A framework for Deep Tech product phases

This article suggests a simple framework for considering the Deep Tech startup product phases and their wider implication for organisational maturity. The context is especially focused new ventures faced with both the typical cold start problem of a new product, along with the uncertainty of frontier technology and all the markets, supply chains, and infrastructure it requires. This exploration is inspired by the running conversations and observations I have working in and around these teams, and the comparisons and contrasts to the typical experiences of product leaders in enterprise or startup organisations.

Whether you’re a founder, a product leader, an investor, or a curious human trying to make sense of what’s real and what’s just hyperbole in emerging and frontier technology, it helps to have a lens to look at the specific phases and wider context of such organisations and their output.

To get there we must first take a quick detour to understand the limits of the more popular framings for technology hype and readiness. After which we can explore the framework that I’ve found helps me not only understand the needs of the product I’m contributing to, but the relationship with the wider organisational evolution in the process.

Perspective one: looking from the outside in

Coming from the widest external public perspective we can see how the mainstream technology coverage is neither interested nor equipped in navigating the nuances of Deep Tech companies. Things are just a little too deep for casual conversation and some liberties must be taken when reporting on popular science. Even accomplished scientists can steer towards dubious populism given enough attention and an audience. Into this space we see the usual collection of management consultants and industry analysts. Both have a crucial role.

Given the brand recognition of firms like McKinsey, we tend to use their figures when it benefits us (like I did in my talk at Open Source Summit in 2023). We might quibble about how accurate their methodology of estimating TAM is, or how much capital inflow is actually being reported (given the private nature of many sovereign funds or the optics-wary caution of startups tapping bridge rounds), but it doesn’t really matter. The work to connect TAM, SAM and SOM to near-term tactical efforts in a long-term industry like quantum computing is as much about rigour and a modifiable model as it is about deterministic outcomes.

So that’s the consulting side. Over on the research and services side we see the likes of Gartner with their iconic hype curves and magic quadrants. These are another mainstay of slide decks, but we should again ask ourselves on the operational side what these really tell us. Gartner is excellent at procurement and the nuances of sourcing, but neither of their views is designed to shape our understanding of Deep Tech progress or our actions at any given stage of getting these things to market.

Take the following chart from Gartner’s 2020 review on artificial intelligence. Used well it is a simple device that encourages consideration of expectations versus maturity of an area of interest. Used poorly it becomes a totem that lives in management slide decks and conveys assumed domain expertise on those who conjure it. Even seeing how well some of the estimates this particular chart made have panned out, it’s once again not operationally useful for product and strategy teams actually building products. We just can’t rely on these tools from the outside in. So let’s go the other way.

Perspective two: looking from the inside out

While there are no singular methods or systems of Deep Tech startup product development used across the board, we tend to inherit a mix of the industries and cultures that we straddle.

From the industrial and manufacturing side we get the likes of Six Sigma, Critical Chain Project Management (CCPM), Lean and Waterfall. All well and good but perhaps not ideal for Deep Tech ventures overcoming the cold start problem of little to no existing operations, market, or even supply chain. This is one of the reasons why veteran manufacturing product leaders often struggle at Deep Tech startups, attempting to optimise what really doesn’t yet exist.

Over on the Silicon Valley side we have the flywheel of cultural adaptation that took Herbert Simon’s seminal “The Sciences of the Artificial” from 1969 and spun out the likes of Design Thinking, Double Diamond, Human-Centered Design, etc. They are all notable attempts at adding nuance to design as a process, but don’t make it easy to true up to whatever high-level guidance the executive team is using. Which is a critical part of the product team’s role.

And of course we’ve also got the frameworks like Technology Readiness Levels (TRLs) and systems like Product Lifecycle Management (PLM). The use of TRLs allows for a structured approach to assess the maturity of a specific technology across a standardised scale, while the PLM system focuses on a structure and guidelines for how product information should be organised, how processes should flow, and how stakeholders should interact.

On paper this pairing should be enough for our purposes as product teams. The use of TRLs are not just the colloquial standard but the requisite measuring tape for government sovereign funds or academic grant committees. And who can argue with one hundred years of PLM development. The challenge is that we find ourselves faced with the same problems as the previous examples. A prescriptive management system will be overkill for a new Deep Tech startup or frontier technology venture navigating from a cold start. And a measure of readiness of a company’s technology is rarely limited to a single system. As soon as there are multiple lines of effort, the TRLs become components or sub-systems in the overall organisational effort.

Mapping any and all of the above to a roadmap is the product team’s job. It’s not uncommon for a Deep Tech startup to maintain both a technical roadmap and a product roadmap. And indeed we can take this framing (and all of the above) and simplify them down into a higher level view with virtually no extra work.

Perspective three: looking by evolutionary stage

When we talk about Deep Tech we are talking about a vector of sorts. There is an implied journey from a scientific discovery to some form of beneficial and commercial product reaching the market. But there’s a few steps between the eureka moment of an interesting discovery and the desired ability to scale and sell a product. Those steps are not just difficult technically but are dramatically different phases organisationally. If we break these down into their logical groupings, we get the four fundamental phases of Science, Technology, Engineering, and Product.

If we plot this out we get a readily simplified view that accommodates all the prior models without becoming too simple to mean anything. It fits well between TRLs (which detail the specific technology) and Gartner’s hype curves (which speak to markets) and whatever flavour of roadmap or horizons. As a simple device this framing helps us not just view but talk about organisational focus and ensure that the executive team is aware of the changes required phase to phase (which is not to be assumed given the likely academic background of many Deep Tech founders).

1. Science

Let’s take a company focused on the science first. A quantum computing company, for example, will be initially founded and staffed by scientists and researchers. The common pattern for this stage is to prove viablity of forther research by the meassure of impact and quantity of papers. Even IBM moved from measuring progress by patents to measuring by papers published. This doesn’t mean any “Product-ish” actitivies are totally absent, just that they must speak to the priority of this phase. For example if there is a marketing and brand effort, it’s with a mind to recruitment and fundraising efforts.

2. Technology

Viable scientific explorations will move into a technology focus with their own set of patterns and heuristics. The organisation may register independent of university of government origins, take on operational independence and legal compliance. Any funding will necessitate a governance overhead, and the distribution of that funding begins the collection of actions that will define the culture of the organisation by what it does (not what it says it is). The goal of “prove that this technology is viable” is a focused one, but one that needs to be mindful of the foundations it creates for the future commercial entity that will build upon it. Early hires are critical for this phase in both contribution, and their awareness of being able to evolve with the organisation or move aside when required.

3. Engineering

The engineering phase is one focused on ways to actually produce for a commercial scale. Even a proven technology may linger here for years, decades, or be discarded for either being unviable to build at scale or simply uneconomical. In our quantum example, the risk of increasing AI discoveries yeilding continual improvements in classic computing performance is one of potentially delaying the economic viablity of quantum-accelerated computing. But in turn the evolution of AI as a tool for synthetic data creation may yield breakthroughs in quantum circuit creation, workflows, algorithm development, etc. Multiple TRL efforts will require a strong control over both the technical and product roadmaps. This phase may require strong support from venture or sovereign capital given the expenditure and likely cycles of refactoring and recruitment.

4. Product

By this stage, the organisation is fundamentally different to the scientific phase, both in focus and team dynamics. It is not uncommon for some or all of the founders to have moved out of key executive roles and assumed board or advisory positions as specialist skills are required. All of the operational capacity of marketing, sales, support, as well as any required supply chain and vendor management, human resources and legal compliance need to be well-established. The focus on repeatable sales cycles and a path to profitability (or other value creation events) will be close to primary. Churn from non-commercial cultural dominance will have been completed by this phase.

Perspective four: all of the above

The key to this framing is that these are phases to consider the organisation’s current and future evolution through. In my opinion this is where Deep Tech startups are unique and these broad strokes are a simple but powerful way to ensure the organisation’s plans (and even brand story) for future possibilities is mindful of each era of change. Like any useful device, it exists not to be clever, but to surface discussion and make decisions.

In my own experience, this framing has helped surface an organisation’s hidden frictions between the back and front office teams (such as R&D and Sales), and forces the hard conversations to be had before they become culturally corrosive and require board intervention. And at the very least, this framing supports the awareness of strategic actions relevant to each phase of a company’s growth, and the need for continual reinvention of systems, processes, and even staffing given the shifting priorities and types of activities required.

What do you think about this “Science to Technology to Engineering to Product” phase idea? Share your thoughts in the comments.