The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Some adtech companies can afford the infrastructure behind the AI boom; there is no evidence that the sector as a whole can. The answer depends on each company’s revenue base, how fully it uses capacity, whether it owns or rents that capacity, and whether AI measurably increases revenue or reduces costs. Advertising-market growth and big infrastructure budgets show the scale of the opportunity, not that AI investments are earning adequate returns.
What “afford” means for an adtech company
Affordability is not simply whether a company can finance more chips or cloud capacity. A sustainable investment has to cover the full cost of operating that capacity and produce a business benefit large enough to justify the expense. That benefit might be better ad performance or yield, more engagement and revenue, a paid AI product, or lower operating costs.
As an Amazon Associate I earn from qualifying purchases.
Public figures do not yet offer a standardized, comparable measure of AI infrastructure returns across adtech. Company disclosures also vary in what they count: capital expenditure is not the same as total infrastructure cost, and a stated infrastructure budget is not necessarily an AI-only budget. The most useful assessment is therefore company-specific.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Advertising growth creates opportunity, not proof of a return
The Interactive Advertising Bureau forecasts 9.5% growth in total U.S. ad spend in 2026. That is a forecast for the U.S. advertising market—not a forecast for adtech-company profits, AI revenue, or returns on infrastructure investment. A growing market may give companies more room to invest, but each company still has to show how its AI spending supports its own economics.
#1 Best Overall
Infrastructure costs go well beyond accelerator purchases
AI capacity carries both capital and operating costs. Alphabet identifies depreciation, energy, equipment, and network capacity as costs associated with its technical infrastructure. It also says serving AI offerings requires more compute than its historical offerings. A company that counts only the purchase price of accelerators will therefore miss important costs of delivering AI services.
Renting capacity adds another dimension. Meta’s 2026 outlook includes third-party cloud spending alongside depreciation and infrastructure operating costs. Comparing a build-versus-rent decision requires looking at the full cost of each route, not simply comparing a hardware bill with a cloud invoice.
What major companies’ spending outlooks do—and do not—show
Meta’s guidance is company-specific
Meta Platforms forecast $115–135 billion in capital expenditures for 2026. The range includes investment for AI efforts and Meta’s core business, so it should not be described as AI-only spending. The company said most expected expense growth in 2026 would come from infrastructure, including third-party cloud spend, depreciation, and infrastructure operating costs.
Meta’s CFO also said: “Despite the meaningful step up in infrastructure investment, in 2026 we expect to deliver operating income that is above 2025 operating income.” That is company guidance, not a realized result or proof that the investment has paid off. It does show how a large platform frames its spending against its own operating-income outlook; it does not establish that smaller adtech businesses can reproduce those economics.
Hyperscaler budgets show scale, not adtech returns
S&P Global Market Intelligence reported that Alphabet, Amazon, and Microsoft collectively projected $495 billion in 2026 capital expenditure on fourth-quarter 2025 calls, up 61% from 2025. The total includes technical infrastructure beyond AI and applies to those three companies, not adtech as a whole. The scale underscores why investors are asking about returns; it does not answer whether adtech companies’ own spending is affordable.
Inference makes utilization and workload mix central
Training models is only one part of AI infrastructure demand. Gartner forecast $42.276 billion in global AI-optimized infrastructure-as-a-service (IaaS) spending in 2026, up 96.4%. Within that forecast, spending on inference was projected at $23.3 billion, compared with $19 billion for training.
Rank #3
These are forecasts for the global AI-optimized IaaS market, not measured adtech-company spending or adtech’s share of the market. They nevertheless point to a practical test for a company considering capacity: how consistently will that capacity serve inference and other paying or business-critical workloads? Capacity that is expensive but underused is harder to justify than capacity that is reliably used to support valuable services.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteAI can change revenue and margins as well as costs
Infrastructure investment is easier to evaluate when a company identifies the mechanism for earning it back. Relevant evidence could include improved ad performance or yield, increased engagement that supports revenue, paid AI services, or demonstrable operating-cost reductions. Spending announcements and market-growth forecasts alone do not establish any of those outcomes.
Alphabet cautions that AI features such as AI Overviews and AI Mode in Search may monetize differently from its historical offerings, affecting revenue growth rates and margin trends. That matters because more compute-intensive services can change both the cost base and the way revenue is generated. Alphabet also describes efforts to improve efficiency through model design and its TPU/GPU infrastructure, illustrating why utilization and technical efficiency belong in the affordability calculation alongside revenue.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.A practical framework for judging affordability
- Start with the company’s revenue base. Separate advertising revenue from cloud, subscription, or other revenue where disclosures permit. Keep geography and reporting periods consistent; market-wide ad-spend growth is context, not a company-level return measure.
- Count the full cost of capacity. Consider capital expenditure, depreciation, energy, equipment, networks, facility operations, and rented cloud capacity rather than treating accelerator purchases as the whole bill.
- Check utilization and workload mix. Ask whether capacity will be used consistently for inference, training, or other business-critical work, and whether those workloads support revenue or essential operations.
- Name the return mechanism. Look for evidence of stronger ad performance or yield, more engagement or revenue, paid products, or lower operating costs. Do not infer returns from spending levels alone.
- Separate guidance from results. Label company outlooks, market forecasts, and scenario estimates accurately. Compare them with realized capital expenditure, margins, utilization, and monetization as results become available.
Why the headline infrastructure numbers cannot settle the question
Long-range estimates are useful for showing the scale of the investment challenge, but they are not adtech funding requirements. Bain estimates, as reported by TechRadar, that AI infrastructure spending could reach $1.5 trillion annually by 2031. Under the assumption that infrastructure equals 25% of sales, that would imply roughly $6 trillion in annual revenue. The revenue figure is scenario arithmetic based on that assumption—not an independent consensus forecast, a regulatory threshold, or a prediction of adtech revenue.
The figures used to discuss the boom cover different markets, companies, cost definitions, and periods. They should not be added together or treated as one measure of adtech spending. They establish that infrastructure investment is accelerating and that returns are under scrutiny; they do not establish a sector-wide affordability verdict.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
