ISSUE 011
Money Up Front
Nine entries of this pillar have now ended in the same place: the only check worth much is somebody actually paying. So this one is about the moment somebody actually pays before you have made the thing. Deposits, pre-orders, a waitlist with a card on file. It is the best signal available to a small business and the research on it is less flattering than I expected, mostly because the money arriving turns out to be the easy half.
An idea worth thinking about
The money arrived and more than a third of them still went over budget
Two researchers, in a paper in Wharton's working paper series, surveyed the people who had run large design, technology and video game crowdfunding campaigns before mid-2012, and went back one to four years later to find out what had become of them. The good news is genuinely good. Over 90 percent of the successful projects were still ongoing ventures. Thirty-two percent reported yearly revenue above $100,000 since their campaign. On average they had added 2.2 employees.
That is the strongest evidence I found anywhere that money up front means something. People who take payment before the product exists are not generally running a fantasy; most of them are still trading years later and a third of them built something with real revenue. If you have been wondering whether to ask for a deposit, this is the paragraph in your favour.
Now the number in the same study that nobody quotes: 37 percent of the projects went over budget. More than a third of the ventures that successfully proved demand, collected the cash and kept trading still got the cost of delivering it wrong. The pre-order answered the question “does anybody want this” and told them nothing whatsoever about “can we make it for what we charged”.
Two things to hold against this. It is a working paper rather than a peer-reviewed article, and it is from January 2014, which makes it twelve years old and describes campaigns run before mid-2012. And it surveyed people who had run large campaigns on one platform, which is a particular kind of person. What survives both caveats is the shape of the finding, and the shape is this: demand and feasibility are two separate questions, and the deposit only answers the first one.
A tool I disappeared into
The thirty-day clock that starts when you take the money
If you take payment for goods ordered by post, over the internet or by telephone, there is a federal rule about what happens next and you can read the whole of it in about fifteen minutes. It is 16 CFR Part 435, and the current text sits on the eCFR website, free, searchable, with every section numbered.
The core of it is section 435.2(a)(1)(ii). If you did not clearly state a shipping time, you must ship within thirty days of receiving a properly completed order — which section 435.1(c) defines as the point where you have the payment and all the information you need to process and ship. You are also not supposed to solicit the order in the first place unless you have a reasonable basis to expect you can ship in the time you stated, and under 435.2(a)(4), if you have no records showing how your shipping system works, you can be presumed not to have had one.
What happens when you slip is the useful part, because it is a procedure rather than a penalty. Under 435.2(b)(1) you must offer the buyer, clearly and without them having to ask, the choice between consenting to the delay and cancelling. If you miss the revised date too, you offer the choice again. If you cancel, 435.1(b) generally requires the refund to go out within seven working days.
Two limits, and they matter before anyone quotes this at a customer. This is not legal advice and I am not a lawyer; it is a trade rule I read. And it is narrower than it sounds — it covers merchandise ordered by mail, internet or telephone, with exclusions at 435.3(a) that include collect-on-delivery orders, seeds and growing plants, and magazine-style subscriptions after the first shipment. It is not a general law of deposits and it does not cover a deposit on a service. Still, if you are about to open pre-orders on a physical product, this is the document that tells you what you have just promised.
Learn something
No more than 32 percent delivered on time, and the paper cannot quite agree with itself
Two researchers at the Technical University of Denmark went through 144 successfully funded Kickstarter hardware campaigns — physical consumer products, the kind with a factory at the end of them — and read every project update and backer comment. Thirty were used to build their model of what goes wrong; 114 were analysed in detail. The paper is open access under a Creative Commons licence, published in Design Science in 2018.
The abstract's finding is that no more than 32 percent of the campaigns managed to deliver the crowdfunded products on time. Of the 114, they report 36 campaigns at 3 to 12 months late and 50 that exceeded a full year of delay. Here is the inconsistency, and I would rather flag it than smooth it: the results text says 28 campaigns delivered on time and calls that 32 percent, while Table 1 lists 36. Thirty-six of 114 is 32 percent; 28 is not. The headline figure matches the table, the sentence does not, and I cannot tell you which count is right.
The part I had not expected is what lateness travelled with. Of the 69 products that did get delivered, 17 — a quarter — were reported to lack one or more of the promised features, and the delayed ones were significantly more likely to be in that group. Usability complaints were also significantly more common for delayed products. The authors attribute most of it to product development rather than to fraud or laziness: manufacturing quality, component sourcing, and designing something that turned out to be hard to actually manufacture.
Their own caveats are stated plainly and I will repeat them. The project updates and descriptions are self-reported by the campaign initiators and therefore carry subjective bias. Two coders worked collaboratively, which they note did not make it possible to estimate a measure of inter-rater reliability. And it is one platform, which they call an obvious limitation for generalisability. Read it anyway, for the failure mode table. It is the most specific list I have seen of the things that go wrong between taking the money and shipping the box.
What we’re seeing
Around nine percent of funded projects failed to deliver. Possible range: five to fourteen.
The most-cited number on this subject comes from a 2015 working paper by Ethan Mollick at Wharton, built on a survey of 47,188 Kickstarter backers. His sentence: among funded projects, a failure to deliver seems relatively rare, accounting for around 9% of all projects, with a possible range of 5% to 14%.
The range is why I put it in the headline. Nine percent gets quoted everywhere as a fact; five to fourteen is nearly a threefold spread, and it is the author's own estimate of how much he does not know. It is also a working paper rather than a peer-reviewed article. Take the 9 percent as a central estimate with wide error bars and you are reading it the way it was written.
Two more findings from the same abstract. Small projects, and to a lesser extent very large ones, were more likely to fail to deliver — so the comforting assumption that a modest ask is a safe ask does not hold. And the demographics of the project creators, including gender, education level and family status, did not significantly affect the chances. Whatever separates the deliverers from the non-deliverers, it is not visible on the person.
Now the limit that governs this entire entry. Three of the four sources here are about Kickstarter, because that is where the data is — a platform with public campaigns, reachable backers and researchers who asked. A reward crowdfunding campaign is not the same thing as taking a 30 percent deposit on a heat pump installation, and I am not going to pretend the numbers transfer. What transfers is the structure: in all of these cases money arrived before the thing existed, and in a measurable share of them the thing never did.
Which lands this pillar somewhere slightly awkward. Across nine entries the conclusion has been that evidence is worth what it has been checked against, and that the only clean check is somebody paying. A deposit is that check. It is incentive-aligned in a way no survey question can be, it costs you nothing to run, and it produces an answer that a stated-intent study would take 300 respondents to approximate badly. And it still only answers half the question. The customers were real. Around nine percent of the projects were never delivered.
So here is the version I would actually run, and the last step is the one people skip. Take deposits, and make them real — a refundable reservation with no card charged is a waitlist wearing a costume, and I would treat the conversion from that to a charged payment as the only number that tells you which one you have. Put the delivery date in writing. Price the deposit at something that stings slightly. Then, three months later, count: of the people who put money down, how many took delivery and paid the balance, and how many asked for it back. That second number is your real validation, and it is the one nobody writes down.
One question
Have you ever asked a customer to pay before you had the thing?
Deposit, pre-order, retainer, anything. What did they say — and if you have never asked, what stops you?
More in this pillar as it's written.
— Shana
Sources link to the original publisher, never to a summary.