Multi-site and follow-the-sun: one queue or many?

Scheduling · ~6 minute read

More sites isn’t automatically more resilience

Operations end up with multiple sites for all sorts of reasons — acquisitions, growth, a deliberate spread for resilience, an outsourcer bolted onto an in-house team. For the planner, the multi-site question isn’t really about buildings; it’s about whether you run one pooled queue or several separate ones, and that single choice quietly sets how many agents you need and how good your worst day can get. Get it right and multiple sites are a genuine advantage; get it wrong and you pay for fragmentation while calling it resilience.

Pooling versus fragmentation

The maths favours pooling. One virtual queue routed across all sites is a single large pool, and large pools are efficient — the power of one means the same service costs proportionally fewer agents than the same demand chopped into separate site queues. Split the queue and each fragment carries its own buffer, so the total headcount climbs for no extra service. But operational reality pushes the other way: skills, language, local knowledge, regulation and the desire for a site to “own” its customers all argue for separation, and a degree of separation is real resilience — if one site goes dark, the others carry on. The planner’s job is to find the line: pool what you can to capture the efficiency, separate only what genuinely needs it, and be honest that every separation has a headcount price.

Follow-the-sun flattens the unsocial hours Site A (local day) Site B (local day) combined — far flatter One site’s 3am is another’s mid-afternoon — cover without the night shift.
Spread sites across time zones and the brutal overnight cover in one region is just daytime staffing in another — the genuine prize of follow-the-sun.

Making multi-site pay

Pool the queue wherever skills and regulation allow, and treat each carve-out as a deliberate, costed exception rather than a default. Where you can spread sites across time zones, use follow-the-sun to cover the unsocial hours with someone else’s daytime — the cleanest way there is to staff a 24-hour operation humanely. Watch the fairness of how work is distributed across sites, because a “virtual” queue that quietly sends all the hard contacts to one site breeds resentment and skews your quality data. And keep the resilience honest: separation only protects you if the surviving sites actually have the skills and headroom to absorb a failure, so plan the fallback, don’t just assume it. Multiple sites are a tool; pooled thoughtfully they cut cost and smooth the day, fragmented carelessly they just multiply your buffers.

Pair this with the power of one, in-house versus outsource, and planning with an outsourcer.