“Asset-Light” Is Part of the Cycle, Not Just an Industry Trait: Why SaaS Doesn’t Make Money

As Lotte Chemical enters a prolonged downturn in the chemical industry, it is said to be accelerating its shift to a capital-light/asset-light structure. On the balance sheet, a company’s assets consist of tangible assets with physical substance, such as plants and production facilities, and intangible assets without it, such as goodwill, copyrights, and software. The “asset-light” strategy that has been in the news lately refers to shedding tangible assets like real estate or factories that have book value but tie up cash.

When many investors take a checkbox-style approach to public or private investments, the first thing they look at is capital efficiency.[1] All else equal, a company that consistently generates cash (free cash flow) is better than one that only consumes it, and a company that can generate more cash over time without large additional outlays is more attractive. This is good for investors and for founders alike, because companies that don’t generate cash have to raise money through additional equity issuance (diluting the founder’s stake) or debt (diluting control if risks materialize).

Take Samsung Electronics, for example. It continually pours large sums into tangible assets such as semiconductor fabrication plants and production equipment. Even after a plant is built, it must periodically upgrade its facilities in the fast-changing semiconductor market. Microsoft, on the other hand, runs primarily on software development and maintenance, so it doesn’t need to make massive expenditures on factories or hardware production lines. As a result, even if they earn the same level of profit, Samsung must keep spending cash to maintain those profits (capital-intensive), whereas Microsoft does not (capital-light).

One reason investors like capital-light businesses is the backdrop that made software companies so eye-catching to investors in the early 2000s. As hardware costs plummeted and open-source solutions proliferated, the initial capital required to start a company kept falling. In theory, that trend has continued to this day. If you know AWS, one major software stack, and how to use ChatGPT or code auto-completion tools, you can build software targeting a specific niche market in a matter of days and for just a few hundred dollars.

The problem is that once anyone can build a new product, the act of “building the product” itself quickly gets commoditized and standardized upward. If you Google “CRM for religious organizations,” you’ll see a flood of companies already advertising to that same niche. The very fact that they’re running ads is telling. Even in an extremely narrow niche, you still need upfront spending—like ad budgets—to pull potential customers into free trials or paid plans.

In other words, even if the cost of building the product itself isn’t that high, once the “cost of acquiring users” rises, that becomes the point where capital investment is required. In reality, there are countless software companies whose core product isn’t that hard to build, but whose main expenses are sales and marketing.

What’s more, for many software companies the ratio of R&D (research and development) spending to revenue doesn’t actually decline much over long periods. Why is that? Take the enterprise market as an example. It’s not that hard to clone an electronic signature service like DocuSign. All you need is to insert a digital signature into a PDF and email the result to all signers. But in practice, that’s only a small fraction of what customers want. You need integrations that continuously connect everything: a repository that automatically stores signed documents, pathways that generate contracts, and rules for who gets notified and how tracking works. At first, it may seem sufficient to offer just “Slack integration for sales-team alerts and contract tracking, Dropbox integration for automatic backup storage, and Salesforce integration to generate inline contracts and share them with the right owner.”

But the problem is that Slack is not a tool with 100% market share. How do you handle companies that use Teams, or email-based organizations that don’t use chat apps at all? And since there are countless CRM products besides Salesforce, over time even a “missionary-only CRM” ends up in a situation where it has to integrate with all sorts of solutions.

In the process of continually expanding these integrations—even if software itself is not capital-intensive and has high margins—new costs keep emerging:

From an accounting standpoint, most of this shows up as operating expenses, with a small portion classified as cost of goods sold (COGS). But from an economic perspective, the act of “building, selling, and retaining customers for a complex product” is the process of creating an intangible asset. This asset has a finite life, just like physical facilities or airplanes. It requires maintenance costs (cash outlays) and amortization (non-cash expenses). And to prevent that asset—and the business built on top of it—from going all the way to zero, you need to spend at least as much cash as the amount represented by its amortization.

Of course, the software industry can still be attractive. Margins are high, pivots are quick, and there are opportunities to enjoy economies of scale such as network effects and lock-in. But as time passes, these advantages are offset by “new, similar companies entering the market and competing to attract customers.” Unless you have exceptional management, strong government protection, or extraordinary luck, the “economic financial statements (earnings structure)” of software companies eventually converge toward those of capital-intensive industries like airlines.[2] In reality, there are airlines that have generated high returns over long periods thanks to good management, strong government backing, luck, or some combination of these.

To understand the early days of any industry, you need to look at “how much it will grow, what impact it will have, and what its unique capital cycle looks like.” But over the long term, you ultimately have to understand “why this industry looks different right now” in order to see the future in which, “after enough time has passed, it ends up looking similar to other industries.”

[1] Capital efficiency is often used as a key metric when assessing whether a company can maintain a high ROI (return on investment).
[2] Even if this doesn’t show up the same way in the accounting books, as competition intensifies and the market matures, intangible assets that require “ongoing investment and maintenance” become the core of actual costs, causing the structure to converge toward that of industries that own physical assets like aircraft or factories.

Even just understanding the above is enough to see why investing in companies that build subscription software (SaaS) is no longer as lucrative as it once was. But since Korean VCs (venture capital firms) are notorious for being bad at investing and behind the times, it seems a more concrete explanation is needed, so I’ve laid it out here.

In the past, thanks to AWS and open-source solutions, you could quickly launch a simple MVP, and SaaS would turn into money right away. Today, however, the market is flooded with similar products, and the features customers want have become more complex. So the old “move fast and break things” strategy no longer works. While it’s still cheap to build an initial product, you keep burning cash on feature expansion, integrations, and marketing.

There has long been a bias that “software is not capital-intensive,” but in reality, the cost of maintaining and servicing intangible assets keeps rising. You have to integrate with all kinds of third parties and APIs, and the customer support team has to deal with unexpected bugs and edge cases. All of this shows up as operating expenses, so even with monthly subscription fees coming in, there’s a high chance you’re still in the red.

These days, many people are looking for subscription products, but there are simply too many competitors. In the end, you end up thinking, “we only grow users if we pour more money into search and social ads.” You lure new sign-ups with coupons, promotions, and referral rewards, but before you recoup those costs, another competitor shows up and entices your customers away. Even if subscription revenue seems to be growing steadily, actual profits are weak because of CAC (customer acquisition cost).

U.S. VCs used to joke that 75% of the money they invest in startups ends up going to Google and Meta ads, but the sad reality is that it’s not really a joke anymore.

When SaaS first emerged, it strongly felt like “a revolution in enterprise software.” But now that it has solidified into a format that “anyone can build,” product quality has converged upward. On top of that, there isn’t much feature differentiation between similar SaaS offerings, so once price competition begins, margins get slashed.

For these reasons, investors are concluding that SaaS startup investments no longer generate the kind of returns they used to. The product itself may not look capital-intensive, but over the long run, its underlying nature is revealed: capital is consumed by “maintenance, operations, and marketing.” In other words, you can no longer invest just because “it’s software, so it’ll be fine.” Only SaaS businesses with truly differentiated technology or strong market dominance are likely to survive.