Key takeaways
- A crowdfunding campaign needs a works-like prototype that actually performs the promised features, not just a looks-like model that shows the finished form.
- EVT, DVT and PVT describe different validation goals, not a fixed unit count; what matters before launch is proving the core function works, not following the sequence perfectly.
- Two well-documented campaigns ran into serious trouble after their campaign videos showed more capability than the prototype actually had, which became the standard backers held the eventual product to.
- How many physical units you need, and how polished they must look, depends on the product’s complexity and cost, not a single rule of thumb.
A crowdfunding campaign needs a works-like prototype: one that actually performs the features you are promising backers, demonstrably, on camera. A looks-like model that only shows the finished form is not enough on its own: Kickstarter’s own rules require a working prototype for exactly this reason, and backers generally judge whether the product works, not whether it looks finished.
Looks-Like, Works-Like and the Combined Prototype
Hardware development commonly splits early prototypes into two different jobs. A “works-like” prototype demonstrates the functionality a design is aiming for, whatever it looks like. A “looks-like” prototype, sometimes called an appearance model, is the opposite: it has no working internals, built only to let people see and hold the intended shape, size and finish. A campaign generally needs something closer to a third category, a prototype that both looks and works convincingly at once, which takes more time and budget to reach than either on its own.
- Works-like: proves the mechanism or electronics function, appearance secondary.
- Looks-like: proves the form, ergonomics and finish, with no working internals.
- Combined: proves both at once, and is correspondingly more expensive to build.
For a campaign video, a works-like prototype that performs the promised function on camera carries more weight with backers, and with the platform’s own rules, than a looks-like model ever will.
Free template: the 9 tests to run before you pay for tooling →Where a Campaign Prototype Sits in the Development Sequence
Engineering validation test (EVT), design validation test (DVT) and production validation test (PVT) describe a common development sequence, moving from an early engineering build toward a unit built on production tooling. In our experience at Inventornest, this sequence is useful for organizing development, but it should not imply that every project progresses without returning to earlier decisions. The important question is what each build must prove before proceeding, not whether the project follows a perfectly linear diagram.
For a crowdfunding campaign, the practical target is usually an EVT-level build that proves the core function works, since full DVT and PVT validation normally happens after the campaign funds the next stage of development, not before it.
Why “Build the Cheapest Prototype First” Is Incomplete Advice
“Build the cheapest prototype first” is common crowdfunding advice, and it is incomplete. A cheap demonstration can prove one function, but it may reveal little about whether the complete product will work. The objective should be to learn something useful before spending more, not simply to produce the cheapest physical object. For a campaign, that means the prototype has to prove the function backers are being asked to fund, not just exist cheaply.
What a Hobby Demo Proves vs an Integrated Prototype
In our experience at Inventornest, a low-cost demonstration can be built for around USD 5,000, which can be useful as a hobby-grade build or a limited proof of concept. An integrated engineering prototype suitable as a foundation for MVP validation, the kind closer to what a serious campaign needs behind its video, may cost around USD 20,000 or more, depending on scope. These examples illustrate different deliverables, not two prices for the same thing.
How Many Physical Units You Actually Need
Campaigns often assume they need a large batch of demo units; in practice, the number depends on complexity and budget. In our experience, a typical first-prototype build produces two physical units: one for the client and one retained internally. For complex or expensive products, the agreed delivery may be one unit; simpler, lower-cost products may allow five or six, depending on scope. The number of physical units is separate from the number of development or PCB iterations behind them, so quoting a unit count alone does not tell you how many design revisions led to it.
When a Minimum Demo Becomes a Liability
A minimum viable product becomes harmful when “minimum” is treated as license to skip testing the product’s essential function. Reducing optional features for a campaign demo is reasonable; ignoring a real performance limitation to make the video work is not. A physical MVP built for a campaign should generate real evidence that the core function works, not merely look complete enough for a few minutes of filming.
Two Documented Cases Where the Video Outran the Prototype
Kickstarter’s own investigative reporting into the Zano drone campaign, which raised millions of dollars, found that the creator later admitted the campaign video’s features “were not fully operational at the time it was shot,” calling the unit shown a “very early stage prototype.” The gap was specific and measurable: the prototype shown in the video weighed roughly 55 grams, and by the time the design reached pre-production its weight had grown to 70 grams as features and heavier materials were added, which cut real flight time to a couple of minutes against a promised 15. The company had already committed to around 20,000 units before a genuine pilot build existed.
A second case followed a similar pattern. Lily Robotics’ throw-and-shoot camera drone campaign drew a 2017 lawsuit from the San Francisco District Attorney’s office, which alleged that the promotional video used a more expensive, professionally operated camera drone rather than a working unit of the consumer product being sold. The product never shipped. Neither case is cited here to accuse the people involved beyond what each platform’s or prosecutor’s own public record states, and both illustrate the same underlying risk: a campaign video that shows more than the prototype can actually do becomes the standard backers hold the eventual product to.
Frequently asked questions
What is the difference between a looks-like and a works-like prototype?
A looks-like prototype, or appearance model, shows the final form without functioning. A works-like prototype performs the intended function, even if it does not look finished. A campaign video needs the works-like version far more than the looks-like one.
Do I need a full DVT or PVT unit before launching a campaign?
Usually not. An EVT-level build that proves the core function works is the realistic target for most campaigns; DVT and PVT validation typically happens after a campaign funds the next stage of development.
How much does a campaign-ready prototype cost?
It depends heavily on scope. A low-cost demonstration can run around USD 5,000 as a hobby-grade build, while an integrated engineering prototype suitable as a foundation for MVP validation can run USD 20,000 or more. These are different deliverables, not two prices for the same build.
How many prototype units do I actually need for a campaign?
Often just one or two. A typical first-prototype build produces two units, though complex or expensive products may need only one, and simpler products can allow several more, depending on scope and budget.
What happens if my campaign video shows more than my prototype can actually do?
Documented cases like the Zano drone and Lily Robotics campaigns show the real risk: backers and, in Lily’s case, a government prosecutor, treated the video as a representation of what the product could do, and the gap between the video and the real prototype became the center of the resulting dispute.
Is a minimum viable prototype the same as a cheap one?
No. A minimum viable prototype should still generate real evidence that the essential function works. Cutting corners on that testing to save money or time is a different thing from reasonably trimming optional features.
Where Inventornest Fits
Deciding which prototype fidelity a campaign actually needs, works-like versus looks-like, EVT versus a production-representative unit, is a feasibility question before it is a filming question, which is why we start there. For the broader engineering checklist behind a campaign-ready product, see hardware product ready for kickstarter, and for the general distinction between a prototype and an MVP outside the crowdfunding context, see prototype vs MVP. Our full walkthrough of the process is at how to crowdfund a hardware product, and our crowdfunding product development service page covers how we scope this work. You can book a consultation to work out what fidelity your own campaign prototype needs.
