The conversation about outsourcing sales almost always starts crooked. The client asks for a team that will "bring in customers" and the provider answers with a quota. Both are talking about the outcome and neither about the work. Six months later the argument is predictable: the provider says the list was cold, the client says the team couldn't sell, and nobody can show which of the two actually happened.
Sales can be outsourced. What cannot be outsourced is responsibility for the offer and for who it gets sold to. The whole difference between a commercial operation that works and one that collapses in quarter two sits in that sentence.
Sales isn't one block, it's three jobs
Treating "sales" as a single thing is the first design error. In practice there are at least three distinct jobs, with profiles, rhythms and measurement that look nothing alike:
- Generation and qualification. Finding who to contact, reaching them, understanding whether they have the problem the product solves, and booking a conversation.
- Closing. Presenting, handling objections, negotiating terms, signing.
- Expansion and retention. Renewing, widening scope, recovering accounts that went quiet.
The first one outsources best, and it's also the one most often postponed in-house, because it's repetitive and thankless. The second outsources well only when price and terms are fixed. The third is rarely worth handing over whole: the relationship with an account that already bought is an asset, not a task.
What delegates well
There's a pattern behind the commercial functions that work when outsourced: repeatable process, written criteria, bounded decisions. With that in hand, these are the ones that come out right most often.
- High-volume cold prospecting. A discipline job, not an improvisation job. Measured by activity and effective conversations.
- Inbound lead qualification. Especially outside the in-house team's hours, where the cost of answering late is losing the lead.
- Booking and confirming meetings. Including no-show reduction, which is an operational process and not a sales talent.
- Following up on quotes already sent. The work an in-house rep postpones whenever a bigger opportunity is open.
- Reactivating an inactive base. High volume, low expected conversion, clear script.
- Transactional selling from a fixed catalogue. When there's no margin to negotiate and no configuration to design.
What those six share isn't simplicity. It's that you can write down in advance what happens in each situation and what happens when an exception shows up.
What's worth keeping in-house
The other side of the list matters just as much. Don't delegate negotiation when there's discretionary margin on the table, because whoever grants a discount has to answer for the margin. Nor key accounts, where the relationship outweighs the process. Nor selling an early-stage product, while the message still changes every week: an external team can't learn a story that hasn't been decided. Nor selling where the rep is also the technical expert designing the solution.
None of that is permanent. A function you don't delegate today can be delegated once the process settles. What doesn't work is handing it over before that.
The list and the offer decide before the team does
This is the point that sinks the most commercial contracts, and it rarely appears in the proposal. An outsourced team amplifies what already works; it doesn't discover what doesn't work yet. If nobody in-house has managed to sell that product to that segment, outsourcing won't find the message: it will repeat it faster and burn through the list while doing so.
Three things have to exist on the client side before signing: a segment defined by verifiable criteria, an offer with price and terms that someone has already sold at least a few times, and a list that can be contacted lawfully. If one is missing, the honest move isn't an annual contract but a bounded pilot, designed the way we lay out in how to design a BPO pilot.
An external sales team doesn't fix a proposal that isn't competitive. It puts it on the phone more often.
What to measure and what not to promise
The measurement mistake is the same one that shows up in collections: asking the provider for an outcome they don't fully control. An outsourced commercial operation is steered with stage indicators, not with a revenue number at month end.
- Effective contacts, not attempts. Attempts measure effort; effective contacts measure how much of the list you can actually reach.
- Held conversations, with a written definition of what counts as a conversation and what counts as an early brush-off.
- Meetings booked and meetings held, always reported separately. The gap between the two is an indicator in itself.
- Opportunities accepted by the in-house team. This is the one that prevents the most arguments, because it forces the client to say whether the opportunity was any good.
- Stage progression, to see whether what enters the funnel moves or stalls.
Closed revenue gets measured too, of course. But attributing all of it to the provider is unfair in both directions: it charges them for a drop that may come from price or product, and it hands them a result that may come from a brand campaign. The same discipline of measuring the work and not only the outcome is in outsourced collections: what to measure and what not to promise.
How to pay without breaking quality
The payment structure shapes the team's behaviour more than any training does. A fixed per-person model puts the risk on the client and earns the right to demand quality of execution. A pure commission model puts all the risk on the provider, and that gets paid for in ways that never show up on the invoice: the provider assigns its best people to the accounts most likely to close, and the pressure to bill pushes practices that later surface as refunds, complaints and early cancellations.
A blended model is usually what keeps the relationship standing: a base that covers the cost of the team, plus a variable tied to verifiable stage indicators rather than revenue alone. How each alternative splits the risk is worked through in BPO pricing models.
That team speaks in your brand's voice
A support agent talks to someone who is already a customer. An outsourced sales team contacts people who didn't ask for the call, and does it using the contracting company's name. The reputational risk is higher and it's managed with three things: a conversation framework instead of a rigid script, monitoring with a rubric that also scores what must not be said, and explicit rules about commercial promises. No rep should be able to commit to a deadline or a discount the contract doesn't allow.
On top of that sits data handling. Contacting a list means defining where the data came from, what purpose it was collected for, how do-not-contact requests are recorded and who answers to the data subject. In Colombia that falls under the personal data protection regime and is worth putting in writing before the first contact; how the roles split is explained in data controller vs. processor in outsourcing. None of this is legal advice: it's the conversation to have with counsel on each side.
The ramp is longer than anyone plans for
A support agent learns to answer what they're asked. A rep has to understand the product, the buyer, the competitor and the objections that appear when the conversation gets difficult. That doesn't happen in a classroom. It happens by talking to real prospects and reviewing those conversations one by one over the first few weeks. Planning a commercial ramp on the same calendar as a support operation is the most common way to conclude, in month two, that the team is no good. The logic of a well-built ramp is in how long a new agent takes to be ready.
How smartBPO works it
We start by splitting the funnel and agreeing in writing which stretch we take on and which we don't, instead of signing up to a headline quota. We ask to see the segment, the offer and the origin of the list before quoting, and if any of that isn't ready we say so and propose a bounded pilot rather than an annual contract. We report effective contacts, conversations, meetings held and opportunities accepted by the client's own team, with each definition agreed at the start. We prefer a blended model over pure commission, because the incentive to close at any cost ends up being paid for out of the client's brand. And we monitor what gets said in each conversation with the same rubric we use for quality in any other operation.