I was talking to a buddy who’s in corporate law and about the value of the services/products/deliverables we provide, and the associated responsibilities (He in law, I in engineering). A big difference in our fields is that, where engineering and project management can commonly be 8-12% of total cost for some of the projects I’ve worked on, the couple millions in legal fees on some of the deals (he works in M&A a lot) he helps shape rarely amount to more than 3% of the transaction value, although things can get blurred with success fees, and percentages increase for smaller deals. It was simply interesting to discuss and exchange thoughts on how the world works and how things are valued.
In discussing how things are valued we inevitably discussed software in the Tech sector. Software products enjoy a unique asymmetry between risk, cost, and potential reward that few other fields can match.
Software is Unique
Its digital nature, particularly on the cloud, enables easy distribution, near-instant updates/fixes. Iterations can be really fast on the basis of a Minimum Viable Product. The starting point is very easy, bugs or missing features aren’t a dead end. Even a very imperfect product can attract early adopters. The process of development, the feedback loop, is the product strategy.
Marginal costs are tiny. Once built, serving a new user can cost pennies. Margins can improve with scale, and ease of scalability can make for modest revenue per user translate into massive profits. I.E. a $5/month SaaS product with 100K users generates $6 M/year with low incremental overhead.
Compare this with building a factory, a power plant, or litigation; the upfront capital for a software MVP is minimal (mostly dev salaries and basic cloud services). Software has a low barrier to entry and supports rapid experimentation.
Of course, there are complexities, especially at scale, but the MVP approach really gets the ball rolling fast. This environment also makes for a lot of competition, and so the quality of a product eventually does matter. Security and trust are super important to safeguard user data, maintaining reliability at scale with low-latency needs investment and expertise, maintaining users with good UX is a must unless you’re a monopoly or so large you’re ubiquitous. There are also ethical and societal impacts from moving fast.
Software definitely shapes society and regulations (in the legal sense) are never quick to adapt. The law is a case-by-case exercise and can’t really outpage tech. Beyond profits, tech has had and will have other consequences: social media, facial recognition, hiring tools, financial systems, etc. Some good, some bad, but consequences and impact nonetheless
Contrasting with Law and Engineering
There isn’t really a minimum viable product for a legal argument, contract clause, or compliance strategy. An “imperfect” legal product, in the sense of its legal impact/factual accuracy/etc. is, at best, ineffective and, at worst, leads to reputational harm, lawsuits, fines, jail time, and can ruin lives. The cost of failure can be pretty severe, and relatively immediate/direct.
You can’t really iterate your way out of a legally binding error like you can a software bug after a commit, learning as you go. Law requires pretty deep, fairly constantly updated expertise in complex, ambiguous, and jurisdiction-specific context. Software engineers don’t really need jurisdiction-specific complex knowledge and can almost work from anywhere based on knowledge of tools and frameworks that is important and can be complex, but is much more accessible than the law.
Many software people are highly compensated given the high margins at play. In top law, the top lawyers are expensive and are a significant investment for the scale of the returns. Moreover, given the stakes and regulations around the practice, malpractice insurance, and compliance overhead is mandatory before one can even offer a service. So, profitability demands high fees or massive scale.
My lawyer friend has a few years under his belt and is actively participating in the drafting stage at this point, but he couldn’t go off on his own with his knowledge and make as much as he does know because his employer offers the scale to lever his salary.
As for engineering, well, physics doesn’t really negotiate. A bridge, grid, aeroplane, or medical device must work within parameters safely, and durably from the onset. Failure is catastrophic: I’ve written a few reports discussing mechanisms of injury and risks including “loss of life”. There’s also environmental damage and massive financial loss. Prototyping is essential and iterative, but the path to an analogous MVP is much longer, more expensive, and much more regulated than software.
The tangible nature of a (traditionally) engineered product demands and implies costs for physical components, manufacturing, supply chains, skilled labor (engineers, technicians, tradespeople), and heavy infrastructure involve massive capital expenditure. As a result, margins are inherently tighter. Moreover, in tight economic times (like at the time of writing), these big capital expenditures are “on hold” because of their risky nature. Which is why software is leading (by a long shot) in the S&P500.
Finally, like Law, Engineering has stringent regulatory standards and associated testing that must be met for some products to even make it to market, and in some places like where I am currently based, malpractice insurance is also mandatory to even offer engineering services. In my field, the bar for valuable is a matter of physics, safety, and the law, not just user adoption.
Aside from the vast differences in costs of creating and delivering a product: a certain amount of the costs in law or engineering (and in pharma, and other sectors) is specifically because of regulations. Now, should these regulations make the product costs more because it’s safer? Or should it trim margins? Probably a bit of both. In engineering, regulations tend to be written in blood, too often spilled in the name of greed.
Maybe one day we’ll have regulations for Tech and software as well. What’s for sure is that software is being injected into traditionally slower-moving fields, although the underlying physical or legal realities impose their own high bars.
While the core legal judgment remains high-stakes, aspects are being productised: document automation (NDAs, leases), legal research AI (Westlaw, LexisNexis), e-discovery tools, and compliance management software. These handle high-volume, lower-risk tasks, allowing lawyers to focus on high-value counsel. There is an MVP-like process for these tools.
In engineering, CAD/CAM, simulation software, rapid prototyping (3D printing), and IoT platforms incorporate software-like iteration into the physical engineering process, accelerating development and reducing costs before physical production. However, the final physical product still faces the ultimate high-stakes test.
Software is powerful and disruptive as a force. It operates by different economic rules, allowing for experimentation and value creation at a pace and scale impossible in most other fields, while simultaneously demanding mastery of its own unique set of complex, high-stakes challenges. Like all tools, it is to be wielded with care.