Innovative Entrepreneurship: Managing Business Risks

What if you could prove your startup's value before having a finished product? A new methodology is transforming how investors evaluate early-stage startups, focusing on strategic validation rather than traditional metrics. While you're sticking to outdated valuation methods, startups with weaker products but smarter validation strategies are attracting smart money that's slipping through your fingers.Don't let your innovative idea get undervalued–learn how to demonstrate its worth at each development stage.

In the rapid world of technological changes, evaluating innovative startups is becoming increasingly challenging. How can you minimise risks while not losing the opportunity to create something truly groundbreaking? Recently, our [Kmbs] graduate Dmytro Shestakov published a book titled "When Businesses Test Hypotheses," in which he presents a cutting-edge method for hypothesis testing that helps evaluate the potential of startups and internal risks for both new projects and existing businesses.

We spoke with Dmytro about innovations, strategic agility, his approach to evaluating innovations, and how entrepreneurs can apply these ideas in a dynamic business environment. Dmytro Shestakov ― founder of technology startups, graduate of MBAF kmbs and the "School of Strategic Architect" program at kmbs, graduate of Executive MBA from London Business School, member of the expert council for evaluating startups of the BRAVE1 defence technology cluster, and expert of NATO’s deep tech accelerator DIANA―offers a new perspective on this challenge.

Dmytro Shestakov, 2024The interview was originally published on Kmbs.

Creating a Culture of Innovative Entrepreneurship

Dmytro, what does innovative entrepreneurship mean to you?

D.S.: Innovative entrepreneurship is about agility and execution. It's the ability to constantly change oneself and one's enterprise, adapting to new circumstances—both internal and external. It's also about continuous learning. For example, how quickly you started using technologies like ChatGPT or Claude from Anthropic.

There's a perception that innovation is something difficult to predict and may have unknown results. Yet you say that innovation is not about the unknown. Why?

D.S.: It hasn't been a black box since roughly the late 60s or early 70s. Silicon Valley, the Israeli economic miracle, the Swiss economy, etc.—these are examples of successfully implemented methodologies for economic development. It's not a coincidence, but the result of specific decisions, such as the creation of DARPA (Defense Advanced Research Projects Agency), and the activation of the innovative economies development model. These approaches were then scaled in Israel, Singapore, Switzerland, Sweden, and now in Britain. DARPA itself is the developer of technologies such as microchips, the Internet, GPS and other satellite technologies, and graphical user interfaces. These are all results of deep scientific research. Such types of innovative developments are the riskiest for investment because they require funding for at least 10 years. This is an area where businesses usually don’t invest due to different time horizons. And from a national security perspective, 20 years is not a big deal.

Large Language Models, i.e., the networks we know, such as ChatGPT and similar, emerged within the DARPA program, and companies like Google, Anthropic, Microsoft gained access to these models through contracts for developing technological solutions for national security needs.

Agility in Risk Management

Why does this work in other countries? How do they use this? How can it be adapted?

D.S.: First and foremost, it's about social awareness and the state's ability to make strategic decisions. Business and the state have common strategic tasks but different horizons for making strategic decisions: for business, national security issues are not a priority, while for the state, it's one of the main tasks.

This is the example of the USA, Israel, Singapore, Sweden, Switzerland, Canada, and now the UK. They have created autonomous state organisations to coordinate and implement risky high-tech projects. For example, ARIA in the UK. These organisations are independent of political campaigns and have autonomy and immunity within the state structure. They are responsible for strategic decisions and setting the agenda in the field of innovation and technology development.

How can a startup apply an agile approach to risk management?

D.S.: It is not a simple question, particularly due to the continuous lack of resources in most startups. The exception is startups that offer innovative products or services based on advanced scientific achievements and technologies, where money problems are usually much less important. For others, it's a challenge: how to balance in conditions of limited resources.

First, it's about multifunctionality, which is related to agility, adaptability, and learnability. Large companies like Google and others are already selecting people not only based on mental capabilities but also on the speed of thinking and adaptation. For startups, this agility is simply necessary for survival.

Dmytro, tell us how the idea to write a book about the methodology for evaluating innovative startups came about.

D.S.: I'm from the fintech industry, I worked on tech startups, but it's a pragmatic business. We used machine learning (ML) and natural language processing (NLP), but the basis still remained economic and mathematical modelling.

The idea of writing the book came because of searching for an answer to the question: how to evaluate innovative projects, particularly in the defence industry. Traditional evaluation models don't work here. I started looking for an answer back in 2009 and began to research more deeply in 2012 when I entered the Doctoral School of Kyiv-Mohyla Academy.

The idea of the Hypothesis Testing Framework itself emerged in 2016 when, under the guidance of former DARPA’s Director Dr Tony Tether, my teacher, we were building our Ukrainian DARPA. We faced the question of how to evaluate innovation and risks. Tony jokingly said that if we solve this problem even partially, we'll go down in history.

How to use this? Read, think, and become methodological and technological. And this is primarily about abstractions, not parts of code. The DARPA model is an abstraction. Currently, we don't have it in our architecture, while it is a central element of national security systems in the USA, UK, Germany, France, Japan, Israel, Switzerland, Canada, and two dozen other countries whose standard of living we aspire to.

Evaluating Innovative Startups 

What is the main idea of your framework for evaluating innovative projects?

D.S.: When we talk about evaluating something, we usually look at money, and at the end result. But how do we evaluate the result when we only have unclear interdependencies and assumptions?

The main idea of the framework is to provide a systematic approach to assessing risks and prospects of innovative startups with a high level of uncertainty, where the result is not yet known and it's difficult to assess whether the technology will work.

The framework I developed allows you to look at the project through the prism of its development stages, that is, through the evolution of solutions. When you gradually validate assumptions and reduce investment risk, you approach an intuitive understanding that each validation step reduces risk. However, traditional methods, such as discounting, often don't reflect this dynamic and therefore aren't suitable for evaluating innovative startups.

Can we say that your approach is also useful for large corporations that want to remain innovative?

D.S.: Yes, but with some caveats. Large corporations can remain innovative by investing in startups and supporting their development, even if they themselves aren't engaged in R&D. This allows them to adapt to changes in the business environment without the need to develop new technologies on their own. However, if a company doesn't update its operational models and processes, it risks becoming less adaptive and failing to respond to the emergence of "black"swans"—unexpected events that can strongly impact the market.

The developed framework helps corporations assess what gaps need to be filled and how to better adapt to the changed reality to remain competitive in a dynamic business environment. Because the corporate world now has problems with agility and adaptability.

What does strategic agility mean to you within a startup and within an existing business?

D.S.: There's no fundamental difference. Strategic agility is the ability to adapt in a timely manner to changes caused by both internal and external factors. An important element of being strategically agile is doing it proactively rather than reactively.

The difference is only in the rigidity of the business and simultaneously its complexity. It's mentally harder for a business to experiment than for a startup. The capital at risk is disproportionately different. Moreover, experimenting is always more expensive than focusing on efficiency, as the average cost of a team capable of working in experiment mode 24/7 is significantly higher.

Furthermore, business is in many ways more complex than a startup. For example, from a risk perspective, there's a large layer of work that arises due to the volatility of demand and opportunities, changes in maturity, and singular transformational changes.

Effective Approaches to Hypothesis Testing

The testing method you describe in your book is based on five hypotheses that consider lifecycle stages. Could you tell us more about this?

D.S.: These are the hypothesis of team competence—both technical and business,—the hypothesis of technological feasibility of the product or service, the hypothesis of its value to the customer, the business model hypothesis, and the hypothesis of market depth. They are then broken down into finer assumptions and analysed from the perspective of startup development stages and lifecycle.

In other words, they aren't new by themselves, but in this configuration, filtered through the stages of a startup's lifecycle, degrees of innovation, and with backward induction logic borrowed from real options, the proposed framework and method of hypothesis testing open up a new perspective for understanding, measuring, and managing startup risks.

You mention that this is a method of testing hypotheses for startups in the context of post-war economic recovery. How exactly will this help in the context of recovery? What are the advantages? How best to use this?

D.S.: From the perspective of behavioural finance, that is, psychology in business, if a person doesn't understand something, they perceive it as danger. Conversely, if something becomes more understandable to you, you treat it with more trust. This is how heuristics and biases work. Add to this the principles of system dynamics or the theory of constraints, and you'll see how the speed of movement and the amount of money in the innovation and startup ecosystem will increase.

In software engineering, there's a concept called abstraction. And it's not what we're used to perceiving as something general and meaningless. On the contrary, it's a mathematical construction that can interact with other abstractions. Software architecture is created on such abstractions, but to do this, you need to know and understand them.

Similarly, the architecture of national economies consists of abstraction models, forming value chains and ecosystems. Obviously, it's more reasonable to rebuild the economy not just as innovation-driven or data-driven, but as knowledge-driven, that is, technologically.

Ensuring Long-term Sustainability

How can you minimise risks in a startup while not losing the opportunity to create something truly breakthrough?

D.S.: You need to clearly distinguish between two fundamentally different types of management approaches to balance between ever-better execution of the current strategy and experimenting to find new opportunities.

In my understanding, a startup is about the business inception stage, constant experimentation until there's a need for operational efficiency during scaling, meaning you grow to the scale-up and growth-up stages.

Accordingly, a startup is about strategic types of uncertainties, not related to the efficiency or quality of the operational part of the business, which is managed as an experiment (preferably controlled and methodological) and requires strategic agility.

That is, you need to determine what strategic risks are involved. And here humanity got stuck. Meaning, a methodological solution was missing. And in my work, I propose such a framework that will make it fairly easy to identify and understand the nature of strategic risks in startups.

What would you advise startups that are testing their hypotheses and trying to enter global markets?

D.S.: Interestingly, startups often surpass businesses technologically, but you shouldn't wait until the product is fully ready or at least minimally viable to start validating hypotheses. Often, startups begin experimenting only after they've invested a lot of time and effort, and this is quite likely to lead to burnout. The startup and innovation industry isn't easy, and those entrepreneurs who survive become a valuable resource for the corporate world, which currently suffers from a lack of agility and adaptability.