The TMT sub-sector map

Why the map matters more here than in other groups

Every coverage group has sub-sectors. Healthcare has pharma, med tech, and providers; industrials has aerospace, building products, and diversified manufacturing. What makes the TMT map unusual is how little the sub-sectors resemble each other economically. A vertical software company and a wireless carrier both get pitched by "TMT bankers," but one is asset-light and grows by spending on customer acquisition, and the other is one of the most capital-intensive businesses in the public markets. An interviewer who hears you describe TMT as a single business model, rather than seven distinct ones sharing an industry label, will assume you have not spent real time with the sector. This article is the map: seven sub-sectors, how each one makes money, and how each one gets valued.

The seven sub-sectors

Software

Software companies sell access to a product, usually under a subscription model (software as a service, or SaaS), rather than a one-time perpetual license. The economics that matter: high gross margins because the marginal cost of serving one more customer is low, heavy reinvestment of that gross profit into sales and marketing to acquire new customers, and revenue that recurs, meaning a customer who signs up keeps paying (and ideally paying more, through upsells) for years rather than making a single purchase.

Because reported earnings are often suppressed on purpose by growth investment, software companies get valued primarily on revenue or annual recurring revenue multiples rather than EBITDA multiples, with the exception of mature, already-profitable software businesses. The full mechanics are in how software companies are valued, and the specific metrics investors use to judge quality within that framework (retention, CAC payback, the magic number) are in the SaaS metrics bankers actually use.

Internet and marketplaces

Internet businesses monetize either a transaction (a marketplace connecting buyers and sellers, taking a cut of each transaction, called a take rate) or attention (an advertising-supported platform, monetizing user engagement through ads). Both business models depend heavily on network effects: a marketplace gets more valuable to buyers as more sellers join, and more valuable to sellers as more buyers join, which is what makes a market-leading marketplace so hard to displace once it reaches scale. Advertising platforms depend on a related but distinct effect, where more users generate more engagement data, which improves ad targeting, which attracts more advertisers, which funds a better product.

Marketplaces are typically valued on a multiple of revenue or gross profit rather than a multiple of gross merchandise value (GMV, the total dollar value of transactions flowing through the platform), since GMV overstates the actual revenue a marketplace keeps. The full breakdown of take rates, GMV, and network effect types is in the internet marketplaces and network effects article.

Semiconductors

Semiconductor companies design and manufacture the chips that power essentially every other category in this map. The sub-sector splits into three business models: fabless companies design chips but outsource manufacturing to a third-party foundry, foundries manufacture chips designed by others without designing their own, and integrated device manufacturers (IDMs) do both design and manufacturing in-house. Semiconductor earnings are notoriously cyclical, driven by the long lead time between deciding to add manufacturing capacity and that capacity actually coming online, which routinely causes the industry to overshoot in both directions.

Because of that cyclicality, semis get valued with a mean-reversion lens: multiples often compress right when earnings peak (because the market is already pricing in the next downturn) and expand when earnings trough. The full explanation, along with the value chain from design through packaging, is in semiconductors: the cycle and the value chain.

Hardware

Hardware covers physical technology products: computers, phones, networking equipment, and similar devices. Margins sit well below software (there is a real, meaningful cost to manufacture each additional unit) but the business is typically less capital intensive than semiconductor manufacturing, since most hardware companies outsource actual production to contract manufacturers rather than owning factories themselves. A hardware company's valuation often hinges on whether it has a services or software attach rate, meaning a recurring revenue stream layered on top of the hardware sale, since that attached revenue tends to command a higher multiple than the hardware itself and can justify valuing the two pieces separately in a sum-of-parts analysis.

IT services

IT services companies sell expertise and labor rather than a product: systems integration (helping a client implement complex software or infrastructure), staff augmentation (providing contract technology workers), managed services (running a client's IT infrastructure on an ongoing contract), and business process outsourcing (running an entire business function, like customer support, on a client's behalf). Margins are thin because the primary cost is people, and the business scales roughly linearly with headcount rather than exponentially the way software does. Consolidation through acquisition is a common growth strategy in this sub-sector precisely because organic growth is capped by how fast a company can hire and train qualified staff. Full detail is in the IT services and outsourcing business models article.

Media

Media companies produce or license content and monetize it through subscriptions, advertising, or licensing fees to other platforms. The distinctive accounting wrinkle bankers need to know is that content costs get capitalized and amortized over the period the content is expected to generate revenue, rather than expensed immediately, which means a media company's reported earnings in any given period depend heavily on assumptions about how long a piece of content will keep earning. Subscriber counts and churn are the operating metrics that matter most for subscription media, echoing the retention logic in software but applied to a very different cost structure.

Telecom

Telecom carriers build and operate networks (wireless, cable, or fixed-line) and sell access to those networks as a recurring subscription. The defining trait of telecom economics is enormous, continuous capital expenditure: carriers spend a large share of revenue every year on network maintenance and upgrades just to hold their competitive position, not even to grow. That capital intensity, paired with subscriber revenue that is unusually stable and contractual, is exactly why telecom carriers can and do carry far more leverage than a software or internet company with the same EBITDA. The interplay between media and telecom economics, including how the two sub-sectors sometimes sit inside the same conglomerate, is covered in the media and telecom economics for bankers article.

Where the lines blur

The map above is a useful default, but real companies straddle categories constantly, and interviewers sometimes hand you a company on purpose that does not fit cleanly into one row, just to see whether you notice. A hardware company that sells a physical device but layers a subscription service on top (extended support, cloud connectivity, ongoing software updates) is really a hybrid of the hardware and software rows, and a careful analyst would value the two revenue streams separately rather than forcing a single multiple onto the whole business. A large diversified technology company might own a cloud software division, a hardware division, and a services division simultaneously, which is exactly the setup that produces the sum-of-parts valuation gaps discussed in TMT deal structures, because the market tends to apply one blended multiple to the whole conglomerate even though its pieces deserve very different multiples individually.

Cybersecurity is a good example of a category that sits across two rows on this map rather than inside one. Some cybersecurity companies are structured and priced almost exactly like enterprise software, subscription revenue, high gross margin, valued on ARR. Others, particularly companies that sell security services delivered by human analysts monitoring a client's systems around the clock, look more like IT services, priced on EBITDA margins and headcount economics. Media and telecom blur together too: a diversified media and telecom conglomerate might own both a content studio and a distribution network, and historically some of the largest media and telecom mergers were explicitly attempts to control both the content and the pipe it travels through, with mixed results. Naming which row (or rows) a company actually belongs in, out loud, before you commit to a valuation approach is a habit that reads as real sector judgment rather than pattern matching off a memorized list.

The map at a glance

Using the map in an interview

The practical use of this map is defensive as much as offensive: it stops you from answering a valuation or deal question with a framework borrowed from the wrong sub-sector. If an interviewer describes a company and you cannot immediately place it on this map, ask a clarifying question about how it makes money before answering, the same way you would clarify unfamiliar terms in any technical question. Naming the sub-sector correctly, out loud, before you launch into a valuation approach signals exactly the kind of structured thinking interviewers are grading for, and it sets up whichever deep-dive article on this list you go to next, whether that is pitching a tech stock or one of the sub-sector economics pages above.

Practice question

A private equity fund shows you two companies with identical revenue and EBITDA: one is an enterprise software company, the other is a regional telecom carrier. Would you value them the same way?

No, and the fact that they have identical revenue and EBITDA today barely matters, because the two businesses sit in completely different places on the TMT map. The software company likely holds that EBITDA level either because it is mature and has already gone through its heavy growth-investment phase, or because it has intentionally suppressed earnings to fund sales and marketing, so I would want to know its growth rate and retention before picking a multiple, and I would sanity-check EV/EBITDA against a revenue or ARR multiple. The telecom carrier's EBITDA is more likely to represent a stable, recurring level of profitability, but I would look past the reported EBITDA to free cash flow after capex, since telecom requires continuous, heavy network investment that software doesn't. I would also expect the telecom carrier to support meaningfully more leverage than the software company for the same EBITDA, because its subscriber revenue is more contractual and predictable, and a lender has more comfort underwriting against that stability. Same starting numbers, different underlying businesses, so the valuation approach and the leverage assumptions both have to flex to match.

What the interviewer is listening for: Whether you resist the trap of treating "same EBITDA" as "same valuation logic," and whether you can name the specific reasons (growth stage, capital intensity, revenue durability) that drive the difference rather than gesturing vaguely at "different industries."