To avoid maintenance debt, IFPA's two-person digital team filters every new request through four tests: substance, scope, distributed ownership, and technical reusability. This ensures every project either creates a lasting capability or can be managed by non-technical colleagues, preserving the team's limited bandwidth for high-impact work.
To combat maintenance drag, IFPA's team automated the tedious process of updating sponsor logos on event pages using their DAM and GraphQL. This eliminated manual resizing and placement, with a long-term plan to shift ownership of this task entirely to the sales team and then to the sponsors themselves, creating a self-service model.
Don't delete content with low human traffic. This content serves a second, crucial audience: AI engines. A niche article may not attract many visitors today, but it can be the perfect answer for an AI search tool tomorrow. Instead of deleting it, simply reduce promotion and let it serve as a long-term asset for AI discovery.
A small team cannot afford to supervise complex AI. The best filter for AI adoption is whether it removes manual friction from an existing process, not if it adds a new system to manage. Focus on solid data foundations first, then use AI for practical tasks like enforcing taxonomy or running analytics queries, avoiding tools that require constant babysitting.
Agility for resource-constrained teams is not about speed or taking on more projects. Instead, agility is the discipline of creating enough operational "breathing room" to focus on the next important thing. This is achieved by solving specific, known problems with practical, right-sized solutions rather than chasing the latest "best-in-class" technology for its own sake.