Nesting: What It Is, and How Long It Should Last

Nesting is the period after classroom training when a new hire starts taking real customer contacts with reduced volume and extra support, before moving to full production. Depending on how complex the job is, it should run from about 5 days to 2 weeks. It's also called incubation, grad bay, transition, or soft landing — the name changes by company, the stage doesn't.

It's the least-written-about stage in contact center onboarding, which is strange given how much rests on it. Ask five operators what nesting is and you'll get five answers, at least two of which describe something else entirely. That vagueness isn't harmless — it's how programs end up reporting a nesting period they don't actually run.

What nesting is — and the three things it gets confused with

The definition has three parts, and all three have to be true. The new hire is handling live customer contacts. The volume or target pressure they're carrying is reduced relative to full production. And there is support available in the moment that a fully-ramped agent doesn't get. Drop any one and it isn't nesting.

• Shadowing is not nesting. In shadowing, the new hire watches someone else work. It's genuinely useful and it belongs in the ramp, but the new hire isn't handling the contact, so nothing about it tells you whether they can. Nesting is the first stage where you find out.
• A reduced-volume week with no support is not nesting. Turning the dial down and walking away produces a slow first week, not a supported one. The support is the mechanism; the reduced volume just creates room for it.
• Going live at full volume is not nesting, whatever it's called on the schedule. This is the most common version. The plan says "nesting week," the queue says otherwise, and the new hire spends it doing the job with no more help than anyone else. That's a first shift with a friendlier label.

How long it should last

The same thing that sets classroom length sets nesting length: measured complexity — how many types of work an agent handles, how many systems they touch in a typical interaction, and how heavy the compliance burden is. Here's the published benchmark:

• Low complexity — about 5 days. One type of work, one or two systems, little or no regulation. Follows roughly 2 weeks of initial training.
• Moderate complexity — about 6 days. One or two types of work, a handful of systems, some compliance requirements. Follows about 3.5 weeks of training.
• High complexity — about 9 days. Multi-skill or meaningfully regulated, with several systems in a typical interaction. Follows about 5.5 weeks of training.
• Very high complexity — about 2 weeks. A universal role across three or more types of work, many systems, heavily regulated. Follows about 7 weeks of training.

Notice that nesting grows with complexity but much more slowly than classroom time does. At low complexity it's about half the length of initial training; at very high complexity it's under a third. That's not an oversight. Nesting is expensive in a way classroom isn't — you're paying for a seat that's producing at a fraction of its rate, and you're spending supervisor attention that has to come from somewhere. And its returns flatten: past roughly two weeks, what's left to learn is learned in production with ordinary coaching, and extending nesting mostly delays the point where the agent starts building real pace.

The detail that decides whether it works: staged volume

If there's one thing to take from this piece, it's this. Contact volume during nesting should be reduced and then deliberately stepped up — not held flat, and certainly not switched on at full pressure on day one. A nesting period that opens at full volume is a shift, not a ramp, and it does almost none of the work it's supposed to do.

The reason is about what the new hire can actually process. In their first days on live contacts, the gap between contacts is where the learning happens — where they look up the thing they didn't know, ask the question they couldn't ask mid-call, and mentally re-run the one that went badly. Full volume removes that gap entirely. The new hire is still having the experiences, but they have no capacity left to convert them into anything. They just get through the day, and the day teaches them nothing except that the job is overwhelming.

Stepping the volume up matters just as much as starting it low. The point of nesting isn't to be a comfortable week; it's to move someone from supported to unsupported in controlled increments, so that by the last day they're working at close to production pressure while the support is still there to catch what breaks. A nesting period that stays gentle all the way through just relocates the shock to the following Monday.

What "extra support" should actually mean

"Support" is where nesting programs get vague, so here is the concrete version. Four things carry most of the weight: a named mentor or buddy the new hire can go to without feeling like they're escalating; a clear path for getting help mid-contact, so a stuck agent has somewhere to go that isn't guessing on air; explicitly reduced performance targets, stated out loud rather than tacitly understood; and scheduled check-ins that happen on the calendar rather than when someone notices a problem.

That last one matters more than it looks. The most common way a struggling new hire gets identified is that they complain or resign — which is to say, the program finds out at the point where it's too late to do anything. Scheduled check-ins during nesting are the cheapest mechanism there is for finding out earlier, and they cost calendar time rather than headcount.

Nesting is where you find out whether the gate was right

There's a second function that rarely gets stated. Nesting is the first time anyone sees the new hire do the actual job, which makes it the first honest test of whatever decision let them out of the classroom. If a program graduates people on a scheduled date rather than against a demonstrated bar, nesting is where that shows up — and if nobody is watching for it, the information is generated and then thrown away.

Which is why having no nesting period at all is one of three structural failures that cap how ready a program's agents can be, regardless of how strong everything else is. Without it, the first time anyone observes a new hire doing the job unaided is in full production, on a real customer, with no mechanism to catch what goes wrong. Adding a week of classroom does not substitute for it. Neither does hiring better.

If you don't have a nesting period

Building one is more a scheduling exercise than a budget one, which is why it's usually the highest-return change available to a short program.

• Start with a week, even at low complexity. Five days is the benchmark floor and it's enough to matter. Don't wait until you can staff the ideal version.
• Set the day-one volume target at something deliberately low, and write down the step-up. Day 1 at a fraction of production, rising across the week. If the step-up isn't written down, it doesn't happen — the queue always wins an argument with an intention.
• Name the support, don't assume it. A specific mentor, a specific escalation path, a stated target reduction, and check-ins on the calendar. "Ask anyone" is not a support structure.
• Watch what nesting tells you about the gate. If most of a class struggles in nesting, the problem isn't nesting — it's what let them out of the classroom, and that's where the fix belongs.

Nesting is the second half of a ramp, and it's hard to size sensibly without the first half. How long contact center training should be covers the full complexity benchmark — classroom weeks, nesting, practice reps and situation coverage — and why complexity is measured from the job rather than inferred from the industry.

If you want to know how your own nesting period compares — its length, whether the volume is genuinely staged, and what support exists in the first four weeks — the free Agent Readiness Assessment scores exactly that as part of an eight-dimension picture of your program. It takes about five minutes and needs no login.