Companies build in-house AI software as pressure grows on enterprise SaaS vendors
Businesses are increasingly using generative AI to create internal software tools that they might otherwise buy from enterprise vendors. The shift is opening a lower-cost path for custom features in areas such as HR, sales, and finance, while raising questions about future add-on revenue for software providers.
Highlights
- McKinsey's survey showed about one-third of 1,719 organizations avoided buying at least one software product or feature by building AI-powered internal tools.
- West Monroe estimates saving over $300,000 annually by developing payroll and workforce insight software internally with ChatGPT and Codex instead of purchasing from vendors.
- Gartner still forecasts nearly 16% global software spending growth to $1.47 trillion in 2024 despite risks to vendor expansion revenue from custom in-house AI tools.
AI-driven internal tools reshape software buying
As first reported by Business Insider, companies are using tools from OpenAI, Microsoft, and Anthropic to develop internal applications instead of paying for some software products or extra features from established vendors such as Workday, SAP, and Salesforce. The article says this does not amount to a broad replacement of full enterprise platforms, but it can reduce the incremental sales those providers depend on for expansion revenue.West Monroe, a Chicago-based consultancy, said it built payroll-checking software and a tool that gives managers workforce insights for its roughly 2,000 employees. Tanya Moore, the company's chief people officer, estimated buying those capabilities from a vendor would have cost more than $300,000 a year, so employees created the tools earlier this year with ChatGPT and Codex.
McKinsey found in a recent survey of 1,719 executives and managers that about a third of organizations decided against buying at least one software product or feature because they could build it with AI tools. Kyle Lagunas, an HR-technology analyst, said the trend is likely to erode part of the traditional vendor "walled garden" by allowing HR teams to self-serve more of their needs.
Omega Venture Partners framed the risk more sharply in an August report, saying the customer may now be the vendor's most dangerous competitor. Even so, Gartner forecasts global software spending will still rise by nearly 16% this year to $1.47 trillion, suggesting the broader market remains on firm footing.
Cost savings come with operational and compliance risks
Companies cited a widening range of practical uses for internally built AI tools. Spotify said its HR Bot, launched in November 2025, answers employee questions using the company's handbook and helps staff check matters such as remaining vacation days and parental leave policies, while Twilio said its in-house sales tool Jarvis has reached about 80% adoption across sales teams and shortened the sales-deal lifecycle by 54% for users versus non-users.Twilio CEO Khozema Shipchandler said the company would probably not adopt a similar outside product if one were available because it has already built its own version. Gray, a design and construction firm in Lexington, Kentucky, expects to avoid paying $1 million to a vendor by starting later this year on an internally developed financial forecasting tool using Microsoft's Copilot and Anthropic's Claude.
Still, maintaining even small in-house tools can be difficult because companies must fix bugs, issue updates, handle feature requests, and keep systems reliable as usage expands. Workday chief technology officer Gabe Monroy warned in a June blog post that errors by AI agents in compliance or pay approvals could lead to audits, regulatory scrutiny, or lawsuits.
Large vendors also argue that years of product development and accumulated data remain a competitive advantage that large language models alone cannot replicate. Moore said West Monroe has already built eight tools instead of buying vendor offerings, with each undergoing months of testing and periodic reviews, but she added that not every business has the expertise or risk tolerance to follow the same approach.
AI-driven fundraising across Asia-Pacific capital markets highlighted how companies are ramping up equity and convertible bond issuance to finance AI-related expansion in chips, data centers, and power infrastructure. Our earlier article noted that this AI-led investment wave is also prompting greater attention to governance and risk controls, as wider use of automated tools can increase exposure to fraud, manipulation, and opaque decision-making.