5 reasons why hardware is so hard

Sad Panda

As a former entrepreneur, I struggled to sit still during my studies at INSEAD this year. To keep myself on my best behavior and not break things was to spend time helping nascent startups understand and overcome their most important problems.

One particular startup I worked with involved a lead scientist and his team at A*Star, a large government R&D center in Singapore, to find a way to commercialize their hardware technology that they had been developing over the past few years. At first glance, the project seemed straightforward: they had proprietary technology with a prototype working in the field that solved a potentially huge customer pain point. All they needed to do was convince their customers that they were the solution to a real problem.

Yet, upon further research, I discovered how tricky it would be. A simple Google search revealed not one, not two, but more than six different competitors in the same space with many other failed attempts since 2006. Despite the large number of players in the market, no one had found a way to convince their target customers the need for the technology.

After discussing with potential customers and end-users, I concluded that any hope to bring the product to market and scale would require a significant pivot for the client. Customers didn’t know why the current product was important and were unwilling to spend the money.

However, persuading the client to change the business model would be a tough task. To help the client understand, I gave him a short list explaining why having hardware technology alone was a difficult go-to-market strategy.

Upfront investment costs are much higher

Recent Kickstarter failures, including the infamous Zano drones and Coolest coolers, are prime examples of the tall hurdle of R&D investment that is required to make marketable products. Moreover, not all costs are monetary. My client invested over three years to develop their product, and they still don’t have revenue. Applying the lean startup method still takes time and money, often more than non-hardware startups.

Hard to deliver

Once you have a product, it’s hard to deliver. Distribution channels can make or break large, established consumer goods companies. Imagine how challenging it would be for a powerless startup. The growth of ecommerce platforms, such as Amazon, Ebay, and Etsy, and crowdfunding platforms, including Kickstarter and Indiegogo, have created alternate avenues, but acquiring new customers is still an uphill battle. If you’re a B2B product, it’s even more difficult.

Easy to copy

After you invest all that time and money into creating your product, it becomes easy to copy. Two CEOs of different tech companies, one in Singapore and one in the US, once told me separately that they do not apply for patents because patent filings expose their technology to competitors (and would-be competitors). A simple case study in the fashion industry includes United Colors of Bennetton which once held a huge cost advantage in their colorful sweater collection until competitors replicated their wool dyeing technology.

Hardware becomes a commodity

When it’s easy to copy, hardware becomes a commodity. A recent example is entrants of Chinese manufacturers in the PC industry. As a result, American tech companies, such as IBM, Dell, and HP, have shifted away from hardware manufacturing to service-based products. Samsung has already begun to shift investments from smart phones, no thanks to Xiaomi and Huawei, towards biotech innovation.

Preventing commoditization is too costly

Maintaining customer relationships with hardware requires large customer service teams (B2C) and account managers (B2B). Conversely, CRM is much easier when the product is a software, a service, or a platform since its naturally embedded into the product. Two Asian examples include OYO Rooms and GrabTaxi: people use these services for the quality assurance that each startup offers.

It was enough to convince the client, because he asked me to begin searching for new business models. Fortunately, I had already started working on something different, and I told him I’d be back in a week with strategies on how to win with hardware.