What Gartner's 2027 software forecast means for buyers

Gartner expects worldwide software spending to grow 14.6% to $1.68 trillion in 2027, according to SaaStr’s breakdown of Gartner’s October 21 IT spending forecast. That sounds like booming demand for apps. The detail that matters for you is price. Gartner’s Lovelock has said software is getting more expensive because of GenAI. SaaStr reads a meaningful share of that 14.6% as the same seats at higher prices, plus AI add-ons and usage meters.

The background you need

What a Gartner forecast is. Gartner is a research firm. Large companies pay it for advice on what to buy. Each quarter it publishes estimates of how much the world will spend on IT, split into categories such as data center systems, devices, IT services and software. These numbers are estimates of buyer spending. They are not a count of users, licenses or apps.

Why spending is not the same as usage. A spending figure mixes two things: how much you buy and what each unit costs. Think of a grocery bill. If it goes up 15%, you might be buying more food, or the same food at higher prices. The total alone does not tell you which. The analogy breaks in one place. Groceries get pricier for reasons outside the store’s control. A software vendor that adds an AI feature and raises the price is also changing the product, so part of the increase buys something new, whether you wanted it or not.

The pricing terms the report assumes.

• A seat is one paid user license. Most SaaS tools you pay for are priced per seat, per month.
• A SKU (stock keeping unit) is a distinct item a vendor sells. An AI SKU or AI add-on is a separately priced AI feature bolted onto a product you already pay for.
• A usage meter charges by consumption, such as API calls or tokens. Tokens are the small chunks of text that AI models read and write. Model providers bill by them.
• A system of record is the tool that holds your source-of-truth data, such as a CRM, an HR system or an accounting package. It is hard to replace, which gives its vendor pricing power.
• FinOps is the practice of tracking and controlling cloud spending in close to real time instead of discovering it on a monthly invoice.
• An AI app dev platform is a tool for building your own AI-powered applications instead of buying a finished product.

Agents. An AI agent is software that takes actions on its own, often by calling other services’ APIs. Unlike a person, it can make calls around the clock. On a usage meter, that changes the math.

What actually changed

Some of this is new, and some was already true.

New in this forecast, according to SaaStr’s reading of Gartner’s numbers:

• Worldwide IT spending is forecast at $7.35 trillion in 2027, up 15%, crossing $7 trillion for the first time.
• Data center systems add $487 billion in 2027, about half of the total increase.
• Software adds $215 billion in one year, second only to data centers.
• Gartner’s 2026 software estimate now stands at $1,464 billion, up $31 billion from the $1,433 billion it forecast a year ago.
• In its September AI forecast, Gartner raised 2026 growth for GenAI models from 110% to 117%, and for AI application development platforms from 28% to 39%.
• New predictions say Global 500 firms will move to real-time AI FinOps by 2028, and that by 2029 four in five new apps will be built for use under a year.

Already true. Software’s share of IT spending is roughly flat, at 22.9% in 2026 and 22.8% in 2027, by SaaStr’s arithmetic. Data centers are taking the share gains, rising from 13.8% to 18.6%. And the price point is not new. Lovelock said it last year: “The cost of software is going up and both the cost of features and functionality is going up as well thanks to GenAI.”

Gartner’s read, per SaaStr, is that enterprises mostly buy simpler embedded AI features from vendors they already use. The AI SKU on the system of record gets bought first.

What it means for developers and small teams

The price lever matters more than the headline growth rate. Here is why. The 14.6% is an average across all software. SaaStr points out that GenAI models above 100% and AI platforms at 39% pull that average up. That leaves many standard business apps growing well below it. If ordinary apps are not seeing much new demand, rising prices on existing customers are one of the few ways their revenue grows. You are the existing customer.

Concretely, expect three things at renewal:

1. Higher per-seat prices for tools you already use, often justified by AI features.
2. AI add-ons sold as a separate line item, or folded into a higher tier so you cannot easily opt out.
3. Usage meters on AI features, billed by calls or tokens on top of your seats.

The third one deserves the most attention. SaaStr notes that agents calling vendor APIs around the clock can multiply a usage bill fast. It also says that once a FinOps team audits those calls, a lot of them turn out to be redundant. Gartner expects large companies to build that audit habit by 2028. A small team has no FinOps department, so you need a lighter version: set spending caps where vendors offer them, and check usage dashboards monthly.

The other shift is build versus buy. With AI app dev platforms growing fast and Gartner predicting short-lived apps, SaaStr argues the departmental tool, internal tracker or niche workflow app now gets built in an afternoon by whoever needs it. For you, that means some small single-purpose subscriptions may be replaceable. The spend does not disappear, though. It moves to the platform and its usage bill.

What we don’t know yet

• How much of the growth is price. SaaStr calls it a meaningful share but gives no number. Its summary of Gartner’s forecast does not split price from volume.
• How the $215 billion divides up. SaaStr says a big slice goes to a short list of model providers, but does not say how big.
• The 2027 rates for models and platforms. The 117% and 39% figures are 2026 growth rates.
• Whether this applies to small buyers. Gartner’s predictions focus on large enterprises such as the Global 500. Whether small-team plans see the same increases is not covered.
• Whether audits actually cut usage revenue. SaaStr’s view that redundant calls get removed is its own argument, not a Gartner figure.