United Airlines is defending itself against a proposed class-action lawsuit built around a wonderfully simple question: When an airline sells you a window seat, is it required to provide a window?

Passengers say they paid extra for seats United identified as window seats, only to board and find themselves sitting beside a solid cabin wall. United asked the court to dismiss the case, arguing that “window” describes the seat’s position relative to the aisle rather than promising an actual window or outside view.

U.S. District Judge James Donato was not persuaded. On July 6, he ruled that the breach-of-contract claims could proceed because United’s reservation screens, ticketing terms and boarding passes represented that the passengers had purchased window seats.

United’s position is essentially that everybody understands “window seat” to mean the seat against the wall, whether or not there is a window. The passengers’ response is that United charged them extra for something specifically described as a window seat, and a large blank section of aircraft fuselage does not meet the ordinary definition of a window.

Separating out the principle that almost any lawsuit causing an airline discomfort is inherently good for society, and that any United executive forced to explain the ordinary meaning of “window” is receiving karmic repayment for the company’s treatment of its Million Milers, this case raises a surprisingly relevant question for the enterprise collaboration industry.

What happens when vendors stop selling products and start selling outcomes?

The Window Has Become the Outcome

For decades, enterprise technology purchasing — specifically AV and collaboration technology purchasing — was based largely on identifiable products, features and specifications. A customer bought a codec, camera, microphone, control panel, license or support contract.

The device either had the promised connection or it did not. The software either supported the stated platform or it did not. The button either worked or somebody had to explain why pressing “Join” required three additional screens, a reboot and a sacrificial offering to the interoperability gods.

There were always broad marketing claims, of course, but the actual purchase could usually be reduced to a list of things on a BOM that were supposed to be delivered and functions that were supposed to work.

That is changing.

With InfoComm 2026 in the rear-view mirror for about a month, collaboration companies participating there increasingly described their value in terms such as better meetings, equitable participation, frictionless experiences, smarter workplaces, reduced complexity, improved productivity, proactive management and AI-generated efficiency.

The product still exists, but it is positioned as the mechanism through which the customer buys a better result.

That is understandable. Features are easier to copy than outcomes. Cameras, room systems and management dashboards can look remarkably similar on a comparison chart. Selling a business outcome moves the conversation above individual specifications and, not coincidentally, helps justify a higher price.

But once the outcome becomes the product, the outcome also begins to resemble that airline window.

If a customer paid more because a vendor promised less friction, better meetings, simpler management or increased productivity, is the vendor obligated to provide those things?

If the customer instead receives the technological equivalent of a blank wall, should it be entitled to refunds or compensation?

The answer cannot simply be that “better meetings” refers to the platform’s position in the collaboration architecture and does not necessarily mean the meetings will actually be better. There are judges and courtrooms around the world every bit as frustrated with technology as the rest of us, and I suspect more than a few would be willing to let a United-style window-seat lawsuit against a collaboration company proceed.

Marketing Promise or Contractual Commitment?

The obvious defense is that business outcomes are influenced by far more than technology.

A platform cannot make an unprepared executive run a useful meeting. A camera cannot force remote participants to pay attention. An AI assistant cannot repair a dysfunctional organization, overcome a bad network or make a room work when the microphone is installed directly above an air-conditioning vent.

Those are legitimate points.

Outcomes depend on product performance, integration, network readiness, user behavior, training, adoption, room design, management processes and organizational culture. No vendor can reasonably guarantee every result when it controls only part of the environment.

The second defense is that phrases such as “better meetings” and “less friction” are marketing language, not measurable contractual commitments. They are aspirations rather than warranties.

That may also be true, but it creates an uncomfortable contradiction.

Vendors want customers to justify purchases using measurable business value. Buyers are encouraged to consider time saved, support costs reduced, adoption increased, downtime prevented, travel eliminated or productivity improved.

Then, if those outcomes fail to appear, the language suddenly becomes inspirational branding that no reasonable buyer should have interpreted literally.

That is the collaboration industry’s version of arguing that a window seat merely means a seat against the wall.

The Contract Still Buys the Seat

The deeper problem is that most enterprise technology contracts are still structured around delivery rather than results.

The hardware arrived. The licenses were activated. The implementation was completed. The rooms passed acceptance testing. The dashboard displayed the devices. The invoices were paid.

From a contractual perspective, the vendor may have done exactly what it agreed to do, even when the promised transformation never materialized.

The sales team sold an outcome, but the contract documented components. The buyer approved the purchase based on a vision, but accepted the deployment based on whether the screens turned on.

This gap is where accountability disappears.

The salesperson can claim the product was delivered correctly. The implementation team can blame the customer’s environment. Support can point to exclusions. IT can blame poor adoption. Users can keep joining from laptops because the room system is confusing.

Everyone can be technically correct while the organization still fails to receive the result that justified the purchase. Meanwhile, the original sales presentation featuring attractive people having life-changing meetings in immaculate rooms has been lost somewhere in the turbulence at thirty thousand feet, usually somewhere between sales and delivery.

Hear Me Out: Outcome Protection

A full money-back guarantee for every disappointing meeting would obviously be absurd. Technology vendors cannot be held financially responsible because Bob scheduled 14 people for a meeting that should have been an email.

But there is a reasonable middle ground between guaranteeing human happiness and accepting no accountability for outcomes at all.

When a vendor makes a specific, measurable outcome central to the sale, that outcome should appear somewhere in the agreement.

The parties should define the starting condition, promised improvement, measurement method, customer dependencies, implementation period and remedy if the result is not achieved.

That remedy need not be a full refund. It could be additional services, corrective engineering, extended support, service credits, delayed payments, retraining or a limited right to terminate.

The point is to make the agreement reflect the value proposition that convinced the customer to buy.

Some outcomes are measurable. Room uptime can be tracked. Call-setup failures can be counted. Support tickets can be compared before and after deployment. Mean time to resolution, device-management labor, adoption rates, meeting-start times and recurring technical failures can all be analyzed.

Other outcomes, such as improved creativity or happier employees, are much harder to isolate.

A vendor should not be required to guarantee something that cannot reasonably be measured. By the same logic, it should not use an unmeasurable claim as the primary justification for a premium price and then act surprised when the buyer asks why the promised outcome never materialized. That would be like paying extra for a window seat to Hawaii, spending the entire flight staring at a blank wall and somehow landing in Cleveland.

Shared Responsibility Still Requires Responsibility

Buyers would also have obligations.

They would need to provide an accurate baseline, meet network and infrastructure requirements, deliver training, follow the implementation plan and avoid changing the environment halfway through the measurement period.

Outcome accountability cannot mean customers receive refunds after ignoring every deployment recommendation. It should mean both parties identify what each controls and what each must do.

That conversation would improve procurement even when no financial guarantee is offered.

Before approving an outcome-based purchase, customers should ask what the promised outcome means, how it will be measured, what conditions are required, who owns each dependency and what happens if the result does not appear.

In other words: Where is the window?

If the answer is clear, measurable and documented, the customer may genuinely be buying an outcome.

If the answer is a collection of glossy phrases about transformation, intelligence, experiences and the future of work, the customer may simply be paying extra for a seat against the wall.

A Window Means a Window

The United case is still at an early stage. The passengers have not won, and United has not been found liable.

But the judge’s decision reflects a basic principle that applies well beyond air travel: Words used to sell a premium product should retain their ordinary meaning after the customer pays.

The enterprise collaboration industry wants to move beyond selling boxes, licenses, features and compatibility announcements. That is probably the correct direction.

Customers care more about whether meetings work than which processor is inside the room system. They care more about reducing support burdens than the number of charts on a dashboard. They care more about accomplishing work than collecting another set of AI features.

But selling outcomes creates a higher standard than selling components.

A vendor cannot claim credit for the business result when closing the deal and retreat to the technical specifications when the result fails to appear.

If the industry wants customers to buy better meetings, less friction, smarter management and improved productivity, it should be prepared to define what those promises mean and accept some responsibility for delivering them.

Otherwise, the customer is not buying an outcome.

It is buying the right to sit beside a blank wall while the salesperson who promised the view stands at the gate, watches your plane take off and laughs while counting the upgrade money. By the time you realize there was never a window, he is already selling the next passenger an equally unverifiable sunset.