So we discussed some of the unique features of tech yesterday - more specifically as it pertains to software. We glossed over some financial characteristics that make tech so lucrative with generous bottom lines. This all begs the question: how come there are only a few options at scale? I don’t mean in terms of market cap., I mean enterprise software, data management systems, cloud providers: options at scale are concentrated in just a few large players.
More users lead to more value which begets more users. Big players have strong sticking power (first-to-snowball). Combine that with the high margins and capital leverage: bigger players can aggressively fund expansions and/or acquisitions. Because growth tends to be explosive, tech companies tend to attract “growth” minded investors. That’s Venture Capital’s main schtick after all. That’s also why valuations in that space can seem incredible for projects that have no income, just a (potentially) very good idea. To keep the numbers growing when they get bigger becomes increasingly difficult (law of large numbers) and acquisitions become the way to pump that % growth for bigger players. As well, it’s a way to eat up potential competition as well.
A Kill Zone
At scale, big incumbents (Meta, Google, Amazon, Microsoft, Apple, Oracle, etc.) can either copy fast and leverage their sticking power or purchase promising startups early. Big capital pools also enable talent buyouts as recruitment can be competitive and can eliminate competing projects. A lot of startups also rely on APIs, or app stores, cloud, and or services offered and controlled by these giants such that incumbents may also have informational and leverage advantages.
The landscape makes for lots of opportunity early, but few independent midsized successes. It is very much a “barbell” ecosystem - tiny startups and huge incumbents with fewer in between.
Antitrust
We’ve previously discussed how the law and policy are slow-moving copmared to tech. Antitrust, historically speaking, was focused on consumer prices, but for software and digital services which are often free or cheap, enforcement lags. The value lies a lot in data and attention by consumers, but laws weren’t written for that. Since the capital and landscape for the emergence of these companies was the US, we will give a cursory look at US laws.
Sherman Act (1890) and Clayton Act (1914)
Designed to combat industrial monopolies (rail, steel, oil) with tangible associated goods, clear pricing, visible market shares. These markets (to this day) have high marginal costs and physical supply chains. Monopoly clearly meant controlling production and setting prices. I’m repeating myself, but this contrasts with digital platforms operating with near-zero marginal costs and the ability for near instant scaling; often by offering a free product. Traditional measures of consumer harm (like high prices) simply don’t apply.
US antitrust laws evolved around a consumer welfare standard based on price and output. In tech, users/consumers often pay with data. Big companies can dominate markets with prices at or near zero, but still harm competitors and consumers. Chiefly via data extraction, reduction in privacy, lock-in effects (it’s really hard to change ERP or HR system once you’ve committed to a platform), and suppression of innovation (consider the incentives for VCs on multiple boards wanting to preserve an initial invesment). The current legal frameworks have a hard time classifying these harms as antitrust violations and, lobbying activities combat political will to do this.
Antitrust laws also require defining a relevant market to demonstrate dominance, but with platforms dipping into a bunch of interconnected services (search, ads, cloud, maps, AI, etc.) boundaries are blurred. Is Google’s “market” search engines?, browser?, mobile OS?, mobile phones?, cloud services?, all digital attention? Traditional frameworks set as tests for legal purposes can’t easily capture these ecosystem effects or cross-market leverage.
We’ve alluded to acquisitions being prevalent in tech along with the resulting consolidation. The Clayton Act aimed to prevent mergers that “substantially reduced competition,” but this is ill-suited for digital dynamics. Many tech acquisitions seemed harmless (Facebook-Instagram, Google-Youtube) - small startups with little to no revenue or market share. Much like traditional valuation based on physical assets and current revenue, the traditional law isn’t built to handle future potential competition (or value) theories or the strategic role of data aggregation. By the time harm is visible, the market’s already consolidated.
Enforcement of these laws, even in clear cases, typically is a matter of lengthy court cases. Rulings can take 5–10 years to come and in the tech sector that may be a completely different landscape or dynamic. There’s little to no deterrence: big firms just move fast and worry about litigation later.
In short, typical antitrust laws assume price-based, static competition in single-sided markets. Modern digital markets are data-driven, dynamic, multi-sided ecosystems, where dominance doesn’t show up as higher prices but as control over data, platforms, and access.
There seems to be recognition of there being an issue or potential for issue given recent shifts. Ignoring the current US administration on that front, but the FTC under Lina Khan and some high profile tech suits under the previous DOJ were clearly trying to establish precedents for future guardrails in Tech. The EU’s Digital Markets Acts, the UK’s Competition and Markets Authority have all moved to better frame consumer protections in the context of tech.
US - FTC & DOJ Under Lina Khan and Jonathan Kanter (DOJ antitrust div.), there was effort to reinterpret old laws to fit digital realities. Focus:
- Platform dominance and ecosystem control (e.g., Amazon, Google cases).
- Monopoly maintenance via data, network effects, and self-preferencing.
- Reassessing mergers — even small acquisitions — for their potential to eliminate nascent competition (e.g., Meta–Within case). The strategy aimed to expand understanding of consumer harm beyond price consideration to include loss of innovation, privacy, and choice.
EU - DMA
- Shifts from traditional case-by-case enforcement to ex-ante regulation — setting clear rules before harm occurs.
- Targets “gatekeepers” (big platforms like Google, Apple, Meta, Amazon) that control core digital services (search, app stores, social networks, messaging).
- Imposes Obligations:
- No self-preferencing (promotion of own products over rivals)
- Interoperability and data portability requirements (combats lock-in)
- Restrictions on combining user data across services The goal being clearly to keep digital markets open and contestable, not just punish abuses after the facts. Analytically this is a good approach to regulation as it aims to keep markets efficient. (As much as overegulation will skew dynamics and kills competition, power begets power and initial advantages in a too free market snowball so you need regulations to preserve competition)
UK - CMA
- Developed a Digital Markets Unit (DMU) to oversee firms with “strategic market status.”
- Focuses on data access, algorithmic transparency, and fair dealing between platforms and businesses.
- Like the DMA, the UK is leaning toward a proactive, rule-based approach, emphasizing intervention before entrenchment.
- The CMA’s tough stance on mergers (e.g., blocking Meta–Giphy, scrutinizing Microsoft–Activision) show a more forward-looking, risk-based model.
Overlapping Boards and Cross-Ownership?
It’s not rare for the same VC firms to fund multiple competitors. A priori this makes sense, the firm is aligned for a kind of product and looking for that unicorn, knowing that most of their investments won’t return a profit. In practice this positions VCs to soften competition or align incentives toward acquisitions.
In principle, section 8 of the Clayton Act restricts directors from sitting on competing boards, but enforcement has been lax. But wait, there’s more! Different persons sitting on competitor’s boards while working at the same VC (or PE or other fund/firm). Now, well-connected directors are a great value add, but that seems to come from anticompetitive reasons.
There’s a nice presentation on this subject in the Harvard Law School Forum on Corporate Governance called Overlapping Directors as a Competition Problem. Please review it, it is articulate, data-drive, and important.
Very big influential players (and their capital) kind of have the power to determine the fate of some of their investments by their sheer involvement. Having Sequoia or YC affiliated with your startup definitely helps the snowball. Beyond direct overlap, the informal networks that occur around these influential players can create a “gentleman’s agreement” culture that discourages true open competition.
If not explicit collusion, there is certainly the appearance, and incentives in place, of a concerted effort to reduce competition.
And so, Software is different. Its economic characteristics tend to naturally push toward oligopoly, even without a conspiracy in place. Scale and network effects yield a winner-takes-most environment. High-margins make for cash-rich incumbents that can outbid everyone. Regulatory lag makes for slow correction. Not to mention, it’s an ecosystem where almost everyone knows everyone else. There’s heavy interconnectedness in tech boards, “elite” schools, venture networks, etc.