Lumen – Hardware is Hard

Problem Discovery and the Ideation Process





Fast facts:

The first product launched in March of 2020 as part of a soft product launch, starting with China, the U.S., Israel, and Canada
Raised $1 million dollars led by Zim Integrated Shipping Services and the ZEP Fund Vintage 3
Officially launched in mid- September of 2020
Table of content

How It All Began

In the case of Lumen, the business began as a family affair. In 2014, Daniel Tal-Mor, Founder and CEO of Lumen, found himself mentally packing his belongings from his office at Conduit, the software company that acquired his first venture, Wibiya (a company started during his Zell year along with co-founder, Dror Ceder (Zell 7 – 2008) and their friend and CTO, Avi Smila). It took a total of three months for his leave to become a reality, and with that, an almost three year chapter in the software industry came to an end. While closing this door, Daniel had already opened a window, or rather, his wife did.

Now a little on Michal Tal Mor (Daniel Tal-Mor’s wife). Michal is a Ph.D graduate with a focus on cardiology and cardiac arrhythmia research. She is also an avid athlete, competing in marathons and ironman competitions. Around the time that Daniel was leaving Wibiya, Michal decided to return to school and study nutrition, along with her twin sister (and marathon training companion), Merav. These identical twin sisters shared twin interests.

Over a family dinner one night, Michal and Merav shared their frustrations about the nutrition industry. From the various dieting misconceptions to the lack of tools needed to easily measure bodily reactions to food, they discussed the many challenges that prevented most people from maintaining a healthy diet that is personalized to their needs. With so many different (and often conflicting) dieting opinions and food trends, one day eating meat can be a necessary component to a healthy diet, and the next day, strict plant-based meals are the “only” way to go.

Daniel was facing similar frustrations. During this period, he had been struggling with his own health dilemma. While building his former startup and becoming a new father, he had gained about 20 pounds. By the time his startup was acquired, Daniel began looking into all kinds of diets but couldn’t find one that resonated with him and that could eventually be integrated into his lifestyle over the long term. Nothing seemed sufficiently data-driven, at least not in a way that assured him that it’s what his body needed and could maintain over time. He kept asking himself, ‘where are the metrics?’

So as it turns out, besides misleading diet trends, it’s also difficult to measure the nutrients any given body needs over time, as it’s not a fixed regimen. But how can someone easily identify what they should be eating in order to give their body what it needs? It was over this discussion and a few glasses of wine that Michal and Merav decided to embark on solving this particular health gap. Naturally over time, Daniel’s interest in their nutrition riddle rose, and he brought the concept to Dror and Avi. The two had their own inner reasons for embarking on this journey. Dror is passionate about all things health-related, and more so, the future of the consumer health and wellness industry, while Avi, who struggled with weight issues as a kid, has a general knack for challenges that are outside of his scope of knowledge. Thus, the original Wibiya team jumped on board to join the twins on the Lumen journey.

Breaking Down the Problem

During their nutrition studies, Michal and Merav analyzed a metric called Respiratory Quotient (RQ), which identifies if a body is currently using fats or carbohydrates for fuel. Diving deeper, when any given body is using energy to function (whether being sluggish or active), it is either burning fats or carbohydrates. If someone is eating carbohydrates during every meal and using little energy, the body will be quick to burn the carbs as fuel instead of fats. However, when a body has used up its carbs, it uses fats for fuel instead, ultimately causing weight loss. That’s the logic behind shedding pounds by exercising and eating less calories than the body is burning. By measuring this metric alone, it would be possible to provide better nutritional guidance to each person on what he or she should be eating depending on an individual's body and health goals. 

However, at the time, in order to define someone's RQ, they needed to go to a clinic and breathe into a mask connected to a metabolic cart continuously for 30 minutes to analyze the Co2(O2) levels. Besides being time-consuming, this process was also expensive and ultimately not scalable. On the bright side, by identifying the main friction points, the team knew exactly what problems they needed to solve. The next steps would be determining how to solve them, and for whom.

And the Product Goes To...

As this was not his first time in the venture creation realm, this time around Daniel was determined not to build a new product blindly. This had been a big learning from the previous venture Joongel that morphed into Wibiya. [See Wibiya Case Study] As a team, they understood that there were several product categories they didn’t want to be associated with.

“Are we a weight loss product? Are we a fitness product? We asked ourselves, ‘who is the audience that will rally around such an offering?’,” Daniel considered. “In weight loss products, there are a lot of mechanics that we don't like – they’re always very restrictive and focused on a short period of time, say two to three months of extreme work, then usually most people revert back to their usual routines. The weight loss mechanics are not really what we aspire to do.”

Instead they imagined someone waking up in the morning, brushing their teeth, and measuring their RQ. With a product that is simple to use, someone could become more aware of their nutrition, allowing them to make more educated decisions around food, and create and follow a plan for each day. With that dream in mind, they agreed to create a B2C lifestyle product.

While it’s still a rather large market segment, Daniel and his team gave themselves focus while not pigeonholing themselves so early in the product creation process. “Even if we'll sell it to doctors, ultimately it's still a consumer product. Figuring out that ritual of how someone gains insight into their body and makes decisions based on that insight, that was on us,” Daniel said. It wasn’t until months later, when Daniel tested out his first beta release, that he segmented his target audience. Lumen’s first obvious audience (about 50% of its customers) is the weight loss audience. This is defined as people who experience hardship losing weight and are in the market for wellness and weight loss plans

and products that can aid them in their mission to shed pounds. Within this 50%, Lumen users are segmented further as 33% obese, 33% overweight, and 33% healthy according to their respective body mass index (BMI). The other audience type accounting for the remaining 50% of Lumen customers are those looking to be informed about their nutrition in order to make wiser decisions, but not for any specific weight loss purpose. In contrast, this user group is not seeking diet products or a strict lifestyle of counting calories. Instead, they believe in science and want to learn about their body more, so they can be empowered to make better decisions about their nutrition. Up until this point, their vision was enough to hit the ground running.

Go with the Pros

Compared to software, building a hardware product is expensive. Every change or addition is a deep dive into the money pocket, a move most startups want to avoid. Daniel thought about the build process carefully, understanding that unlike his process with Wibiya, he wanted to deliver enough value that a consumer would spend money to buy his product, and to do this, he needed experts. With that in mind, in 2014, Daniel and his team decided to hire Frog Design, an industrial design consultancy. They flew to Milan to present to Frog Design their concept of a device that can measure a person’s metabolism. And so began a series of workshops where, together, they defined the device’s design, use, and narrative.

“If you do go and create a crazy technology that will take time and investment, it's worthwhile to make sure that people want it, and that you have the business story and the usage story already in place,” Daniel said. From understanding the market to creating a brand, Frog Design helped Daniel and his team come up with their business story. This was a lesson learned from the first iteration with Joongel, and then Wibiya, where the trial and error was done without any professional design or market guidance. At the end of an iterative process, the product and company took the name Lumen.

Instead of jumping straight into production mode, they wanted to test their assumptions in the real world. They needed a way that didn’t require any technical development, was accessible to a large group of people, and was most of all, cheap. Their solution was a phone box.

Fake It Til You Make It

Daniel and the team figured they didn’t need a completed beta, or really any working prototype to gauge potential customer interest. With that in mind, the team designed a few phone boxes promoting the Lumen idea and put a small flashlight in each. They also built an online form that asked what a customer ate in the last 24 hours. Depending on their answers, it would show three possible outcomes: their body will likely be burning fat, carbohydrates, or half and half. The flashlight didn’t actually connect to the form, but they were prepared to make it seem like the light could capture a customer's RQ. Once they had the not-quite- real testing material ready, the team created a small stand at a local gym and approached people to pitch the “Lumen” device, with the goal of finding out if those people would be willing to buy it. Of course the team had no clue how they would eventually get a device to properly measure someone’s RQ, but the point was, would they even care to do so? And importantly, would they pay for it?

As it turns out, people were amazed, and following their experience quickly offered their credit cards in order to buy the product. Unable to sell them any product, the team took their emails to let them know when the final product would be ready. “That really was a proof of concept. Those small experiments are extremely meaningful to hold on to during this long road, to know that this is worth it, and that if we build it, they will actually buy it,” Daniel said.

Hardware Mode

“Building hardware is a nightmare.” Daniel can speak on this topic for minutes on end. It took five years and three prototypes until Lumen had a working and viable product. For the first year and a half, Daniel and team worked off technology assumptions that didn’t pan out. His mind was set on reaching the same quality RQ capacity as the gold standard machines that were used in clinics, but without having to breathe into a mask connected to a machine for 30 minutes. In order to accomplish this, they needed a cheap CO2O2 sensor, which they assumed would be easy to source and then merge into a small device, add some machine learning, and poof, this invention would provide a RQ result with just a few breaths. Well, that elegant plan crumbled.

At the time, there was no cheap CO2O2 sensor that functioned to Lumen’s requirements. All of the off-the- shelf sensors were either extremely expensive, big, or consumed too much energy. As for machine learning, it repeatedly failed to shorten the time to measure RQ compared to the clinical machines, and ultimately proved irrelevant to the Lumen cause.

Daniel and his team were not ready to give up just yet. After a year and a half of researching and testing sensors, Michal and Merav made a breakthrough finding: a method that with a single breath using a different breathing maneuver (inhaling, holding that breath in your lungs, and exhaling), Lumen could create a correlating metric to that of the gold standard machines. With that finding, they would be able to solve the speed issue, but what about finding the right sensors? Lumen started prototyping with an expensive sensor, but Daniel knew he had to get the cost down in order to release an affordable lifestyle product. Within such a limited market, they decided the only realistic solution would be to build the right sensor themselves. This detour, though, had considerable price and time implications.

Daniel traveled far and wide to meet every sensor manufacturer in the world, from Sweden to Scotland, Germany, the United States, and China. Their determination eventually led them to sign a joint development agreement with a U.S. manufacturer that specializes in CO2 sensors for environmental purposes and the automobile industry. It seemd like an odd choice, considering Lumen required medical specs for its sensors (with the ability to deal with human breath and humidity), but there was not much of a paved path to follow. The Lumen team was inventing new technology, and they went with the manufacturer they thought was the best fit to get the job done.

“People say hardware development is hard, but it was the development of the sensor itself, of the underlying technology, that was our main barrier,” Daniel said. After two and a half years and numerous trials and errors, Lumen got its sensor.

Within those two plus years, Daniel and his team were busy prototyping and learning from potential customers. By the time the sensor was finalized, Lumen was on its third prototype and ready to go to market. “It's hard for me to say we could have learned the market sooner. A lot of the things you can't really learn until you give the real experience to the user,” Daniel said.

Hindsight Is 20/20 (Even in 2020)

Daniel is the sort of founder who believes that dreams and targets are not mutually exclusive. He’s practical but leaves room for imagination. While embracing the milestones Lumen has achieved, Daniel acknowledges that had he known how bumpy and risky the road would be, he might not have done it. There were no guarantees that they would succeed in making a functional product to detect RQ levels, nothing besides their passion and resilience. The truth is, inventing new hardware technology is really, really hard, and sometimes inventors and founders alike come up short. In the case of Lumen, it just took longer than expected – six years longer to be precise. The saying that hardware is hard... holds true in this case. What fueled Daniel and his team was their unswerving eye on the value a tool like Lumen could offer. Their goal was never to be a device for nutritionists or for the science community, but rather, an invaluable tool for the everyday person struggling to maintain a healthier lifestyle. Even though the decision to focus their go-to-market (GTM) strategy on the general public, and not on healthcare practitioners, was not an easy one, it was the universality of the benefit and the audacity of the goal that kept the team going.

“There aren’t many things that can keep you in the desert of entrepreneurial technology building and looking at the horizon, and saying it’s worthwhile,” Daniel said. “But you need to find those levers, whatever they are, whether it’s the strength of a team, or the uniqueness of a product you're building.” Daniel recognizes he had both a strong team with a can-do culture and a unique product with an aspirational value proposition to make people’s lives better, and that’s what got him through those six years in his entrepreneurial quest.

Discussion Questions

  1. What makes building a hardware startup so different from building a software startup?
  2. What lessons learned from their previous startup did Daniel and Dror put into action the second time around. Did it make it any easier that they were second-time entrepreneurs?
  3. Was their phone box “concept test” a good way to validate the need?
    What other ways could they have tested the concept and assumptions before actually building anything?
  4. When the team could not find a suitable sensor, what did they do? What were the pros and cons of this decision?
  5. What kind of customer did Lumen choose to target, and how is their market segmented? Why did they choose this specific approach?
    Do you think it was the right one?