AssociationAI / AI Literacy
Trihelix AI team Published

Article

Beginner

How to Evaluate a "Proprietary Platform" Claim

A vendor calls its product a proprietary platform. Economists have a name for what you are buying: a credence good. Here is why the demo proves less than you think, and the questions that tell the difference.

Vendor Evaluation Software Procurement Platforms

The pitch arrives as a slide deck: a proprietary platform, built for associations, an operating system for your marketing. The salesperson is sharp, the screenshots are beautiful, the reference client sounds like you. Your board asks whether to sign.

Here is the uncomfortable truth we start from, in Joseph’s words: good salespeople don’t mean good products. Period. This is a guide to telling the difference, written for the staffer who has to make the call without an engineering team to back them up.

Why the pitch proves nothing

Economists have a name for what you are being sold: a credence good. The seller knows more than you do about what you need and about what you received, before the sale and often after it. A 2017 survey of the research literature puts the consequence bluntly: that asymmetry “creates strong incentives for misbehavior on the sellers’ side.” Worse, in the experiments the survey reviews, letting the buyer check the work afterward helped far less than theory predicted.

Software makes it worse, because you cannot pop the hood. A study of enterprise software markets describes the setup: vendors face “the considerably high switching costs for customers and their inability to perfectly observe the real quality of the system at purchase time,” which “gives vendors additional incentives to exaggerate the quality of their software products,” hoping that when you discover the low quality, “it is too expensive and too late to switch to a competitor.” The paper asks outright: should customers believe the vendor-speak about their high product quality?

The demo is necessary but not sufficient

The standard advice is to demand a live demo, and it is good advice. TechSoup’s nonprofit software guide says most software companies offer a free product demonstration, and that by seeing the tool in action you can “start imagining how to adapt it to your organization’s needs.” It goes further: take a free trial, or pay for one month, to see “how easy it is to use and how it fits with the other technology you have.” Make the demo a requirement, not a favor.

But a demo shows you the theater, not the machinery. A recorded walkthrough is a commercial. The questions below are the machinery check: they ask for things a real platform can show and a story cannot.

The machinery check

What is actually proprietary? Ask which parts were built in-house and which are licensed, open-source, or white-labeled. An assembled platform can still be good, but then you are buying integration and service, and the price should reflect that. “Proprietary” should mean something specific, or it means nothing.

Can I see it run, live, on my data? Not a recording. A live session on a scrubbed copy of your member list or event data, run by someone who can answer questions. If the answer is no, ask what a one-month trial costs.

What happens if we leave? Ask how you export your audiences, your data, and your configurations, and in what format. The price of lock-in is paid at exit, when a small association can least afford it. If the export path is vague, the platform is a lease on your own data.

Show me the margin. If the platform resells ad inventory or services underneath the dashboard, ask what share of your dollar reaches the actual media versus fees and markup. And if the pitch includes buying ad impressions “in bulk” at a discount, know that the market’s own protocol prices each impression in its own real-time auction. We walk through those mechanics in our guide to whether your association should pay for a demand-side platform.

Who operates it day to day? A platform nobody on your staff can operate is a service contract wearing a software costume. Price it as one.

A worked example

This is a teaching example, not a case study. A vendor pitches your association a “proprietary audience platform” for $2,000 a month. You ask what is proprietary: a dashboard over licensed data segments. You ask for a live demo on your data: you get a recorded walkthrough. You ask about export: audiences download as a spreadsheet, but the scoring stays behind. Nothing here is dishonest. But what is being sold is a $2,000 dashboard plus a service relationship, not a platform. Price it against a contractor and a spreadsheet before you price it against software.

The stakes are real

Claiming technology you do not have is not a gray area, and regulators have started collecting fines for it. In March 2024 the Securities and Exchange Commission brought its first “AI washing” cases, charging two investment advisers with marketing AI capabilities they did not have; the firms paid $400,000 in combined penalties to settle. The SEC chair’s warning travels well beyond finance: companies “should not mislead the public by saying they are using an AI model when they are not.” A “proprietary platform” claim is the same kind of claim. The pitch is marketing. The substantiation is the product.

The caution

The burden of proof is on the seller, and the proof is never the pitch. Ask the questions, write down the answers, and compare what is promised against what is demonstrated. A vendor with a real platform will welcome the inspection. One without it is selling you the salesperson. And now you know the difference.

Sources