Build, Buy, Borrow

The options

• Build. Develop current employees through upskilling, reskilling, rotations and promotion. Slower, often cheaper per head, and good for retention and culture. Works when you have time and people with adjacent skills.
• Buy. Hire externally, either permanent or senior specialists. Faster for skills you do not have at all, but expensive in a tight market and with real integration risk.
• Borrow. Use the contingent workforce: contractors, staffing agencies, consultants, outsourcing partners or gig platforms. Fast and flexible, with no long-term commitment, but higher hourly cost and knowledge that walks out at the end of the contract.
• Bot. Automate or use AI for the task itself, such as self-service, workflow automation or AI agents handling routine contacts. It reduces the need for people in a role rather than finding them.

Some versions add bind (retain the people you have) and bounce (move out roles you no longer need). The core idea stays the same: decide deliberately how each gap will be filled rather than defaulting to external hiring.

How to choose

For each skill or capacity gap from your workforce planning, ask four questions:

1. How soon is it needed? Under three months usually means buy or borrow. A year or more makes build realistic.
2. How long will it be needed? Permanent, core needs justify build or buy. Temporary peaks and projects suit borrow.
3. How scarce is it outside? If the market is thin and expensive, building from adjacent internal talent may be the only practical route.
4. Is it core to how you compete? Knowledge you want to keep in-house argues for build or buy. Commodity work can be borrowed or automated.

Most plans use a mix. A useful rule of thumb many workforce planners use: build for depth, buy for speed and new capability, borrow for peaks and uncertainty, automate the repetitive.

Worked example

A health insurer's member services operation needs 120 more agents for open enrollment season, 15 agents who can handle Medicare Advantage questions (a licensed, specialist skill), and three workforce analysts. It also wants to reduce routine ID-card requests.

• Borrow 80 seasonal agents through a staffing partner for the 12-week peak, since the need disappears in February.
• Buy 40 permanent agents to cover growth and backfill expected attrition.
• Build the 15 Medicare-capable agents by licensing and training high performers from the commercial team, which also gives those agents a career step. Backfill them with part of the 40 permanent hires.
• Buy the three workforce analysts, since nobody internal has forecasting experience.
• Bot the ID-card requests through self-service and an automated channel, which removes an estimated 8 percent of call volume and reduces the seasonal borrow from 80 to about 70.

The same gaps, filled by external hiring alone, would cost more, take longer for the licensed roles and leave 80 surplus agents after the peak.

Cost and risk trade-offs

Each option has a different cost profile. Build costs training time, lower productivity during ramp-up and a backfill for the person who moved. Buy costs recruiting fees or effort, often a pay premium over internal staff, and the risk of a bad hire. Borrow costs a markup on pay, commonly 25 to 60 percent for agency staff depending on role and market, plus onboarding time each time a worker changes. Bot costs implementation and maintenance, and it moves work rather than always removing it, since exceptions still need people.

There is also legal risk in borrow. Long-term contractors doing the same work as employees under your direction can be misclassified, and heavy reliance on agency workers creates joint-employer and co-employment questions. Rules vary by state and country.

Common mistakes

• Defaulting to buy for every gap because it feels fastest, then wondering why internal mobility and engagement are low.
• Committing to build without giving people the time and budget to learn.
• Borrowing for years for core roles, paying markups indefinitely.
• Counting bot savings before the automation actually handles the volume.
• Deciding role by role rather than looking at the full gap across the plan.