Over the years lean testing methodologies have become an entrepreneurial staple for software startups. But some, including entrepreneur, investor, and best-selling author, Ben Horowitz, have argued that the lean methodology doesn’t always work. In particular, hardware startups face unique challenges. Because hardware products contain multiple parts that need to function together, all parts must be developed and tested simultaneously, which prevents low-fi testing. Instead of the quick iterative testing and sales cycles that software startups can follow, a hardware startup that sells to enterprise customers faces cycles that last many months or even years. Engineer and entrepreneur Joris Poort discovered this during the decade in which he’s built Rescale—the world’s leading scientific and engineering simulation and high-performance computing (HPC). Under Poort’s leadership as co-founder and CEO, Rescale has grown to 150 employees and has five offices worldwide. In 2019, the company was valued at half a billion dollars. Along his ten-year journey, Poort acquired valuable insights about starting and scaling a deeply technical hardware startup that sells to enterprise clients. In a conversation with Shikhar Ghosh, he offers tips on market testing and sales cycles for hardware startups working with enterprise customers.
Tips for Hardware Startups Working with Enterprise Customers
- Lean Startup Methods Don’t Easily Apply to Market Testing for Hardware Products
- When Market Testing, Consider Co-Developing Products with Your Customer
- Understand the Long Cycles and Expectations for Enterprise Sales
- Create Key Metrics to Evaluate Progress during Long Sales Cycles
- Set the Right Sales Pursuits for Your Startup’s Stage
Dubbed a “silent genius” by the press, Joris Poort co-founded Rescale with Adam McKenzie, who became CTO, in 2011. The company produces software platforms and hardware infrastructure that aggregate and optimize computing power. It enables large enterprises in science, engineering, and business to solve their biggest problems faster. Building on Poort’s and McKenzie’s background as engineers at Boeing, Rescale takes “all the different applications that are typically in an R&D environment, like designing an aircraft or a car, and help run them more efficiently through optimizing” server configurations, operating systems, and more “in a cloud SaaS environment.” In 2019, with the release of ScaleX and ScaleX Government, Rescale became the first company to deliver a completely cloud-based HPC platform for the public sector which allows federal engineers, scientists, and researchers to convert dated, fixed IT infrastructure to on-demand cloud architectures that comply with requirements.
Lean Startup Methods Don’t Easily Apply to Market Testing for Hardware Products
Creating low-fi prototypes to conduct widespread testing with customers, then iterating on feedback—as you would following a lean approach—isn’t a viable option for technical hardware products. “It’s very hard to build a large, complex system through quick, minimal iterations” with individual parts of the whole a product, Poort notes. He observes, most founders of tech hardware companies have experienced the problem themselves, often at a deep level. For instance, as an engineer at Boeing, he learned not only “how an aircraft is developed” but also the complex systems already in place there. Having an intimate understanding of the space can give hardware founders invaluable insights into “similar types of customers and understanding what type of challenges they have.”
Understanding what type of challenges they have, and developing technology yourself, you have to have a good vision and a long-term alignment to solving those pain points.
A Different Approach to Market Testing
When operating a technical startup, Poort recommends conducting market testing by working with B2B customers to co-develop the product. When developing technology, especially for enterprise clients, Poort stresses, “you have to have a good vision and a long-term alignment to solving those pain points.” He notes the difficulty of adding incremental value to a big complex system. “You can’t just insert a widget and make a meaningful impact.”
Instead, he recommends a co-development approach that involves working with a B2B customer directly. If possible, create a professional services arrangement—they can begin testing your product as you develop it and you can, in exchange, receive some payment and feedback. That way, Poort explains, “You keep iterating with that customer.”
You want to have a strong vision for what problems you’re trying to solve, at scale, for the market. Having a couple of customers who help both finance and help inform how you should be solving these problems can be really helpful.
Co-Develop Products with Enterprise Customers
But co-developing a product poses the risk that you may end up building a product that’s customized to meet the needs of one customer. To offset that risk, Poort stresses, “you do want to have a strong vision for what problems you’re trying to solve, at scale, for the market. But having a couple of customers who help both finance and help inform how you should be solving these problems can be really helpful early on.”
Two companies—Blue Origin and SpaceX—gave Rescale the opportunity to solve some of their problems. “They were able to iterate with us quickly,” Poort recalls. “There were not many other examples like that in aerospace, at that time.” Larger companies who have processes or bureaucracies in place have more difficulty with a co-development arrangement. “The key thing is building enough trust with the customer that you can really co-developed something, as opposed to saying, ‘Here’s a final product. Do you want to buy my widget or not?'”
The key thing is building enough trust with the customer, they can really co-development something, as opposed to saying, ‘Here’s a final product.
Developing a Product Together Can Build Trust
One of the best ways to build trust, Poort learned, is by doing some custom development or expressing a willingness to “own and solve their specific problem.” Often, you can recoup the time and effort required for a customized effort by using it to create and commercialize a product that fulfills the needs of a broader customer base.
Get the trust from the customer by being willing to do some custom development—to a certain extent, to own and solve their specific problem. This is paid back through the ability then commercialize something that’s a broader product applicable for many customers.
Understand the Long Cycles and Expectations for Enterprise Sales
Development for hardware products, especially if you’re building to solve a particular problem for an enterprise customer, entails long cycles. You run the risk of investing time, money, and effort before you’re certain if you’ll secure a sale at the end of the process. Poort acknowledges, “The feedback loop is quite long.” He smiles, “procurement at a large aerospace company is where really good startups go to die.” Some enterprise customers “can take 18 to 24 months from when they decide they want to buy something to when they actually buy it.”
Enterprise sales often come in “large deal sizes that are in the seven, eight, nine figures.” But Poort learned, “You only want to start entering that process when you’re really ready to do that.” Typically, “you only get to that deal if you can survive that process” and your product tests at an enterprise-grade which means meeting the service level agreements (SLA) that an enterprise customer would expect, such as security protocols, privacy, and how you handle data, and support, for starters.
Make Sure Your Startup Is Equipped to Deliver
“You can look at that as an advantage or a massive disadvantage” Poort notes. Rescale has met the highest industry compliance standards in medical devices, aerospace, and defense. The security certifications can take two or three years to obtain. So you have to be willing to invest the time in that, to have a shot at that type of market. “If your software is instrumental and mission-critical, it’s great because customers will pay you a lot to have that perform well.”
At the same time, many enterprise SLA requirements come with a set of responsibilities, Poort notes, that present challenges for smaller companies. For instance, can you provide 24/7 staff support? Are you always on call? He advises, “You have to make the right investments that put you on the right path. Then, ideally, start generating some revenue with your customers before you’re really ready to be the entire system for that company.”
That is the chicken and egg thing you have to solve. How can you make those right investments without really being ready and being able to take the revenue from those customers in the short-term?
Create Key Metrics to Evaluate Progress during Long Cycles
If you’re in the early stages and cash-constrained, how can you measure progress during such a long sales cycle? Poort admits that assessing progress is challenging with a small data set, especially if you have only one salesperson. He encourages founders to “have at least two salespeople, so you’re not 100% dependent on one person doing the selling.” That way, ” you solve for the variable of an underperforming salesperson.”
Larger enterprise sales, Poort discovered, have longer sales cycles which make the sales process more challenging initially. “If your deal size is two, three million, it is really hard because you’re always going to be in these really long cycles.” An enterprise sales cycle can take three to nine months, from signing an NDA to through production. He identifies fifteen steps in the buying cycle. These steps involve micro-commitments made by the seller to the customer, but also by the customer to the seller.
In a long sales process, Poort explains, it’s essential to maintain any promises you make to the customer. But “The best sales process tests the customer every step of the way” to ascertain, “Are they really serious about doing this thing?” Most hardware companies will follow unique sales processes. This can “be a huge competitive advantage for a company if you understand how your customer wants to buy your product” Poort notes. “The seller learns to figure out the process and determine what steps will work to sell the product. Your startups need to “figure out in that maze of all these different routes you can take, which ones work and which ones don’t.”
Poort emphasizes that the discoveries made during the sales process become part of your customer relationship management and should be captured in a CRM system. That way, as you grow, you “can execute systematically with every new seller that gets plugged into your sales team. They can get trained on that process, they can execute, and you continue innovating and improving that.” Often, Poort observes, salespeople develop “happy ears. They only hear the things that they want to hear.” That’s why it’s imperative to test your customers’ intention to purchase at “every step of the way” and he encourages, “try to disprove that” intention.
Often salespeople have “happy ears.” They only hear the things that they want to hear. But it’s really important to be testing every step of the way. Is the customer really going to go buy this thing? And try to disprove that.
Paid pilots can play a big role, Poort suggests. “If a customer’s committed to paying for the pilot, they’ve invested something. It can be any amount of money. But at least they’ve had to figure out how to write a check to go engage on that.”
If a customer’s committed to paying for the pilot, they’ve invested something. It can be any amount of money. But at least they’ve had to figure out how to write a check to go engage on that.
Set the Right Sales Pursuits for Your Stage
Most companies have multiple stakeholders and decision-makers. Identifying key decision-makers is one of the most important and valuable parts of the sales process. “A lot of people can say ‘no,'” Poort notes, but only a few people are typically authorized to say yes. Additionally, during the sales process, you should figure out the budget and timeline. Many different sales methodologies exist. Poort recommends “trying to disprove that you’re going to make that sale” because “you can waste a huge amount of time on a customer hoping.”
It may be tempting to pursue a sale because you foresee a big deal at the end if you are successful. But make sure that your startup can survive over that timeline. For instance, Poort notes, “a $20 million deal” may take you “four years in the sales process.” Ask yourself if that timeline makes sense at your stage. Do you have the staff and runway you need to survive to that point? If you don’t “then it’s not a very useful sales pursuit.”
If you have a $20 million deal that’s going to take you four years in the sales process, but you’re not sure if your company can really survive to that point, that’s not a very useful sales pursuit.
Poort stresses the importance of pursuing the right sales targets for your company’s stage. Ideally, each sales cycle involves a payback period. Each seller has a quota, “so you’re trying to measure what that sales productivity is, as they’re ramping.” If you don’t start seeing that productivity, Poort notes, “then, you want to make some changes.”
Interested in learning more about the sales process in general? In What Founders Need to Understand about the Sales Process sales guru Lou Shipley reviews fundamentals all founder-CEOs should understand about the sales process. For more insights on sales, including launching your product in a foreign territory and culture, see our full-length interview with Lou Shipley accessible here. Want to learn more about the unique challenges that founders of hardware startups face? See the additional resources below.
Bringing a Hardware Product to Market Navigating the Wild Ride from Concept to Mass Production by Elaine Chen provides an invaluable overview of the unique challenges hardware founders face when moving from idea to mass production. Chen, who is a Senior Lecturer and Entrepreneur-in-Residence at the Martin Trust Center for MIT Entrepreneurship, wrote about the book to help founders with an engineering background navigate “hurdles before getting to a saleable minimum viable product” when producing hardware products for the first time.
The Case for the Fat Startup by best-selling author and co-founder of Andreessen Horowitz, Ben Horowitz, examines the limitations of lean startup methodology as a one-size-fits-all model. He notes, “by making “running lean” an end, you may lose your opportunity to win the market, either because you fail to fund the R&D necessary to find product/market fit or you let a competitor out-execute you in taking the market. Sometimes running fat is the right thing to do.”
In “Repeat after me: Hardware is NOT software,” Elaine Chen, Senior Lecturer and Entrepreneur-in-Residence at the Martin Trust Center for MIT Entrepreneurship, cautions that, unlike software, “Creating and testing part prototypes along the way provides only partial benefit, as each individual part could be fine, yet the entire assembly might not work when it’s put together” so, essentially, “all of the parts have to be released at once, at the end of the design period.”
In “The New Path To Starting A Hardware Startup: Lessons Learned From A Crowdfunding Approach” reviews how launching a crowdfunding campaign can help hardware startups by providing capital to get underway. However, it reviews the risks or downsides of using crowdfunding—”Compared to creative project endeavors or software startups, if you’re a hardware company, you have to do so much more to deliver a product: you have to build a hardware prototype, develop the firmware and often create an app which works with your product.”
Bolt is a pre-seed and seed-stage venture firm focused on investing at the intersection of the digital and physical worlds. Based in San Franciso and Boston, Bolt invests “early and with conviction.” The firm has invested in over 70 companies and led co-led the first round of financing in over 90% of its portfolio companies.
Hardware.co is a Berlin-based accelerator and global community for entrepreneurs dedicated to the creation of innovative hardware products and companies. Started in Berlin in 2013, the program supports every stage of hardware development, from ideation, through prototyping and investments. It offers multiple channels including an accelerator, lab, meetups, and online platform.