10 tips for de-risking hardware products

How to test and evaluate demand for hardware products before you crank up the factory

Manufacturing real-life, tangible objects that you can touch is often a lot riskier than developing software. Once you’ve created 10,000 thingamajigs, it’s far harder to make changes to them than in the software world, where you can push an update if you want to tweak something.

In the world of manufacturing, then, the question is: How can I make sure that I’m building the right thing for the right audience?

Last week, when I wrote about Prelaunch.com’s $1.5 million fundraise, I asked company founder Narek Vardanyan what he thinks are the biggest pitfalls in hardware development.

Measuring the right users

To truly understand what your customers want, Vardanyan recommends studying what your potential customers actually do, not what they say they are going to do.

In an ideal world, that means getting them to buy or at least put down a deposit, for your product. Real buying intent is more valuable than someone simply saying, “Yes, I would buy this thing.”

“You need to make decisions based on people’s actual behavior. You need to make sure that the data you’re tracking is coming from the right types of people,” Vardanyan said. “Working with people who are putting down money acts like a filter: You keep only the people who really want to risk their money. In other words, your potential customers.”

Measuring basic interest

Once you’ve selected the right subset of customers, the people who want to buy your product, the next step is to measure the right metrics.

“Sometimes founders share their products on social media and will see clicks, engagement, etc. They’ll think they are doing great, but those metrics can lie,” Vardanyan said. “You may have an interesting concept, but your pricing may be wrong, for example. People may be interested, but they may not buy because it doesn’t work at the price point you’ve selected.”

The solution is to create a two-page campaign. The first page will showcase just the idea or the product’s concept at a high level. The trick here is to not go into too much detail and just show a rough outline.

At the end of the page, you can add a call to action: “Learn more” could work well as would “Find out the price,” “Join the waiting list” or “Reserve yours!”

If you get a positive response at this stage, meaning people are clicking on a call to action, you’ll know you have something that they might be interested in. Someone giving you their email address signals real curiosity: They are signing up to be kept in the loop.

Testing the pricing

Once the basic measurement of interest is in place, you can start experimenting with pricing. On a price-testing page, you can get really granular with the specifications of a product. To capture real buying intent, it’s helpful if you get people to put down some real money.

“You can say, for example, that the price of this product is $100. You give the customers an offer: If you put down $10 today, you get a $20 discount when the product is released,” Vardanyan told me. “This measures how many people are ready to shift from interest to real buying intent and helps you compare your user engagement against benchmarks to see how well you are doing. If you’re doing terribly or very well, you will know; those extremes are easy to spot. In 95% of cases, it isn’t that clear-cut.”

If the product validation is going well but your would-be customers aren’t converting into actual customers, it means there’s a problem. If people aren’t willing to pay what you want to charge, it’s rather likely that the product will fail in the market.

A/B testing

There are many product validation techniques out there. Prelaunch.com specializes specifically in the kind of testing that hardware product designers want, but there are other options as well. For example, you can run manual tests or leverage the plethora of A/B testing platforms for marketing and sales.

“The trick is to tweak and test one parameter at a time in order to understand what works and what doesn’t. You can adjust the pricing, the messaging, the audience, the reservation amount,” Vardanyan said. “There are all these types of metrics, and we track 35 different data points on our platform: How much time they spent on the page, whether they have comments or questions, how engaged they are, how they scroll around on the page.”

Vardanyan added that in its three years, the team has found some interesting correlations between people’s interests, their behaviors and their buying intent. “We provide creators with benchmarks, and tell them whether a particular metric is doing well, or if we see behavior that indicates that something needs to be tweaked,” he said.

Start basic

If people arrive at your landing page and leave right away, it means that you have a fundamental problem: The person you hoped would be a customer is not interested in your value proposition.

That might be a design challenge or you may have created something that your target audience is fundamentally not interested in. Sometimes, you may find that European audiences respond very positively, while Americans may respond with a “meh.” That might mean that the product wouldn’t work in the U.S. or there might be something about the language or positioning that doesn’t appeal to the U.S. market.

It’s crucial to understand what the issue is, because if the marketing doesn’t work, there’s no point in putting the product in a warehouse in the States; nobody will buy it.

Prelaunch.com has an answer to this problem: It created a marketplace of sorts. “We have consumers and we have creators,” Vardanyan said. “For the creators, we provide a lot of data, benchmarks and guidance. We give you comparisons with similar products and show you how you are doing. That helps you understand whether you need to improve something. We have an AI that helps you analyze survey data, create new test pages and notice if something changes. Change the color of the product to blue or to red and see whether the performance spikes or drops, just to name an example.”

You don’t need a tool like Prelaunch to implement a lot of these approaches, however. Segmenting customers based on their geography, gender, interests, hobbies and other things can help guide the marketing messaging, the packaging or help you understand your target customers.

Running shadow campaigns

Another approach is to do a “fake door test” or a “shadow campaign,” where you create product pages without mentioning brands or with a fake brand and a fake product. You’ll be able to emulate a real e-commerce store experience and drive traffic to it.

If people try to buy these made-up products, you just give them a refund. The product doesn’t exist yet, but you can get really powerful data about what works and what doesn’t. And, of course, it’s very easy to change the price, look or specifications of a fake product.

Cancel if you must

It can be heartbreaking to realize that people don’t want the product you’ve been working on for so long. If that’s the case, having the conviction to cancel it could be the best path forward.

“Sometimes, creators are really committed to the product. They are willing to change everything about it to find product-market fit. They may change the name, materials, the colors, the price; it can become a totally new product,” Vardanyan said. “But in other cases, creators start with testing and use marketing tests to decide where to commit design resources. If they are doing a lot of testing and see that it isn’t working, it makes no sense to move forward with it.”

Trust the data

Creators might discover that the opportunity is far bigger than they originally had hoped, and that will create a different problem. If they were planning to manufacture 2,000 units, but the data supports that they can sell 200,000, the manufacturing costs are going to be far higher, which will send the risk shooting up.

“Sometimes, the entrepreneurs don’t believe their own concepts. They may have a product that was kind of a hobby or a side hustle, and they run a test. They might be amazed at how well it goes. In fact, we have a few success stories that started that way,” Vardanyan said.

“Sometimes it’s the pricing; the research indicates that you can charge more, but it’s hard to convince founders to increase their prices. In many cases, it turns out that increasing the price is the right thing to do and the market is ready to pay the higher amount. We’ve had a number of cases where the increased price launched into the market and benefited from the additional 50% additional margin. A lot of surprises can be hidden in the data.”

Trusting the research and the data you collect means that you have to take action. That can be hard if you have a preconceived idea of what a particular product is and who it is for.

Start testing sooner than you think

Another common pitfall founders are susceptible to is testing too late. You do need something to test, but that something can be pretty light; a mock-up or a market idea could be enough to get meaningful data back.

“We always recommend having a prototype or a very realistic mock-up. That way, you have enough information that people can relate to. That goes throughout the product development cycle: When testing features, it’s useful to test the right feature before you have made final decisions about the product,” Vardanyan said.

“Even after you have made all the other decisions, you will still want to test the colors and other final tweaks to the product. Another late-stage thing you can test is the packaging, to see how customers react. What you test depends on the product and who your customers are.”

Crowdfunding is often too late

I was curious whether a platform like Prelaunch would be similar to crowdfunding. But running a Kickstarter or Indiegogo campaign often focuses on sales, and usually most features of the product have already been locked in. If nobody backs the campaign, there’s no way to know why and whether a relatively simple adjustment to the product could have made a difference.

“Crowdfunding is more focused on fundraising, and you only get one shot. To do a Kickstarter, you need to have a product that’s more or less ready to manufacture. If it isn’t selling as well as you hoped, you still have to produce it and deliver it to your backers or cancel the whole project,” Vardanyan said.

“Product validation happens at a much earlier phase [with crowdfunding]. On the consumer side, crowdfunding is much riskier, too: Many times, when you back a product on Kickstarter, it gets delayed or people don’t deliver at all.”

Prelaunch solves that problem by holding preorders in escrow. If the product doesn’t launch, the people who reserved it get their money back.

“Don’t get me wrong: For 99% of crowdfunding campaigns, the intentions are good and everybody wants to deliver. But most of them are first-time creators,” Vardanyan said. “They don’t understand how to manufacture and they don’t have relationships with the manufacturers.”