How many BPO vendors should a company use for customer support?

There is no universal right number. What determines whether a multi-vendor setup works is governance capacity: whether someone owns a consistent cadence of SLA review, performance comparison, and escalation across every vendor. A single vendor with no governance is riskier than three vendors with strong governance.

Why do companies keep underperforming BPO vendors?

Often because of the sunk cost fallacy: the time already invested in training a vendor on the product and processes feels like a reason to continue, even though that investment cannot be recovered regardless of the decision going forward. The relevant question is whether the current vendor would still be chosen today, not how much has already been spent.

What should trigger a BPO vendor review?

A structured review should be triggered by SLA misses sustained over a full quarter, a documented rise in repeat contacts traceable to a specific vendor, or a renewal date approaching without a comparative benchmark against alternatives having been run in over a year.

Vendor Governance · CX Operations

How many CX vendors are too many? A governance framework, not a number

By Roberto La Rosa · Fractional CX Director · July 2026 · 7 min read

GMs ask me this a lot: how many BPO vendors should we be running? It is the wrong question, and answering it with a number is how vendor chaos starts in the first place.

I have governed four BPO providers across three markets at once, run full RFQ and vendor selection processes, and migrated an internal contact center to an external provider without a service gap. The number of vendors was never what determined whether those setups worked. Governance capacity was.

Why the vendor count is the wrong metric

A single vendor with no structured review cadence is riskier than three vendors under tight governance. The count feels like the variable that matters because it is the one you can point to in a slide. What actually predicts outcomes is whether someone owns, on a fixed schedule, the comparison of SLA performance, cost per contact, and quality across every provider you use.

Without that ownership, a single vendor drifts unnoticed because there is nothing to compare it against. With multiple vendors and no governance, you get the same drift, multiplied, plus the coordination cost of managing several relationships that nobody is actually managing.

The bias that keeps a bad vendor on contract

I have sat in enough renewal conversations to recognize the pattern immediately. A vendor has missed SLA for months. The instinct in the room is always some version of: we have already invested so much in training them on our product, replacing them now would cost too much.

That reasoning has a name: the sunk cost fallacy. The time and money already spent on that vendor relationship are gone regardless of what you decide today. They cannot be recovered by staying, and they are not saved by leaving. The only question worth asking is: knowing what you know now, would you choose this vendor again?

The cost already sunk into a vendor relationship is not in the contract. It is in the way you look at the decision.

The same bias shows up on underperforming individual agents, on platforms chosen years ago and never revisited, on processes everyone privately knows are broken but nobody touches because "we built it." Vendor governance is really a discipline of asking that question on a schedule, before the sunk cost has a chance to accumulate into something that feels too large to walk away from.

A governance framework that actually works

What I put in place, across every multi-vendor operation I have run, has the same four components regardless of how many providers are involved.

A fixed review cadence. Monthly for operational metrics, quarterly for strategic performance. Not "as needed," because "as needed" means never until something breaks.
Comparable metrics across every vendor. Same SLA definitions, same quality scoring, same cost-per-contact calculation. If your vendors report different numbers in different formats, you cannot actually compare them, which means you are not really governing them.
Incentives aligned to outcomes, not activity. A vendor paid purely per contact handled is incentivized to close fast, not to resolve. Renegotiate SLAs so quality and resolution carry real weight in the commercial terms, not just volume.
A defined transition trigger. Decide in advance what sustained underperformance looks like, in writing, before you are inside an emotional renewal conversation. A trigger you set calmly in January is more reliable than a judgment you make under pressure in November.

What a real vendor transition looks like

Vendor transitions are treated as high-risk events, and done without a governance structure behind them, they are. Done with one, they are routine. I have led a full vendor transition with zero operational disruption, maintaining SLA continuity throughout the migration, because the groundwork, parallel run periods, knowledge transfer protocols, and a clear cutover plan, existed before the transition decision was even made.

The organizations that fear vendor transitions the most are usually the ones with the weakest governance, because for them a transition means building the entire evaluation framework from scratch under time pressure, at the exact moment they can least afford to get it wrong.

When you actually need more vendors, or fewer

More vendors make sense when you are entering new markets with different language or regulatory requirements no single provider covers well, or when concentration risk on one provider becomes a genuine continuity threat. Fewer vendors make sense when governance capacity is the binding constraint: if nobody has the bandwidth to run the cadence above properly, consolidating to fewer, better-governed relationships beats spreading thin across many.

This is also where the fractional model earns its keep. Vendor governance is exactly the kind of ongoing, structural responsibility that a full-time hire is expensive for and a one-off consultant cannot sustain. I wrote about that trade-off in more detail in Fractional CX Director: when you need one, what it costs, and when you don't.

Roberto La Rosa Fractional CX Director and founder of Happiness Harbor, based in Milan. 20+ years leading customer operations across European scale-ups, with teams of over 1,850 agents in 4 countries. He started his career as an agent. Author of "AI Alone Is Not Enough" and the white paper "CX Is Not Delegated. It Is Built." His work has been published in CMI Magazine. Currently studying organizational psychology. LinkedIn