AI Spend Splits Enterprise, Agency Buyers
An "enterprise divergence" is emerging in how companies pay for and implement AI, according to a recent analysis. Large enterprises are increasingly building their own custom AI layers, while smaller agencies rely on best-of-breed SaaS solutions, creating distinct market segments for software vendors.
- The shift to in-house AI development is significant, with 47% of enterprise AI solutions being built internally in 2024, a dramatic increase from just 20% in 2023. This move is driven by the need for deep customization and control over proprietary data, with the average custom AI project costing between $500,000 and $1 million. - Enterprise AI spending saw a massive sixfold increase in 2024, reaching $13.8 billion. Large enterprises allocate about 8-12% of their total IT budgets to AI initiatives, a stark contrast to smaller businesses whose entire IT budget may only be 3-6% of their revenue. - For agencies, AI adoption is widespread, with 91% actively using or exploring generative AI tools. The most common applications are brainstorming and content creation, though there are underutilized opportunities in workflow automation (44.4% adoption) and SEO strategy (31% adoption). - The procurement cycle for enterprise AI solutions is lengthy, averaging 6-18 months and involving multiple stakeholders and security reviews. In contrast, sales cycles for SMBs, including agencies, are much faster, typically lasting 1-3 months due to fewer decision-makers. - By building custom AI, enterprises treat it as a long-term capital expenditure (CapEx) for a predictable total cost of ownership. Agencies and SMBs prefer the operational expense (OpEx) model of SaaS subscriptions, which offers fixed monthly costs and avoids large upfront investments. - While OpenAI remains a major player, its enterprise market share fell from 50% to 34% in 2024. Competitors like Anthropic have doubled their market share, indicating a diversifying vendor landscape for both off-the-shelf and custom solutions. - Agencies are beginning to leverage specialized AI-powered "deep research" tools from companies like Perplexity to automate the process of gathering insights for client pitches and brand strategy, compressing hours of work into minutes. - The decision to build versus buy often hinges on whether AI is a core competency. For enterprises, a custom AI solution can be a key competitive differentiator, while agencies use off-the-shelf AI to enhance non-core functions like customer support or marketing automation.
