearing three times. If the assistant pulled Square’s domain, it got Square’s tool. Zoho Bookings, also at three, doesn’t have a dedicated source domain in the top tier, so it probably rode in through the same review sites and forum threads that boosted the leaders, just less consistently. SimplyBook.me, the third product at three mentions, shows a separate ambiguity: the data includes both “SimplyBook.me” and “simplybook.me” listed as distinct products, one appearing three times, once as a separate lower-case entry. Treating them as the same entity would push its count to four, but the raw output suggests the assistants sometimes treated them as variants. That muddies any simple attribution.
Then there is the long tail. Appointlet and Appointy were both named twice, and Lunacal was named twice despite its domain touching five answers. The rest, from AnswerConnect to YouCanBook.me (with varying capitalisation), each appeared once. The one-offs map cleanly to the low-frequency domains. PracticeSuite appeared when practicesuite.com was a source. Picktime surfaced alongside picktime.com. Koalendar appeared only in the “best free” query, where setmore.com was a source and where Square Booking, a product name that doesn’t appear elsewhere, got a single mention. These products were invisible unless a domain that talked about them made it into the assistant’s retrieval set.
What the measurement cannot tell you is the mechanism. A domain appearing as a source does not mean the assistant adopted its recommendations wholesale. We know the assistants pulled content from these pages, but we do not know which parts of a page influenced the final list. A site might mention ten tools, but only three survive in the assistant’s output. Lunacal.ai’s own site fed five answers, yet Lunacal was named only twice. That could mean the assistant picked up the site for reasons other than recommending its product, perhaps because it ranked for a broader term and the assistant summarised a page that mentioned Lunacal among many alternatives. Or it could mean the assistant actively demoted a vendor’s self-promotion. The data gives no answer.
What is clear is that the contest for assistant visibility in this category is a contest for domain presence on the pages the assistants tend to retrieve. emailtooltester.com is the closest thing to a gatekeeper right now. Calendly.com, squareup.com, and a handful of others occasionally open side doors. Getting onto those domains is a necessary but not sufficient condition. For a product manager reading this, the gap between a five-answer tool and a two-answer tool is not a gap in scheduling capability. It is a gap in how many of those five high-frequency domains mention the product prominently and how well those mentions align with what the assistants’ models decide to retain.
You are defending a placement on one review site, not a category lead
Acuity Scheduling, Calendly, and Setmore each appeared in five of the six answers we collected. The only source that fed every single answer was emailtooltester.com. That is not a coincidence. Those three products are not riding broad assistant awareness of their brand, a deep moat of user reviews, or a dozen independent endorsements. They are riding one domain that the assistants all pulled from, regardless of how the question was phrased. If you sell one of these, what you are actually defending is your spot on that specific page.
The position looks safer than it is because five out of six feels like consensus. It is not consensus. It is a single editorial decision retransmitted through ChatGPT, Gemini, and Perplexity. Remove Acuity Scheduling from emailtooltester.com’s top picks for a month, and you would see that product slip from five answers to two or three overnight, replaced by whatever that site chooses to feature instead. That is the difference between a durable product recommendation and a snapshot of one site’s rankings at crawl time.
The next tier is even more fragile. SimplyBook.me, Square Appointments, and Zoho Bookings each showed up in three answers, exactly half. They missed the other half because the source mix shifted. In the query “best appointment scheduling software for small teams,” none of those three appeared; the assistants leaned on connecteam.com, youcanbook.me, and ruby.com. To stay in three answers, these products must hold their ground across a looser set of sources (reddit.com, zapier.com, and lunacal.ai each fed five answers), not just one. A single change in a zapier.com roundup or a reddit thread could push Zoho Bookings down to a single mention, invisible to the assistant that previously recommended it for “most affordable” or “recommendations.”
For the fifteen products that were named only once (Baluu, Bookwhen, Booksy, Koalendar, PracticeSuite and others), there is no position to defend. They surfaced because a single query triggered a domain that happened to include them. That mention can vanish without any product change. These companies may not even know they appeared. If they acted on it, they would be chasing a ghost.
The clearest threat to the current leaders is what the data shows for Lunacal. Lunacal’s own domain fed five of the six answers, yet the product was named only twice. Source presence does not guarantee product mention. But if Lunacal secured a spot on emailtooltester.com, the one domain that consistently converts to a named recommendation, its mention count could jump from two to four or five almost instantly, displacing one of the incumbents. That is the attack surface. Any competitor that can earn a feature on the review site that underpins every assistant’s answer can overturn the leaderboard without changing a single line of product code. The current top three are defending a single placement. They lose it not when a rival builds a better scheduler, but when that rival gets a better mention on emailtooltester.com.
If you were not named, you are simply not on the pages the assistants used
Three products appeared five times, another three appeared three times, and then over a dozen names surfaced once or twice and immediately dropped out of the pattern. If you sell appointment scheduling software and your product is entirely absent from the six answers we collected, the reason is not ambiguous. You are missing from the specific pages that ChatGPT, Gemini, and Perplexity retrieved for these queries. Visibility in this category is not a brand-building exercise or a reward for a better product. It is a placement problem on a tiny number of source pages.
The most important domain is emailtooltester.com. It fed all six answers. Acuity Scheduling, Calendly, and Setmore, the three products that appeared five times, are almost certainly listed on that site’s comparison or roundup article. If your product is not on that page, no assistant will surface it no matter how someone asks the question. The other high-frequency sources, zapier.com, reddit.com, lunacal.ai, waitwellsoftware.com, and a handful of vendor or forum pages, each appeared in two to five answers, but none of them offered the blanket coverage of emailtooltester.com. Getting onto those secondary pages can earn you a mention in some answers but never all of them. For a seller not named at all, the first concrete step is not social media, not SEO, not a product launch. It is to get onto the specific article on emailtooltester.com that the assistants are pulling from. That means outreach to the site owner, submitting a review request, possibly offering a demo account, and asking to be included. This is not a passive process. Unless you are already on that page, no amount of product improvement will move you into these assistant recommendations.
The mechanism also clarifies what does not move the needle. Driving traffic to your own site does not matter here because the assistants rarely pulled vendor domains as a primary source. Only two vendor domains, acuityscheduling.com and setmore.com, appeared even once, and those appearances were late in the data and did not correlate with a product naming advantage. Acuity Scheduling was named five times but its own domain surfaced in only one answer. So ranking your own site for “best appointment scheduling software” will not get you recommended. The assistants are not scanning the open web at query time; they are retrieving specific pages they have been trained to or that appear in their search indices for these queries. Those pages are predominantly third-party comparison lists, not property pages or blog posts from the software companies themselves. Similarly, advertising in the assistant interfaces does not exist as an option, and paying for inclusion in these source pages is often not a straightforward transaction. The work is closer to PR: convincing a reviewer or editor at emailtooltester.com that your product belongs in their list. If they update that page and you are added, you would likely see your name appear in responses across all three assistants, because six of six answers already used that domain.
There is a second constraint, visible in the data from lunacal.ai. That vendor’s domain fed five of the six answers, yet the product Lunacal was named in only two. Being on a source domain is not sufficient; you must be in the specific block of text that the assistant extracts and treats as a recommendation list. If the page mentions you in a hosted banner, a sidebar, or a separate paragraph about alternatives, the assistant might skip it. The retrieval step that picks out product names is unreliable. So even getting onto emailtooltester.com requires checking how the page is structured. If your listing appears below a fold or after the main ranked items, you might still be invisible. The work, therefore, is not just “get listed on this site.” It is “get listed inside the main comparison table or the top-N list that the assistant is parsing.” That is a negotiation over page layout, not just a name drop.
The pattern leaves sellers with a stark conclusion. The questions we asked covered broad and narrow intents, from “best free” to “most affordable” to “for small teams.” The source pool barely changed. The pages that shaped the answers are extremely stable across queries. If you are new or overlooked, your path to appearing is narrow: identify the exact pages on emailtooltester.com that rank for these terms in the assistants’ backends, get your product into the primary list on those pages, and verify that the extraction picks you up by running test queries. Nothing broader than that will have any effect.
Running the measurement yourself: six questions, three assistants, and one tedious afternoon
The single most useful thing we learned running this is that you can reproduce it for any software category with nothing but a browser, a spreadsheet, and a stomach for grunt work. Pick a category, write six questions the way a real buyer might phrase them, and ask those same questions to ChatGPT, Gemini, and Perplexity. That is it. A morning spent copy-pasting will surface the same concentration of sources and the same handful of product names that a more elaborate crawl would spit out. The tedium is the price of seeing which specific pages actually feed the answers.
Start with the questions. You want enough variation that the assistants cannot simply rerank the same list each time. We used “best appointment scheduling software,” “best appointment scheduling software for small teams,” “what appointment scheduling software should I use,” “best free appointment scheduling software,” “most affordable appointment scheduling software,” and “appointment scheduling software recommendations.” The exact wording matters less than covering broad picks, small-team picks, free, cheap, and an open-ended recommendation. Six is a workable number. Fewer and you miss edge cases; more and the manual tally becomes draining.
Open three browser tabs, one for each assistant, and paste the first question into all three. Copy every product name the assistant returns into a spreadsheet row labelled with the question. If the assistant cites sources, copy those URLs into a separate column. Repeat for each question. You will do this eighteen times (six questions times three assistants). What you discover quickly is that the answer formats are inconsistent. One assistant gives a bullet list with bolded names; another gives a paragraph; a third splits recommendations into tiers. You will be trimming stray asterisks, deciding whether “Square” and “Square Appointments” are the same product, and trying not to lose your place as you scroll back to the first tab. You will also notice that some assistants show source links only as clickable numbers. That means you either click through to log the domain or you skip the source layer entirely and still get a useful product count, though you lose the most revealing part of the picture.
The product tally is where the real monotony sets in. You comb through the spreadsheet, consolidate variants, and count how many of the six question rows mention each product. For us that meant spotting that Acuity Scheduling appeared in five rows, while Booksy appeared in one. You do the same for source domains. You will see emailtooltester.com pop up again and again. You will see a vendor’s domain, too, and then notice that the vendor’s own product barely surfaced relative to the domain’s presence. That exact gap only becomes visible because you are staring at a grid of counts and you cannot unsee it. If you skip the source logging to save an hour, you lose the entire story of why products rank where they do. Doing it by hand means you feel in your fingers how few pages actually prop up the whole edifice. That feeling, not the numbers, is what makes you rearrange a marketing budget.