"With AI making development so much easier, discussions came up internally about whether we should stop outsourcing and switch entirely to in-house development. But since we haven't been able to hire software engineers in the first place, the question remains: who is actually going to build it?"—we received this inquiry from a business owner evaluating a new operational system for a company of several dozen employees. As word spreads that AI enables development with lean teams, SMBs face the pressing question: should we build in-house?
In July 2026, research firm Gartner projected that by 2029, 60% of organizations will fully deploy smaller software engineering teams (up from 15% in 2026). This reflects team restructuring designed to capitalize on the complementary strengths of humans and AI. This shift directly impacts SMB decisions between building in-house and outsourcing. This article explores how to translate Gartner's prediction into practical business decision-making.
What "smaller teams" really means
What Gartner refers to as "smaller teams" is not merely about headcounts. It describes restructuring development—where roles were traditionally segmented into system designers, builders, and testers—into a model where compact teams leverage AI assistance to handle everything from domain comprehension and architectural design to AI agent oversight across the entire lifecycle. Teams of four to five, or even two to three members, are appearing today, and this pattern is projected to become widespread.
The key takeaway for SMBs is the reality that smaller teams can now manage a broader scope of development than ever before. Initiatives that once required maintaining dedicated departments are coming within reach through lean staffing combined with AI. A comprehensive view of integrating AI agents into operations is discussed in our article on operational utilization of AI agents. The trend of expanding in-house development through no-code and AI is also explored in in-house development in the era of AI agents.
The barriers of recruitment and talent development have not disappeared
At the same time, the same forecast issues a critical warning that must not be overlooked. Gartner explicitly cautions against using this trend as an excuse to cut entry-level hiring. Organizations that restrict junior recruitment assuming AI can replace entry-level responsibilities will severely deplete their talent pipelines by 2028. Institutional knowledge transfer stalls, internal talent fails to develop, and companies end up reliant solely on scarce, expensive senior professionals—a danger that strikes SMBs especially hard.
In short, being able to run smaller teams with AI does not mean people are unnecessary. Even when teams are compact, you must recruit, train, and maintain individuals capable of leveraging AI effectively. The question raised by the business owner above—"who is going to do it?"—hits the nail on the head. AI takes over parts of development, but it does not take over hiring and nurturing the talent who direct it.
Re-evaluating in-housing vs. outsourcing for SMBs
Weighing these two factors—that lean teams can accomplish more, yet hiring and training hurdles remain—reveals that the choices for SMBs are no longer a simple binary.
| Approach | Best suited for | Key considerations |
|---|---|---|
| In-house focus | Core business systems require continuous iteration, and the company has bandwidth to recruit and develop staff | Bears hiring, training, and turnover risks internally; watch out for knowledge silos |
| Outsource focus | Development volume does not justify dedicated staff, but quality standards cannot be compromised | Complete delegation often turns systems into unmaintainable black boxes |
| Collaborative partnership (hybrid) | The organization wants to build internal capabilities, but transitioning everything at once is unfeasible | Clearly define the boundary between internal responsibilities and external support |
For many SMBs, the realistic fit is the third option: collaborative partnership. Rather than maintaining everything internally or delegating everything blindly, the business works with external partners while gradually building internal operational capability. In an era where lean teams thrive on AI, this model of building internal capacity through collaborative guidance is exceptionally well-suited. Establishing clear boundaries between in-housing and outsourcing is also examined in our article on deciding between in-housing and outsourcing.
Case study: A company that grew by letting go of "total in-housing"
Consider a concrete example. The company referenced at the start (name withheld) initially leaned toward shifting entirely to in-house development because "AI makes it possible." However, facing persistent engineering hiring challenges, leadership revised their strategy.
What they did was redraw the line between in-house and outsourced responsibilities. Day-to-day minor modifications and routine operations were handled internally by existing staff using AI assistance, while foundational architecture and complex challenges were addressed alongside an external partner. Crucially, outsourced components were never left as black boxes; they were handed over so internal team members understood the codebase. Letting go of the ideal of doing everything in-house proved decisive. By avoiding plans dependent on talent they could not hire and establishing boundaries viable for their current team, development progressed and operational expertise steadily accumulated inside the organization. For approaches to ongoing operations that avoid passive black boxes, refer to key points for clients in maintenance and operations contracts.
Start by identifying the boundaries your company can sustain
The trend toward smaller development teams is a distinct tailwind for SMBs. Lean teams can now maintain a wider range of software than ever before. However, this neither means that people are unneeded nor that everything should be built in-house. What matters is looking squarely at hiring and training realities and identifying where to draw the boundary of what your organization can realistically sustain.
If you struggle to recruit talent while pursuing in-house development, find your systems black-boxed due to complete outsourcing reliance, or wish to structure a lean operational model powered by AI, feel free to contact GleamHub for a free consultation on system development and in-housing support. From defining boundaries between in-house and outsourced work to transparent handovers and establishing AI-enabled workflows, we partner with you to support your team and business.
Sources
- Gartner Predicts 60% of Organizations Will Adopt Smaller Software Engineering Teams by 2029 | Gartner
- Gartner Predicts 60% of Organizations Will Fully Deploy Smaller Software Engineering Teams by 2029 | Publickey
- 60% of organisations to adopt smaller software engineering teams by 2029: Gartner | People Matters
