This post is written to crystallize my own thoughts, and isn't particularly groundbreaking. I have no particular experience in product management and appreciate thoughts/insights from anyone who does.
A concept that comes to mind for me quite a bit when evaluating early stage startups and more mature companies alike is what I call "the product curve." The product curve is a simple graph, with a measure of product efficacy on the y axis and a measure of resources expended on the x axis. Here's a sample one:
The graph shows a product that is comprised of various small, independent features, each of which is equally useful. As a result, the returns to future product investment are linear: more engineering hours = more utility for customers. As easy as it is to conceptualize, it is pretty tricky to find real world examples of products that act this way. In truth, the above chart is more representative of labor than a product (more hours spent digging a ditch = a longer ditch).
More commonly, the curve looks something like this next graph:
The early features of a product are extremely useful as the team building it focuses on high value use cases that are broadly applicable. In the case of CRM software, everyone needs a central contact record system, so building one makes the product far more useful. Over time, for various reasons, incremental hours become less productive. Some of these relate to the product-space itself. As development progresses, it is harder and harder to find new features that most users need and instead the team focuses on demands from smaller subsets. At the same time, as organizations grow and the user base expands, inertia and sclerosis get in the way, and each marginal hour spent becomes less effective. This is why massive companies with comparatively unlimited engineering budgets often have bad software- layers of bureaucracy and lack of focus slow down decision making and execution.
Of course, even the above graph is pretty silly (models are to be used, not believed). Product advances are discontinuous, and the most prominent discontinuity happens right at the beginning. In early development (labeled A on the graph below), the product isn't complete or in production- every line of code gets it closer to being effective for customers, but doesn't actually translate to efficacy in a measurable way. It is impossible to generate revenue from the product at this point. In this case, the graph ends up looking like this:
The nature of period "A" is important, because it tells us how long/how many resources it will take for a company to start generating some revenue. In the last 15 years or so, this period has compressed- with many startups getting to phase "B" on a very small amount of money (<$1m), and then iterating through various step functions (major new features, modules, etc.) until the product is built out. The lean startup methodology in action.
As the founder of a startup, the length of A (and the height of B) are crucially important variables that dictate how much money needs to be raised to de-risk the business in front of future financings, how to plan sales hiring, how to message to investors, etc.
For instance, at Matrix, we invest in many startups that are doing nuts and bolts research on really difficult problems in machine learning and materials science. For these startups, phase "A" is long, unpredictable and barren, and phase "B" is often awesome, with a meaningful technological advantage and game-changing potential. Startups like this are tricky for VCs- the "A" phase is expensive and doesn't necessarily come with lots of proofpoints (because no customers are using the product). We need to believe that the company can make enough technical progress to potentially raise additional capital before generating revenue, and that "B" is going to be amazing. There will always be exciting new SaaS applications or consumer apps that get to our phase with an MVP and some early customers, but as bottlenecks build up on the infrastructure side (from the slowing of Moore's law to batteries that can't keep our devices running) there is a steady supply of startups with this look:
It is worth talking about right half of the charts- the potential for long duration improvement in a product. It is a simple question of the derivative of the curve- can the product continue getting better for customers over time, and how fast? If it isn't getting better fast enough (and the company is profitable or clearly on a path to being profitable) competition will come and the marketplace will get crowded with lookalike products. In the chart above, I show basic research giving way to incremental improvement. However, that isn't necessarily a given. In the very worst case, the company ends up with a commodity product that is totally undifferentiated once a few other competitors get through their "A" phase, and pricing power erodes.
Lastly, here's what the chart looks like for network effect businesses (it isn't too different from the functions used to model growth of algae in a pond). Network effects businesses are often tricky to evaluate in the early days, because until the network effect takes off and goes exponential progress is slower. As the network builds, customer acquisition gets easier, not harder and the product gets exponentially more useful until each extra user either isn't that helpful (because liquidity is built) or because there aren't many more users to add.
Each chart in this post could merit a blog post on its own (and maybe I'll get them all written one day) but for me the key takeaway is that understanding where a startup is at on the product curve can have big implications for what its trajectory is like. Some of the more impressive startups have a product that is "fully fleshed out" despite limited funding, which I tend to view as a bad sign. On the other hand, networks effects businesses and basic research businesses often deserve a pass. I have tons of case studies in mind of both startups and mature companies where changes in the product curve where crucially important to understanding the story. As investors, we're all comfortable looking at bookings growth and other topline metrics to get a sense of how a go-to-market is functioning and this will always be more abstract/hypothetical than that, but there are big insights available from considering what the product curve for a given company or product might look like over time.