Efficient deep tech product development framework (Part 1)

In my previous 3 posts, I explored nine uncertainties that deep tech product development faces. In this article, I would discuss how deep tech product development can happen efficiently despite the uncertainties.

Summary of uncertainties

To roll-in all the solutions for deep tech product development, a two-part stage-gate process with iterative properties is most ideal. The first part is a feasibility stage-gate process while the second part is development stage-gate process. Tasks to be performed at these stage gate processes differ as the levels of uncertainties present are different. The goal of this first part stage-gate process is risk assessment and risk mitigation so that the full-blown development stage-gate process (part two) is in a better position for success.

Two part stage-gate process with iterative properties

Types of development work

Before diving into the framework and the detailed work involved at each stage-gate process, it's beneficial to first explore the types of tasks the development team can tackle. Generally, there are only five types of development work. Any technical endeavor undertaken by the product development team can be categorized into one of the following areas:

1. Discovery: Discovery work is rife with the highest level of theoretical and practical uncertainties. These projects often make slow progress until a breakthrough occurs, involving considerable trial and error. Estimating the time required for discovery and understanding the underlying technology beforehand is challenging. However, if a breakthrough happens, it can be tremendously rewarding and valuable. This type of development work is primarily found in academic settings and is less common in corporate environments (except for upgraded iterations of existing products).
2. Repurposing: This type of work involves applying technologies known to work in other fields or applications to a new product. Just because a technology has proven effective in other products doesn't guarantee success in a new product, so significant risks are involved. However, since the principle behind the technology is already known, the development process has a clearer direction and a greater chance of success. Many deep tech product developments incorporate repurposed technology.
3. Building: Similar to constructing a building or developing a software application, building-type development work entails assembling technological modules that are relatively well-understood. While individual modules are no longer shrouded in technological mystery, integrating them requires design finesse and careful consideration. There's usually a defined workflow for how the work should progress, and development unfolds in a linear and more predictable manner.
4. Validation: This type of development work aims to tailor products or technologies to meet specific use cases or performance requirements. In most cases, the product or technology is already known to be feasible. The main challenge here is quantifying the performance of the product over time, batches, or samples. Typically, the development team needs to conduct multiple tests, comparing against controls or competitors' products to validate the technology. A product specification or data sheet usually emerges from this work.
5. Optimization: Optimization development work involves making subtle adjustments to the product or technology to enhance performance metrics. For instance, in semiconductor dry etching, reducing dry etch taper or within-wafer non-uniformity falls under optimization work. The scope of optimization work is usually well-defined, as the adjustable parameters are known before the work begins. The focus lies in tweaking these parameters to achieve optimal performance.

Feasibility stage-gate process

Conceptualization: Establishing the Foundation

The conceptualization phase is about laying the groundwork for the project. It involves sketching out how the product will come together, assessing risks, determining its value, and clarifying its focus.

1. Market Identification: This step acknowledges the influence of market dynamics on product development. Defining the target market early on is crucial, even though complete market validation may not be possible until prototype release or product launch. This exercise helps assess the viability of the product and sets a clear direction. It's recommended to allocate a budget of no more than 1% of the Total Addressable Market (TAM) for this phase (In fact, no more than 0.5% will be better).
2. Product Value Chain Identification: This involves identifying the various elements of the value chain, both upstream and downstream which might already be present, requires development or might be complex deep tech by their own right. The semmering railway project is a good example of deep tech products (back then tunnel boring technologies are not as advanced) that cannot provide immediate value but requires other supporting elements (locomotive) to eventually fulfil the value proposition. By understanding how value is created and exchanged within the ecosystem, gaps can be identified so the product can be scoped to maximise success.
3. Modularization: Breaking down the product into smaller modules using the box and arrow diagram helps define its components and their functions. While the specifics of what goes inside the boxes may evolve, clarity in input-output specifications is essential to help define the problem that needs solving.
4. Assumptions: Listing and assessing assumptions in a table key to understanding project risks. Gather as many experts as possible to brainstorm all the possible assumptions and include a column in the table to map the assumptions back to its relevant module. Perform a simple assessment of the listed assumptions based on their impact, degree of uncertainty, ease of development and type of development.

Planning: Setting the Stage for Feasibility

During the planning phase, logistical and technical aspects are addressed to prepare for the feasibility study.

1. Module Details: This step involves delving into the mechanistic details of the technology inside the box and arrow diagram. It should explain how the inputs are converted to output within the box and outline technological possibilities and alternatives.
2. Shortlisting Assumptions: Along with box and arrow updates, new assumptions may arise. Continuously updating and prioritizing assumptions ensures that focus remains on key areas during the feasibility study. From the sorted list of assumptions, pick out the assumptions to be addressed during the feasibility validation stage and these will be the scope for feasibility study. Assumptions should not be confused with building a prototype. Rather, it should be a quick test to know if the technology (which is a component of the overall product) is feasible and ready to be deployed into the product. As such, generally only assumptions that are breakthroughs or repurposing in nature are targeted for validation.
3. Skill Set Identification: Shortlisted assumptions are the scope of the feasibility study. Identifying the skills needed for assumption validation helps assemble a capable team. Grouping skills based on requirements ensures efficient resource allocation. Apply good project management planning skills here to prevent overloading of burden on a single person.
4. Operational Planning: Planning timelines and budgets is crucial for managing resources effectively. This feasibility stage-gate process is aimed at waste minimisation and loss reduction. Aligning budget allocations with the TAM prevents unnecessary investments into products with low ROI. Adopting an output-centric approach by limiting number of output before kill/continue review sessions minimizes timeline uncertainty.
5. Sunset criteria: Whether a technology work is more important than how well the technology work during feasibility studies. The goal of this feasibility stage-gate process is to answer questions of whether an idea or technology is even possible. Metrics can be second-rated but if the technology can produce the desired phenomena, the validation is successful. Binary sunset criteria are preferred to avoid endless development spiral.

Feasibility Study: Testing Assumptions

The feasibility study phase involves validating assumptions and addressing technical challenges. Management should adopt a less hands-on approach and hand the reins over to the development engineers and scientists as they are the experts in their fields. There are just 2 types of meetings necessary at this stage.

1. Logistics Meeting: Regular meetings to track progress and prioritize tasks. In a ‘checklist’ format, run through the assumptions and note their status. Ensuring clear communication of weekly budgeting and timeline updates also helps fosters trust and urgency. A quick 30min session weekly will suffice for this type of meeting.
2. Technical Meeting: Addressing technical issues and challenges as they arise is essential for progress. Technical issues generally impact budget, timeline and might invalidate the entire product and call for timely intervention. Thus, these are on-demand meetings spun out of the logistic weekly meetings when items are stalled or utilizing too many resources. Technical meetings run for as long as it takes for the issue to be resolved or validation work terminated. On a slightly different note, when an assumption validation work is completed successfully, it should also be presented in such a technical meeting to relevant stakeholders as a check to ensure that the results are watertight.

The feasibility stage-gate process provides a structured approach to managing innovative deep tech product development without undue exposure to technological risks. By systematically progressing through these phases, organizations can boost the probability of product success with less initial investment.