A significant shift is occurring in how organizations manage their software stacks. For decades, the default business model was to renew existing software subscriptions, but the cost of building custom software has fallen sharply since 2024. This change is altering the math of software procurement, leading many teams to evaluate whether they should continue paying for bloated platforms or build their own replacements. This trend is particularly visible in the tech community, where sentiment analysis of public discussions reveals deep dissatisfaction with major incumbents.
This image illustrates the broader trend of organizations evaluating whether to build or buy software solutions.
Analysis of public sentiment highlights specific friction points that drive users toward alternatives. Major platforms like Adobe, Salesforce, and Jira are frequently criticized for 'lock-in,' 'clunky workflows,' and 'intrusive AI' features. For example, users of Adobe and Salesforce have expressed frustration with licensing restrictions and poor quality, while Jira users often cite manual task transfers and bloated workflows as primary reasons for seeking other options. These complaints suggest that for many users, the perceived value of these established tools is being eroded by poor user experience and restrictive pricing models.
However, the transition from 'buying' to 'building' is rarely a complete replacement of a core system. Research indicates that 'build part of it' is the most common strategy; many teams choose to keep their core infrastructure while replacing specific, high-cost features or seats with custom-built tools. While 32% of organizations using AI report skipping software purchases to build custom features instead, the actual financial return is complex. Many documented builds currently report no immediate cost savings, highlighting a significant trade-off: while custom tools can eliminate subscription fees, they introduce the ongoing maintenance and complexity of internal software management. Because realized build payback is difficult to measure, the decision to build often hinges on whether the specific pain points of a SaaS tool outweigh the long-term overhead of maintaining a custom solution.