You probably don't need a bigger team. You need a bench.
Permanent headcount is the right answer for stable workloads. For everything else, a vetted bench is faster, cheaper, and lower-risk.
When teams hit a capacity wall, the default move is to hire. More headcount, more output, problem solved. It's the only model most leaders have ever used, so it's the model they reach for — even when it's the wrong tool for the job.
Permanent headcount is genuinely the right answer for stable, predictable workloads. The work isn't going away, the volume is consistent, the skill set is core to the business. Hire. Onboard. Build the muscle.
For almost everything else, it's the wrong shape. Spikes in demand. Specialist work you need twice a year. Exploratory builds you're not sure will survive the next quarter. Capacity experiments where you don't yet know if the work is real. Hiring permanent staff for any of these means you're 90 days from useful output, you're carrying the cost forever, and if the work disappears you've got a redundancy conversation instead of an end-of-engagement.
A small, vetted bench of external specialists solves a different shape of problem. They're spinnable in a week, scale down to zero with no friction, and you're paying for output rather than presence. The catch is the bench has to be vetted in advance — bench-by-Google when you're already underwater is just a worse, slower hire.
Here's the test I use: for the next piece of work you're about to hire for, ask whether the workload will still exist, at the same volume, in 18 months. If you can't say yes with conviction, hiring is the expensive way to find out. A bench is the cheap way.