This is what it takes to turn scientific research into a commercial software product in a repeatable and scalable way

In early 2018, we started Ukko Agro Inc. with the goal to bring proactive decision-making to the commercial farms in sustainably growing more crops. Since then, Ukko Agro has released a predictive analytics platform that offers several predictive analytics products relied upon by the commercial farms and enterprises across the agriculture value chain in sustainably growing successful crops in the United States, Canada, Sweden, Brazil and Argentina.

In building predictive analytics products, we combine foundational agriculture science (independent research) and data science to model solutions that are explainable, can adapt to numerous in-field variabilities, and most importantly can be trusted by commercial farms.

Universities and corporations around the globe invest billions of dollars every year in research activities. A significant small portion of those activities is turned successfully into commercial products with a justified ROI (Return on Investment).

Have you ever wondered what it takes to turn scientific research into commercial applications in a repeatable and scalable way?

A disciplined approach is key to solving this complex challenge. Along the way of building a predictive analytics platform with several products and growing the company, we identified a repeatable and scalable way of turning scientific research into commercial products with predictable timelines and ROI.

In this article, this disciplined approach is organized by: Predictable Product Innovation, People, Processes, and Technology aspects with examples of how we did it.

  1. Predictable Product Innovation — How do you decide where to invest/divest for a predictable ROI and realistic timeline?

Deciding where to invest/divest next is one of the most critical decisions a company makes to innovate and grow. Staying laser-focused in doing so while observing the opportunities to exploit new trends and technologies can be even harder.

As a co-founder, in balancing the company’s ambition, market opportunities and company’s financial health, I learned to rely on a series of qualifying questions in this decision making:

  • Does this research activity align with the company mission (in our case, mission to grow more crops sustainably)?

It is crucial to deliver on the core value promised by a company. While the Technology and processes continue to evolve over time, the company’s core value remains its north star.

  • Does this activity create a new technology from the ground up?

Without finite milestones and timelines, this activity may or may not be as enticing as some of the other activities that can offer a higher and clear ROI.

  • Will this activity result in delivering an existing research/technology to the target customers in a simple, reliable, and cost-effective way? As a result, will it enable the target customer to either save time or save money or increase revenue or achieve a combination of these?

The answer to this loaded question, in many cases, surfaced an activity with a predictable time frame and outcome. The ROI depends on understanding the gaps in the existing research/technology in building a simple, reliable and scalable product. With experience, we also learned to prioritize revenue generating opportunities over cost reduction opportunities for a given customer segment.

  • Does this activity glue together or enhance existing independent scientific research to solve core problems for the target customers?

Often such opportunities resulted in identifying activities with a high ROI and a predictable time frame. While we still ended up spending time to understand the gaps in each independent research/technology in making the final call, this activity generates most IP (often patentable IP) for the company. The key to observing high ROI in this activity is balancing the predictability of required effort and the value delivered to the target customer.

2. People — What skills do you need in a team? How do you keep moving forward when you can’t have all the skills?

  • Identify product dimensions

One of the most important aspects of building a team for successfully turning scientific research into a commercial application is grasping the product dimensions in order to translate the skills a company is going to need in successfully building and scaling a product. In our case, the product dimensions were in the field of Climatology, Plant Physiology, Plant Pathology, Epidemiology, Crop Modelling, Computer Vision, Remote Sensing, Crop Economics, Behavioural Economics, IoT, Data Engineering, and Software Development

These dimensions can be identified at the intersection of the target customer’s core problem, market size (demand), and the current state of scientific research.

  • Establish advisory groups

The good news at this point maybe that you have figured out what skills are needed to build and grow the product. The not-so-great news maybe you can’t hire them all due to a budget challenge or issue with the right fit or sometimes full-time hiring needs may not be there yet.

At Ukko Agro, we solved this challenge by identifying the most common denominator skills for the near term in building and scaling the product. Product teams were established around these skills. To continue moving forward, we established technical and commercial advisory groups with advanced skills. These groups guide our team at both tactical and strategic levels. This strategy multiplied Ukko Agro’s team capability by several folds in building and scaling Applied Research software products. With the advisory groups’ help, we have a clear plan to grow the business and product in locked steps.

3. Processes — How do you say ‘no’ more often than ‘yes’ to stay focused?

One of the biggest challenges any growing company faces is knowing where to focus and where not to focus. In turning scientific research into commercial products, this challenge further intensifies. The processes need to be uniquely designed in such a way they can analyze all complexities and variability in answering in a clear way — where should the entire company’s energy and focus go. When executed well, this focus (ideal case one, max 3) means most of the efforts within the company will be channelled in a single direction in creating value. As a co-founder and the CTO wearing multiple hats, this was one of the toughest challenges I experienced.

Since then, we have been applying the following two highly effective techniques to answer a single-dimensional question — is the outcome at a given time/situation worth the effort:

  • 80–20 Rule for the business strategy

Where will the 80% of revenue come from in the immediate, near and medium-term? What 20% revenue-generating opportunities (today) can turn into 80% revenue-generating possibilities in the medium to long term? Answers to these questions have enabled us to identify and prioritize the market opportunities.

  • Risk Acceptance Discovery

For the identified market opportunity, we developed a short risk acceptance criteria. The 5 sections we cover are:

Business viability risk — answers revenue potential by target geography

Value risk — identifies core and extended problems that need to be solved for the customer and end-user (in case they are different)

Usability and Engineering Risk — defines: What does success look like? What a Minimum Viable Product (MVP) and scaled-up solution may look like? Can a successful MVP/scaled-up solution be built at an acceptable quality, time and cost?

Scientific feasibility risk — Identifies the gaps in the target base research, and to what extent it can be implemented/integrated

IoT Risk — Identifies the IoT limitations in balancing customer experience and predictive model requirements for building MVP/scaled-up solution

The answer to these questions enables us to evaluate and prioritize the opportunity size and the probability of success in implementing a solution.

4. Technology — How do you strike a balance in prioritizing delivering a strong customer experience, building on limited scientific research, managing acceptable accuracy/recall rates for the predictive models, and implementing around the technical feasibility?

Building scientific research-based software products is more art than science when it comes to striking the balance. With scientific research-based software products, the biggest challenge is managing complexities and variabilities individually and collectively. They have to be managed such that the product as a whole delivers on the business model in the simplest possible way.

As the CTO, achieving this balance is one of the most complex challenges I deal with on a daily basis. Each product (current or new) has its own challenges and quirks. In striking the balance, we adapted to a Concentric Circles Approach. Under this approach, in chronological order, we:

  • Identify the customer and end-user for the given product.
  • Identify the core value (ideal scenario one, maximum 3) to be delivered.
  • Identify the behavioural side of customer and end-user to design the solution with the solution adoption goals in mind.
  • Design MVP/scaled-up solution (all software, data science, IoT and scientific research domains) around this planned adoption. This is the step at which a reasonable balance is achieved. Solution at this point represents the innermost circle in the concentric circle approach. The next iteration represents a circle around the innermost circle, hence the term concentric circles.
  • In the next iteration, around the core problem, identify the next top secondary problem/challenge/opportunity, repeat the building process as suggested in the previous steps.

In addition to the above, there are many more considerations, techniques, and hard work and dedication of all our team members go into building scientific research-based software products. This was an attempt to summarize some of the key techniques we have been applying at Ukko Agro in building products in a repeatable and scalable way.

Please feel free to reach out directly to discuss any of these techniques in detail or share your viewpoints/questions.

Comments are closed.