Knowledge is power in deep tech, so start

Asking

Better

Questions

“Uproot your questions from their ground and the dangling roots will be seen. More questions!”

― Frank Herbert, Chapterhouse: Dune

Welcome! If you are a deep tech entrepreneur, this resource is for you.

I do not claim to have all the answers, but I have learned that the path to getting those answers necessarily involves knowing which questions to ask.

This evergreen website contains key questions any deep tech entrepreneur should be asking themselves.  Each question is something I have grappled with in my own entrepreneurial journey. If you expand any one of them, you will see a mini case study or example from my experience with additional nuances to consider.

Not all of them will apply to you, so use your own filter to make the most out of the resource. Many of these questions don’t have a definitive answer; they are meant to be asked over and over to guide your learning in an ever-evolving world.

Enjoy,

Ioana Knopf 

Do we really understand our customer’s reality?

We can never know too much about our customers or our customer’s customers, and – in truth – we rarely know enough. Moreover, the customer’s reality is a dynamic thing. Priorities, expectations, and behaviors change, often unexpectedly or in unexpected ways. These sorts of questions aren’t just the first to ask but also the most important to ask over and over again.

  • If solving what you propose to solve isn’t an immediate or near-term industry priority, it doesn’t matter how novel or potentially valuable your technology might be – you will face significant headwinds.

    Trade shows can be a great way to get a pulse of the industry and talk to many relevant stakeholders in one place. I used this strategy when investigating various application areas for my carbon-negative minerals, one of which was using them as plastics additives to lower the overall carbon footprint of plastic products. I learned that despite most companies having net-zero targets, reducing greenhouse gas (GHG) emissions was not a high priority. The plastic waste pollution issue is much more visible and immediate, so plastic recycling remains the major sustainability initiative across the board. And while using recycled content also lowers GHG footprint, offering a solution that simply lowered emissions without recycling missed the mark. Additionally, some folks were actively averse to using my additive if it would impact the recyclability of the final plastic product. Gale-force headwinds!

  • Many entrepreneurs build technologies that aim to replace or alter established industrial processes or products. In order to gain widespread adoption, we must admit that our true competition is the status quo. If we want to have a chance at replacing it, we need to understand WHY something is the default option. 

    There are many possibilities to consider beyond cost and performance, including supply-chain constraints, brand allegiance, regulatory risk, integration with a different process or product, geographic proximity, and more. Moreover, the “WHY” might differ between companies or even within the same company. Getting to the real “WHY” – the one that really matters – requires us to dig deep.

    I remember walking up to a trade-show booth where I saw a giant banner advertising the use of a material to lower GHG emissions and make a more sustainable product. I was curious to learn about how this company was able to find climate-conscious customers, but when I asked the salesperson why people buy the material, he answered, “cost,” without skipping a beat. The product was cheap, and that’s why people bought it. The material wasn’t really what I thought it would be, and there were no climate-conscious customers anywhere in sight.

  • We’ve all made the classic competitive-landscape slide to demonstrate how our startup is differentiated from all the others out there trying to tackle the same problem. That doesn’t mean we’ve considered all the alternatives.

    Aside from competing with the incumbent solution, we may be competing with the customer’s preference to do nothing at all. A customer may have a problem they’re not motivated to address. They may prefer to just live with it. 

    Furthermore, the customer may be able to eliminate the problem you’re proposing to solve by making different choices upstream. One of the big challenges in the plastic recycling space is that we cannot mechanically recycle food-contaminated flexible packaging. One way to tackle this problem would be to develop a chemical method for recycling that plastic. As I dug deeper into material sourcing, however, I realized that many food chains were moving away from plastic packaging and opting for paper or other compostable options. They didn’t need to adopt a new technology. They were solving their problem by making different material choices upstream.

  • This is a common trap for hard-tech entrepreneurs. We are motivated by the societal impact that our technologies can manifest, but we quickly run into collective-action problems if our solutions target issues like reducing GHG emissions. My personal experience was that everyone wanted to decarbonize but few were willing to pay for it. Instead of asking “what problem are we solving for our customers?” we should be asking “what problem are we solving for our customers that they are willing to pay for?”

    It’s imperative to distinguish between our own reasons for building a product and the customer’s potential motivations for adopting it. In an ideal world, our motivations are aligned, but that is rarely the case. We may be providing a cost savings, or a performance enhancement, or supply-chain reliability, or a liability reduction. We need to identify the value proposition that matters to our customers and make it the core of our messaging. 

    In my own search for an alternative, non-climate-related value proposition, I came up empty. My material’s differentiator is that it’s carbon-negative. It performs similar to other minerals on the market; there was no other advantage I could offer to motivate my prospective customers.

  • The question of whether our customer will purchase our solution is inextricably linked to how they would purchase it. Does your business model fit into existing procurement processes? If not, you have to overcome significant barriers to adoption unrelated to your technology. 

    Relatedly, it’s imperative we understand the sales cycles of the industries we’re targeting. At one point, I was investigating markets suitable for precipitated calcium carbonate (PCC), a more premium version of the carbon-negative minerals I could produce. The paper industry is the largest market for PCC, taking advantage of its whiteness and brightness properties. What I discovered is that PCC plants are often colocated with paper mills, and they provide PCC based on 10-year supply contracts. This created a significant barrier to entry I couldn’t reasonably overcome.

Do we know our industry and its value chain end-to-end?

Developing a working understanding of the market or industry we operate in goes well beyond compound annual growth rates and most of the metrics that populate pitch decks. Understanding the value chain, end-to-end, is no small feat. If we are coming from outside and looking in, we can more easily see the forest for the trees, but if we don’t cultivate deep relationships with insiders, we overlook critical details. Understanding the value chain in its entirety requires us to cultivate the skill of zooming in and out.

  • It’s not enough to understand what problem we are trying to solve for our customers because we are often aiming to sell into a complex value chain with many different stakeholders.

    Your customer is not a monolith. To understand your value creation potential, you need to zoom into the nuances of their business. Consider which department(s) has the problem or which other stakeholders in the organization support or oppose the adoption of your solution. Who affects their decisions? Whose decisions do they, in turn, influence or dictate?

    Your customer’s reality extends far beyond their own business. It frequently changes in response to the needs and decisions of stakeholders both downstream and upstream from them in the value chain. Understanding the end-to-end value chain requires you to both zoom out and consider the priorities of the industry at large and zoom in on specific points of the value chain to identify how these priorities manifest and why specific decisions are being made. 

    To further complicate matters, it is not uncommon for the people in the value chain who would purchase your solution to be different from those who set the specifications for the product. Moreover, the champion for adoption of your solution may be someone else entirely. 

    In the early days of my customer discovery journey, I was investigating applications for my carbon-negative minerals in construction materials such as concrete. It was a critical revelation to learn that concrete manufacturers, my hypothesized customers, don’t have much, if any, flexibility in choosing what types of additives they can use. As it turned out, the material specifications for these projects were set much earlier by architects and structural engineers.

  • The materials needed to manufacture your product can become a bottleneck for your business in many different ways. 

    Cost is the first and most obvious thing to investigate, but it’s also something that can become quite nuanced. In my case, silicate minerals are often painted with a broad brush in the academic literature as being cheap and earth abundant, thus the perfect candidates for use in CO2 mineralization. In reality, the silicate mineral I was using proved to be neither cheap nor abundant. The internet was full of misleading data from market reports, so estimating costs in the minerals industry ended up being a significant challenge. It took me a while to realize the best approach to this challenge was to engage a highly experienced consultant. 

    Even for the same chemical composition, materials cost can vary widely depending on characteristics like purity grades or particle sizes. Using optimistic costs at scale in your techno-economic models could be counterproductive if those materials don’t have the right specifications for use in your process.

    Another thing to consider is whether your starting materials will truly be available at commercial scale. There are some very compelling maps from the USGS and similar sources which show the widespread worldwide deposits of minerals such as silicates, which could be used in CO2 mineralization; however, there are just two active mines for calcium silicate (wollastonite) in the US. The lack of current supply severely impacted the potential reach of my technology. From a cost and climate standpoint, it makes little sense to transport these types of minerals across large distances.

  • Everyone likes the idea of waste-to-value, but in practice it comes with many challenges that need thoughtful consideration.

    The first is the assumption that waste would be a free raw material. In some cases you might be able to get the supplier to pay you for taking hazardous waste off their hands. Alternatively, they may be happy to give it to you for free because it’s cheaper than paying for disposal. Then again, your supplier may come to understand that their waste is a revenue stream once you prove its value as a raw material. Assuming one of these scenarios and finding another to be true could break your techno-economic model. 

    I investigated using mining tailings as a starting material for my mineralization process. Many of the people who encouraged me to pursue this direction assumed that all mining tailings are environmental liabilities and that companies would pay me to take them off their hands. In reality, the specific types of tailings I needed were pretty innocuous, so these suppliers did not have a big incentive to get rid of the material without charging at least a modest price. 

    It’s easy to make assumptions about impurities as well and whether they would interfere with a desired process or product. It’s also fairly common that waste streams from different suppliers have different compositions, and this can lead to a lack of reproducibility in your processes or products. In my case, using mining waste (tailings) as a raw material would have resulted in uncontrolled levels of iron or other color-causing impurities, making the product unsuitable for paints and coatings without complex and cost-prohibitive additional purification.

    There are many other considerations including regulations for the type of waste you are evaluating, inherent limitations on scaling your own production because of the cap on how much waste can be sourced, geographical challenges including plant colocation or material transport. And the list doesn’t stop there.

What is our plan for world domination?

There is an enormous abyss between lab-scale prototypes and full-scale production facilities. We need some sense of the steps required to move from the abyss to market traction. Our trajectory is rarely a linear one, not just because of technology but also because of the many people, partners, and stakeholders who shape and reshape the market landscape on a daily basis. There are forces at play that you need to constantly monitor. 

  • Competing with incumbent products on cost is challenging even at scale, but it’s doubly challenging when your production volume is low. Even when we’re not aiming to produce a commodity, it still behooves us to ask whether there is a more premium version of the product with a smaller market but higher price point. 

    When selecting a beachhead market for my carbon-negative mineral product, pricing considerations were top of mind and came as a tradeoff with market size. According to some reports, the global calcium carbonate market was valued around $48bn in 2024. The challenge is that the majority of that market consists of ground calcium carbonate (GCC), a mineral typically sold for $100/ton or less depending on the purity and use case. A small portion of this market (~$2bn in 2024) consists of precipitated calcium carbonate (PCC), a synthetic limestone used as a high-performance mineral additive. PCC pricing also varies depending on purity, morphology and particle size, but could be $500/ton or more in select applications.

  • Even if the technical team is sold on your data and samples, there are many other things that could become barriers to supplying your customer with a material.

    Of course, there is the issue of scale. If you can only make 1 kg of material at a time, and they use 1000 tons per day, then you face an immediate barrier to adoption. Nobody is expecting you to be at scale on day one, but you do need to have a clear definition of what scale means in your industry. When engaging with customers, your goal is to understand not only what qualifies as meaningful scale for a pilot run but also what a minimum order quantity would be for someone to incorporate you into their operations. 

    When investigating the use of my carbon-negative minerals as additives for concrete, I quickly realized that the volumes of material concrete manufacturers require are massive. One customer said they were producing thousands of tons of concrete per day, and they ordered additives in units of rail cars. 

    Regulatory barriers can be another issue. The construction industry is highly regulated, so being able to meet existing ASTM quality standards is necessary for adoption in this market. I won’t go into detail here, but sometimes this requirement is extra challenging because the standards are written with the incumbent solutions in mind. For example, many concrete standards are based on specific chemical compositions, so a new material, by definition, can be disqualified even if it meets all the necessary performance specifications. 

    Finally, it’s often the case that a specific quality certification is required to become a supplier. When investigating applications for my carbon-negative minerals in the paints and coatings industry, I spoke to a large paint manufacturer with a policy that requires their suppliers to be ISO 9001 certified, which inherently disqualifies most early stage startups.

  • The capital stack for early stage, deep-tech startups is often a blend of non-dilutive and equity financing. Regardless of which funding sources you choose to pursue (and which you secure), you should be periodically evaluating the alignment between the funders’ priorities and your business goals. Venture investors will likely advocate for aggressive growth plans capable of generating meaningful returns on the timescales of their fund. This could be amazing fuel for your startup and accelerate your growth if aligned with your plan, but it might not be the right fit for the type of business you are trying to build.  

    Grant funding is particularly attractive in being non-dilutive, but many grants have substantial administrative overhead and can become misaligned with your plans depending on the contract flexibility. There’s no expectation for 100% alignment between the grant call and your technical-development roadmap, but if the overlap is slim, then taking on the additional contract can change your trajectory and slow you down even if it is an additional source of capital. With limited time resources as a solo founder, I found myself having to be selective about which grant programs I applied to in order to maximize my chances of being awarded the funding but also minimize the possibility of veering off course.

    Finally, business priorities can change substantially between the time you first submit a grant application and when awards are made. It becomes that much more important to ask yourself whether pursuing the grant still makes sense in your new business context or whether it would change your trajectory and become a distraction.

  • We must pay particular attention to timelines when it comes to commercializing deep-tech innovations in a startup with limited resources. Your funding and commercialization strategy will dictate which milestones you need to hit at each stage of the business. The rapid progress expected of startups in 1-2 year timeframes can only be achieved if your tech iteration cycle is short enough to allow you to answer enough key questions and retire sufficient technical risk at each stage. 

    So what is the timescale of your tech iteration cycle? For example, if you are making a new type of plastic that is derived from biomaterials but still has the strength of conventional plastics, you might have an initial tech iteration cycle of just 1-2 weeks between formulating, manufacturing some test samples, measuring their strength, and then taking the learnings from your data to come up with improved formulations that go back into the testing loop. When you get to later phases of development, this cycle will change and become 2-3 months if, for example, you are now blow molding plastic bottles from this new material and then testing their performance and collecting additional data and customer feedback. 

    Is the timescale of your tech iteration cycle short enough? I’d caution against false precision here. Even a back of the envelope calculation can be helpful. How many experiments do you estimate you need in order to achieve your milestones at this stage? Let’s say in the sustainable plastic example above we need to run 200 experiments total, and we can run 20 experiments at once. This means we need 10 cycles (yes I know science doesn’t work this way in practice but this is meant to be an order of magnitude calculation). If each cycle is 3 months, then we are at 30 months which may exceed the startup’s runway. In this case, we’d need to find strategies to shorten the iteration cycle in order to meet the critical milestones that support our commercialization strategy.

    How can you shorten your tech iteration cycle? One successful example from a fellow entrepreneur comes from the field of agricultural robotics. When developing robots that interact with crops, deploying your prototype robots into the field at partner farms initially seems like a great way to test the system's real-world performance in the target operating environment. However, the need to travel to distant sites, ship precious prototypes, and align with seasonal cycles of crop availability severely limit how often the deployment cycle can be repeated to 1-2 times per year.  

    One way to shorten the iteration cycle might be to build and operate a small-scale test farm producing crops of interest right where your engineering development takes place, but this is still limited to the speed at which plants grow if new crops are required to conduct each destructive test, as is the case in systems that prune or harvest. To even further reduce the cycle from months to days, a simulant is needed. Yes, physical simulants in the form of artificial plants and virtual environments differ from the 'real' problem, but with careful planning and creative design they can provide a relative sense of how your underlying KPIs are evolving with each system change. This allows for more deliberate, controlled, and selective use of lengthy testing in real-world environments.

    Reducing the iteration timescale from 1-2 cycles per year to 1-2 cycles per week can be transformational for a startup, especially in the early stages of technology development when there’s lots of uncertainty and many variables to test. The key is that you can’t just accept the timescale limitations of one means of testing. It’s incumbent upon you to find viable ways to drive down the duration of your tech iteration cycles to match the business reality of your venture. In some cases, it might be the most important innovation work that you do.

What is our theory of change?

You are likely making something that replaces an existing product or process in some way. So why would people change what they currently do? 

When we talk about change and impact, we often get too philosophical, too fast. Ideals are good, and deep-tech entrepreneurs need to be driven by a deep sense of purpose; nevertheless, there are pragmatic ways of thinking about change and impact that we have to attend to.

  • Asking this question excites my inner scientist. It feels a bit like searching for literature precedents for the synthesis of a new molecule. What precedents are there for similar products gaining adoption in this market? What can we learn from the past that can inform how we approach this problem? 

    This exercise requires a great deal of creativity. We have to think broadly about how to define “similar products,” and we need to understand processes of adoption in general, not just for the product in question. 

    It’s crucial to be equally curious when trying to identify key decision makers. Regulators have been the key decision makers in the concrete industry’s adoption of Portland Limestone Cement (PLC), a lower-emissions product. Widespread adoption of PLC required acceptance from building codes and departments of transportation. 

    Zooming out and considering macro-level forces also helps here. Have there been any recent regulatory changes? Are current conditions driving adoption for your solution or creating unanticipated challenges? 

    One example that has received significant media attention is PFAS-free technologies for non-stick cookware or textiles. Regulatory changes, both implemented and anticipated, have significantly accelerated the adoption of non-fluorinated alternative materials.

  • You probably already have a target data sheet with all the specifications that your product needs to hit; however, targets alone are not sufficient to understand the customer’s definition of what it means to meet performance metrics. 

    Understanding what level of repeatability is needed in your industry is key. A one time test might be considered “promising,” but little more than an academic result. Do you need to show testing on three different material lots or three hundred? Will this testing take two days or two years to complete? Do you need to demonstrate that you have a stable process that can deliver consistent products across time, environmental conditions, or geographies? 

    When investigating using my carbon-negative minerals in the paints and coatings industry, color characteristics were an immediate concern. It wasn’t that I couldn’t meet an absolute whiteness metric; rather, I couldn’t guarantee color consistency across materials sources from different mines with different geologies. 

    Third-party certifications can be a powerful tool to instill customer confidence. Can you meet an ASTM standard or obtain independent testing that demonstrates your product’s performance? What other evidence do your customers require to be convinced that your technical solution really works? I can’t stress enough how important it is to make sure you and your customers have the same definition of what it means for something to “work.” In most cases, seeing is believing. When I was working in additive manufacturing, we invited customers to witness live demos of complex objects being printed. There was no substitute for this live experience.

  • Life Cycle Analyses (LCAs) have become more and more prevalent, and they are a critical step in understanding the full impact of our products. There are formal methodologies for conducting LCAs on mature products and processes, but for early-stage startups even back of the envelope calculations can be extremely powerful in determining the true impact of the solutions we are proposing. 

    When claiming an improvement, it’s important to determine a meaningful comparison point to the status quo. To illustrate this, consider a carbon-negative mineral embedded into a plastic product. A carbon-negative mineral lowers the emissions associated with the raw materials of the product in two ways: by displacing a portion of the plastic volume, and by adding additional carbon offset through the carbon sequestered inside the mineral. Let’s say the virgin plastic has a global warming potential (GWP) per kg material of 2 kg CO2e, and the mineral is -0.3 kg CO2e.  A back of the envelope calculation suggests that a formulation with 20% carbon-negative minerals would have a GWP of 1.54 kg CO2e, a 23% reduction in emission compared to the virgin plastic. That’s great, right? Actually, it’s just an apples to oranges comparison. The true comparison is with the same plastic compounded with 20% regular mineral filler, which could have a GWP around +0.3 kg CO2e. The resulting benchmark plastic GWP would be 1.66 kg CO2e.  The carbon-negative mineral version is better than the benchmark by only 7%. 

    We have to look across the whole value chain to understand the true impact of a product. Yes, adding a mineral filler to a plastic product can effectively lower the GWP of the product by displacing some of the high GWP plastic volume; however, the mineral has a much higher density than the plastic. If the product has the same volume, this means our mineral-modified version is significantly heavier, so transportation of this product to downstream processing and its final customer is going to produce more emissions than the equivalent plastic-only product. 

    Understanding your raw material supply chain is a necessary step for estimating the true GWP of your product. If your process is very low energy but requires the use of a new reagent which itself takes more energy to produce, the net effect will not be beneficial. As a hypothetical, imagine we can sequester CO2 in mineral form using a base like calcium oxide. That process alone might indeed capture CO2, but the reagent calcium oxide is manufactured in a high-temperature process that releases CO2 from limestone (calcium carbonate). The overall LCA would show net emissions, not consumption, of CO2.

Get in touch!

Do you have more questions of your own that should be added to this resource? Or questions about the questions here? I’d love to hear more about your journey and any feedback you have.

Reach out to hello@askbetterquestions.tech