The questions buyers actually ask.

You compare it against what other buyers actually pay for the same role, in the same country, at the same service level. That sounds obvious, but few buyers can do it, because the data sits with the providers rather than with the people buying.

Before you conclude anything from a comparison, normalise three things. Shift pattern, because a 24/7 rate is not comparable with 8x5. Language, because a bilingual seat costs materially more than a monolingual one in the same city. And inclusions, meaning whether QA, team leads, training and attrition backfill sit inside the rate or get billed on top. Two quotes that look far apart are often the same number once those three are lined up.

Most contracts measure the wrong thing. They govern handle time, adherence and internal quality scores, which tell you whether the operation is busy, not whether it worked.

Ask a different question. Did the customer's problem get solved, did they come back, did they buy again, did they stay. If none of those appear anywhere in your reporting, you do not have a performance picture, you have an activity picture. The quickest test is to open your last quarterly review pack and count how many of the metrics would change if every customer left tomorrow.

Two checks will tell you. First, could your provider hit every KPI in the contract while your business got worse? If the answer is yes, you are governing proxies. Second, is there a bonus or a penalty attached to anything a customer would recognise as a result? Usually there is not.

KPIs also should not be fixed for the life of the deal. What matters in year one of a transition is not what matters in year three of a steady state, and a contract that cannot move its targets tends to get quietly ignored by both sides instead of renegotiated.

Not by changing the invoice. The most common failure is bolting an outcome metric onto an existing per-seat deal, which leaves both sides with the old incentives and gives them a new thing to argue about.

It is a change in how the two sides work together: what data the provider gets, what they are allowed to change, who decides, and what happens when the number moves for reasons outside their control. Start with one measurable business outcome, agree in writing how it will be attributed, and run it alongside the existing model before you replace anything.

Decide what you are actually buying before you write it. Most RFPs ask every provider to price the same specification, which guarantees they all come back looking similar and the decision collapses onto rate.

Ask instead for the things that separate them. How they staff a ramp. What they do when quality drops. What their attrition is on comparable work and how they backfill it. And what they would change about your specification. Leave room for them to disagree with you, because the provider who tells you your spec is wrong is usually the one worth talking to.

Work out the real cost of what you run today first, because most in-house comparisons are wrong. The internal number usually leaves out recruitment, attrition, management overhead, real estate, technology, and the time your own leadership spends on it.

Then ask whether the work is a source of advantage or a cost to be run well. If it genuinely differentiates you, outsourcing it rarely ends well whatever the price. If it does not, the only real question is who runs it best, and that one is answerable.

Business process outsourcing. Paying another company to run part of your operation rather than running it yourself. In practice that usually means customer service, technical support, sales, back office processing, finance and accounting, or content moderation, delivered from that company's sites with their staff.

You will hear the same arrangement called different things depending on who you ask and where they are: outsourcing, contact centre or call centre services, customer experience or CX, customer success, shared services, or simply your supplier, vendor or outsourcing partner. They broadly describe the same thing.

It is not the same as staff augmentation, where you rent people who still work under your management. In BPO the provider owns the process and is accountable for the result, which is why the contract matters more than the rate does.

It means another company's employees do work your customers experience as yours. Practically that involves a contract, a rate, a set of service levels, a governance rhythm, and a transition period in which your process is documented, moved and rebuilt somewhere else.

The work can stay in your own country, move to a nearby one, or go much further away. Onshore, nearshore and offshore are just labels for how far it travels, and each one changes the cost, the timezone and the risk differently.

The transition is the part most people underestimate. Whether the work is cheaper is decided by the rate. Whether it is any good is decided by how well it was moved.

One where at least part of what you pay is tied to a result the business recognises, rather than to hours, seats or headcount. Retention, resolution, conversion, cost per resolved case.

The real distinction is not the pricing mechanic, it is who carries the risk. In a per-seat contract you carry the risk of the operation not working, because you pay for capacity whatever it produces. In a genuine outcome contract the provider carries part of that risk, which is why they will want more control and better data in return. That trade is the whole negotiation.

A rate card is the price per person, per hour or month or seat, usually broken down by role and location. It is the number most buyers negotiate hardest on, and the number that predicts the least.

The real cost is in what the card does not show. What you pay when volumes miss forecast. What a change request costs mid-term. What you carry during transition and ramp. How attrition backfill is charged. And what you spend internally managing the relationship. A lower rate card with worse terms is routinely the more expensive contract.