The short answer
Hire permanently for your baseline, which is the work that exists every month regardless of what is happening. Use augmentation for anything above it. That way the peaks do not turn into redundancies when they pass.
Most teams size for the peak because the peak is when the decision gets made, under pressure, with a backlog that looks permanent.
Find your actual baseline
- Maintenance and support that happens whatever else is going on
- Security patching and dependency upgrades
- The steady flow of small product improvements
- Incident response and the work that follows it
- Whatever keeps existing customers from leaving
That is your floor, and it is usually smaller than the current team. Everything above it is project work, which is where flexibility belongs.
The cost of getting it wrong
| Sized for | In a peak | In a trough |
|---|---|---|
| Peak demand | Fine | Paying for idle capacity, or redundancies |
| Baseline only | Everything is late | Efficient |
| Baseline plus augmentation | Absorbs it | Engagement ends, no redundancy |
The bottom row is not free. Augmentation costs more per day and repeats onboarding. It buys you the ability to be wrong about demand without it costing someone their job.
Keep the permanent core strong
Flexing works only if the permanent team holds the knowledge. If augmented people outnumber the people who understand the product, the engagement ending takes the knowledge with it.
Practical rule: your own people should be able to run everything in production and review everything that gets merged. If that stops being true, the balance has tipped too far.
Plan the ramp honestly
- Augmentation is quicker to start than hiring, not instant. Allow weeks.
- Onboarding costs your team time every cycle, so fewer and longer beats many and short.
- Notice periods mean ramping down takes weeks too.
- Knowledge transfer has to happen before the ramp-down, not during it.
- A short trough is usually cheaper to absorb than to react to.
When permanent hiring is the right answer
If the work above your baseline has been there for a year and shows no sign of ending, it is not a peak, it is your new baseline. Keeping it on augmentation indefinitely costs more and leaves knowledge outside the company.
The honest signal is renewing the same engagement repeatedly. At the second or third renewal, the question is whether you are buying flexibility you are not actually using.