We Read the Public Pages of 27 Virtual Team Building Vendors. Eight Offer a Report. None Say What Is In It.

Of 27 virtual team building vendors whose public pages we read in September 2026, eight offer buyers some form of measurement after the event. None of the eight publish what that measurement contains. No list of metrics, no sample report, no method. Two thirds of the category makes no outcome claim at all, and the single largest number the category quotes was produced by a correlation study that never measured team building.

We are a vendor in this category. We are building a post-event report and planned to call it a difference. Before claiming that, it seemed worth checking whether anyone else already had one. The answer we wanted was "nobody." The answer we got was less convenient and more useful.

What we did, and what would have proved us wrong

We assembled 32 vendors from three sources: a competitor list maintained internally, a set of category incumbents, and the entrants that launched or relaunched between 2024 and 2026. For each one we read the public homepage and, where it existed, the product or pricing page. We asked two fixed questions of every page.

First: what does this vendor claim about outcomes? Morale, connection, engagement, retention, productivity, money. Second: what does this vendor offer the buyer after the event? A report, a dashboard, participation data, a survey, anything.

Answers of "nothing here" were recorded as data, not as a failed read. Usable evidence came back from 27 of the 32. The five we could not reach are listed at the bottom of this piece, along with the one that matters most.

Two hypotheses went in, both able to fail. The first was that vendors overpromise on outcomes. That would be disproved if most of them promised nothing. The second was that essentially nobody sells a post-event report. That would be disproved if a meaningful share of them did.

Both failed. The first failed completely. The second failed in a way that turned out to be more interesting than the version we expected.

Finding one: two thirds of the category promises nothing

Eighteen of 27 vendors make no outcome claim of any kind. Not a percentage, not a soft one. They sell the experience: the host was great, the puzzles were clever, fourteen thousand teams have booked us, five thousand reviews.

That was not what we expected to find. We had prepared to argue with a category that oversells. Instead we found a category that mostly declines to enter the argument.

Seven more make qualitative claims with no number attached. GOTO Events writes that "team building improves communication, loyalty and performance." BreakoutIQ promises to "boost morale, retention and engagement." TeamBonding lists "increased employee engagement" and "boost morale, motivation, and productivity." These are positions, not measurements, and none of the three offers a way to check them.

Two vendors publish actual numbers. They are worth taking separately.

Finding two: the vendors who promise most measure least

Outback Team Building states that "highly engaged teams are up to 23% more productive," that "a disengaged employee costs roughly a third of their salary in lost productivity," and that "replacing one employee costs between half and double their annual salary." Their public pages offer the buyer no measurement instrument of any kind. The claim is borrowed, and there is nothing on offer that would let a buyer test it against their own event.

Teamazing publishes four figures on its product page: "10x Return on Investment," "+30% Leadership Impact," "48% Daily Engagement," "+50% Daily Participation." None carries a sample size, a method, a year, or a source. Teamazing is the only vendor in the 27 that both quotes an outcome number and sells an analytics product, which makes the missing methodology more noticeable rather than less.

Across all 27, there is not one vendor that states a measurable outcome and also hands the buyer a tool capable of testing that specific claim.

About the 23% figure, since the whole category runs on it

The "23% more productive" number comes from Gallup's Q12 meta-analysis. It is real, it is large, and it is widely cited across HR. It is also frequently used to mean something it does not say.

The study compares business units scoring high on engagement against business units scoring low, and reports the difference in outcomes. It is correlational. It does not establish which direction the causation runs, and it contains no measurement of team building whatsoever. Profitable, well-run units may generate engagement rather than the reverse; the design cannot separate those. Gallup also sells the instrument that measures engagement, which is a commercial interest in the size of the effect and belongs in the same sentence as the number.

None of that makes the figure worthless. It makes it a statement about engagement scores in general, not a forecast of what happens after your team plays a game on a Thursday. Placing it beside a booking button implies a chain of causation that the study does not contain.

Finding three: eight vendors offer measurement, and the split is structural

This is where our own assumption broke. We had it recorded internally that nobody in this category sells a report. Eight of 27 do, which is 30%.

Read the right-hand column twice. That is the actual finding. Eight vendors have decided that measurement belongs on the sales page. Zero have decided it belongs in the product documentation. The report exists as a row in a comparison table and not yet as a thing a buyer can inspect before signing.

The second pattern in that list is a business-model split, and it is cleaner than we expected. Five of the eight are self-serve software: Scavify, GooseChase, Office Olympics, QuestWorks, SquadGame. All five report participation, points, task completion, in-game behaviour. Only three of the 22 hosted-event vendors offer anything at all.

The reason is architectural rather than philosophical. A browser-based game generates data as a by-product of being software. A hosted call in Zoom leaves almost nothing behind to count. Vendors running live events are not declining to measure out of modesty. Most of them have nothing to measure with.

Which is an uncomfortable conclusion for us to publish, because it means our own report is protected by our architecture rather than by our idea, and architecture can be rebuilt. Three competitors have started.

Finding four: we ran the audit on ourselves, then published our own arithmetic

Anyone can check this in thirty seconds, so we will go first.

The homepage of this website runs a calculator built on the Gallup correlations discussed above. It takes a team size and an average salary, multiplies by 0.17, and returns a number in dollars. At its default settings it produces a figure in the millions.

When we applied our own two audit questions to it, it failed both. It quoted a number without saying what the number could not show, and it presented a correlation from a different research question in a format that reads as a forecast about buying our product.

There were two ways to fix that. Delete it, or show the arithmetic. We chose the second, and the reasoning matters more than the choice.

The finding of this audit is not that vendors should stop using numbers. It is that the category publishes conclusions and withholds methods. Deleting our calculator would have removed one example of the problem and demonstrated nothing about the solution. So the calculator stays, and it now carries what none of the 27 vendors provides.

Here is what is on the page next to it as of today:

• The formula, written out. team size × average salary × 0.17 for the productivity figure, team size × average salary × 0.18 × 0.59 for the retention figure. Every constant is visible, including the 18% baseline turnover assumption, which is an industry average and not your number.
• What the output actually is. The result is the modelled value gap between a high-engagement and a low-engagement organisation at your headcount and salary. That is the comparison the Gallup meta-analysis makes.
• What the output is not. It is not a prediction of what happens after you run an event with us. No study in this category, ours included, has established that chain. The calculator models a gap; it does not claim we close it.
• The source, and its interest. A direct link to the meta-analysis, and a line noting that Gallup sells the engagement instrument the research is built on.

The number on our homepage did not change. What changed is that a reader can now see where it comes from, what assumption is baked into it, and where it stops being evidence. That took one afternoon of work.

We are not claiming this makes our calculator correct. A correlation used carefully is still a correlation. We are claiming it is now auditable, which is a lower bar than correct and a higher bar than anything else we found in 27 attempts.

If a competitor reads this and does the same thing to their own page, the audit will have been worth more than the traffic.

What to ask a vendor before you sign

The practical version of all of the above, for anyone currently comparing quotes. None of these questions require you to believe or disbelieve anything about team building. They only test whether a vendor can show its work.

1. Show me a sample report from a real event, with the client anonymised. Not a feature list, an actual artefact. Eight vendors currently claim a report. If one can produce a sample on request, that is genuinely new information about them.
2. What exactly is measured, and is it something people said or something people did? A satisfaction survey and a participation log are both legitimate, and they are not the same evidence. Ask which one you are getting.
3. Where does that number come from, and what is the sample size? Applies to every figure on the site. If a vendor quotes a percentage without an n, treat it as a position rather than a measurement.
4. What is your participation rate across events at my company size? Attendance is the one outcome every vendor can measure and almost none publish. The reluctance is informative.
5. What do I have at 60 days that I did not have before? If the honest answer is a photo album and a good memory, that can be a completely reasonable purchase. It is a different purchase than the one the comparison table described.

The limits of this audit

Public pages only. A vendor who sends a thorough report on request but never mentions it on the website is counted here as offering nothing. That means the true share offering measurement is higher than 30%, not lower. It does not weaken the central finding, which is about what the category publishes: if the report is not described on the site, it is not being sold as a difference.

Five vendors could not be read. Trivana and Wildgoose returned errors. Two others served empty pages. The significant gap is TeamBuilding.com, one of the largest players in the category, which returned HTTP 403 on two separate URLs and is therefore absent from all counts above. Their inclusion could move the numbers.

We also did not verify a single vendor claim against a customer. Everything above is what companies say about themselves on their own websites, including the paragraph about us. That is the whole point of the exercise, and it is also its ceiling.

The full method, the vendor-by-vendor table, and the fixed prompt used on every page are available if you want to reproduce it. We would be interested to see someone run it again in six months, particularly on the eight, and particularly on the column that is currently empty for all of them.

Frequently asked questions