Most companies that stand up an outbound call center do it in the wrong order. They buy a dialer, hire six agents, write a script over a weekend, then go looking for a list to call. Six weeks later the connect rate is 4% and someone says outbound is dead.
Outbound is not dead. It is unforgiving. The channel punishes bad targeting harder than any other, because every mistake costs a live human minute rather than an impression. When the list is right and the offer matches the moment, it still produces the cheapest qualified meetings in B2B, on a schedule you control.
This guide covers whether to build or outsource, how to construct a list worth calling, what a script that survives contact looks like, the dialer and compliance decisions that quietly determine your connect rate, what it costs, and the numbers that tell you whether it pays.
What an outbound call center actually does
An outbound call center is a team, in-house or contracted, whose agents initiate calls to a defined list rather than answering calls that come in. That single difference changes how you staff it, how you measure it, what software it needs, and which regulations apply.
The work falls into distinct programmes, and mixing them on one team is a common, expensive mistake:
- Cold prospecting and appointment setting. Calling companies that have not asked to be contacted, qualifying them against a short set of criteria, and booking a meeting for a closer. Highest skill, lowest connect rate.
- Speed-to-lead follow-up. Calling inbound leads within minutes of a form fill. Connect rates are several times higher than cold, and it is the easiest programme to justify financially.
- Lead reactivation. Old, closed-lost, or dormant records. Cheap list, warm-ish context, variable results.
- Retention and renewal calls. Existing customers approaching renewal or showing usage decline.
- Research and enrichment. Verifying titles, org structure, and current tooling so other channels work better.
Speed-to-lead and reactivation almost always beat cold prospecting on cost per meeting, and most teams underinvest in both because cold calling feels like the "real" outbound work. If inbound leads sit in a queue for hours before anyone calls, fix that before building a cold programme. Our guide to lead routing software covers the plumbing that makes minute-level follow-up possible.
Outbound and inbound are different businesses
Teams that already run an inbound operation often assume they can extend it. The two share a phone system and almost nothing else.
Inbound agents handle demand that has already declared itself, so the operational challenge is capacity planning against unpredictable peaks. If that is the problem you have, our breakdown of inbound call center services covers the staffing models and service levels that matter.
Outbound agents create demand. Nobody is waiting for the call. The challenge is list supply and morale: keeping qualified records in front of agents, and keeping people motivated through a rejection rate that would break most sales hires. Costs diverge too. Inbound is priced against volume you cannot control; outbound is priced against dials you choose to make, which is easier to model and easier to waste.
The practical consequence is that you should not ask the same agents to do both. Blended teams drift toward inbound because answering a ringing phone is easier than making the next cold dial, and the outbound numbers quietly collapse.
Build in-house or outsource
The honest version of this decision comes down to how complicated your qualification is.
Outsource when the conversation is short and the qualifying criteria fit on an index card. Appointment setting for a well-defined service, speed-to-lead follow-up, list verification, event follow-up, and reactivation campaigns all travel well to a vendor. A good outbound call center partner will have trained agents dialing within two to three weeks, absorbs the hiring and attrition risk, and can scale from two seats to twenty without you signing a lease.
Build in-house when the first call requires real product judgment, when the buyer is senior enough to notice a rented voice, or when your compliance exposure is high enough that you want direct control of what gets said. Regulated categories tend to end up in-house for exactly that reason.
The hybrid that works in practice is to run cold prospecting with a vendor to find which segments respond, then bring the winner in-house once you know it converts. You buy learning speed without committing headcount to a segment that might not work.
Either way, the same diligence that applies to picking a B2B demand generation agency applies here: ask for real dial and connect data from a comparable account, ask who specifically will be on your programme, and ask what happens to meetings that turn out to be junk.
The list is still the biggest lever
Nothing moves results as much as who you call. A mediocre script against a precise list beats a brilliant script against a broad one, every time.
The test for list quality is whether you can complete this sentence about every record: "Everyone on this list is a company that just ___." If the blank has to be filled with something static like "has between 50 and 500 employees," it is a demographic slice, not a reason to call. If it can be filled with something recent, you have a list.
Triggers that reliably produce conversations:
- A new hire in the function that owns your problem, usually within their first 90 days
- A funding round, acquisition, or market entry that creates both budget and pressure
- A technology change on their site, or job postings naming the tool you replace
- A public complaint, review, or job ad describing the exact symptom you fix
Size matters less than most people expect. A list of 400 companies where every record has a genuine trigger outperforms 8,000 pulled from a database filter, and it costs less to work. Agents can also personalise honestly, which is what separates a call that continues from one that ends in nine seconds.
Data hygiene is unglamorous and decisive. Direct dials beat switchboards by a wide margin, and mobile numbers connect better than desk lines while carrying more compliance risk, so you need to know which is which. A list that is 30% wrong numbers does not just waste 30% of dials, it destroys agent rhythm, which costs more than the dials.
What a script that survives contact looks like
A script is not a monologue to be recited. It is a decision tree with pre-written language for the branches agents hit most often.
The opening has one job: earn the next twenty seconds. Name the person, say who you are, and give the specific reason you are calling this company rather than any other. "I saw you posted for a demand gen lead last week" outperforms any clever hook, because it answers the question the prospect is actually asking: why me, why now.
The middle should be a question, not a pitch. One question that only lands if your hypothesis is right and that surfaces the problem in the prospect's words. A real answer gives you a conversation; a deflection tells you the hypothesis was wrong for that segment.
The close is a specific, small ask. A 15-minute call on a named day beats "would you be interested in learning more," which invites a no by default.
Objection handling deserves more preparation than the opening. Write out the eight objections you actually hear, in the words prospects use, and draft a response for each that acknowledges the objection before answering it. Agents who improvise here default to arguing, and arguing loses.
Record calls and review them weekly, because the gap between the script as written and the script as spoken is where the coaching leverage lives. The structure we describe in the sales email guide mirrors the call structure closely, and running both channels off one message tends to lift both.
Dialers, compliance, and the stack
Three technology decisions shape everything else.
Dialer mode. Preview dialers show the agent the record before connecting, which suits complex B2B where research matters. Power dialers work a list at a fixed pace with no research time. Predictive dialers place several calls per available agent and connect the answered ones, maximising talk time and carrying the most regulatory baggage. For most B2B appointment setting, preview or power is right; predictive belongs to high-volume, low-complexity campaigns.
Caller ID and number reputation. Carrier analytics flag numbers that place high volumes of short-duration calls, and a flagged number shows up as "Spam Likely" on the handset. That label alone can halve a connect rate. Rotate numbers, register them properly, and monitor reputation as a standing metric rather than discovering the problem after a month of bad results.
CRM integration. Every dial, disposition, and recording should land against the right CRM record automatically. If agents copy outcomes by hand, the data will be wrong within a fortnight.
On compliance, the obligations are real and the penalties are per-call. In the US that means checking the National Do Not Call Registry and applicable state lists, honouring internal opt-outs permanently, respecting calling hours in the prospect's own time zone, disclosing recording where consent laws require it, and knowing that mobile numbers dialed by automated equipment carry extra exposure. B2B has carve-outs, but far fewer than most sales leaders assume. Write the rules into the dialer configuration rather than the training deck: rules configured in software get followed, rules taught in onboarding do not.
Outbound call center pricing
Pricing models fall into three shapes, and each one moves the risk somewhere different.
Per hour is the most common outsourced model: roughly $9 to $16 per agent hour offshore, $16 to $28 nearshore, and $28 to $50 or more for US-based agents on complex B2B programmes. You carry the performance risk, which is fine once you know the list converts.
Per appointment runs $150 to $500 per booked, qualified meeting depending on target seniority and how tight the criteria are. The vendor carries the risk and prices it in. Watch the definition of "qualified" closely; disputes over show rates are the most common reason these deals fail.
Per seat, per month for dedicated agents lands between $2,500 and $7,000 depending on geography and hours covered. It suits long-running programmes where product knowledge needs to accumulate.
In-house economics look different. Budget $75,000 to $110,000 fully loaded per SDR in most US markets once salary, commission, benefits, tooling, and management overhead are counted, plus $100 to $200 per agent per month for the dialer and data stack, plus two to three months of ramp before that person produces at target. A three-person pod is a $300,000-a-year commitment before it books its first meeting.
Whichever model you pick, convert it to one comparable number: cost per qualified meeting, then cost per closed deal. Run those through the customer acquisition cost calculator alongside your other channels, and check the result against lifetime value with the LTV calculator before you scale. An outbound programme at $400 per meeting is excellent for a $60,000 contract and ruinous for a $3,000 one. Our piece on marketing customer acquisition cost covers what to include and what teams routinely leave out.
The numbers that tell you whether it works
Most outbound reporting drowns in dial counts. Five metrics carry the signal.
Connect rate (conversations divided by dials) tells you whether your data and caller reputation are sound. Below 5% on direct dials, fix the list before touching the script.
Conversation-to-meeting rate tells you whether the offer fits the segment. Healthy B2B programmes land between 8% and 20%. Strong connects with a weak rate here means your hypothesis about the buyer's problem is wrong.
Show rate tells you whether the meetings are real. Under 70% usually means agents are booking anyone who says yes to get off the phone.
Meeting-to-opportunity rate is where vendor-booked meetings quietly fail, and the number to build any per-appointment contract around.
Cost per qualified meeting, then cost per closed deal, is the only number that belongs in a budget conversation.
Track these by segment, not in aggregate. Outbound results are almost never uniformly good or bad; one segment usually carries the programme while others drag the average down, and an aggregate report hides exactly the decision you need to make. The conversion rate calculator is a quick way to compare stage-by-stage rates across segments.
When outbound is the wrong channel
Outbound fails predictably in a few situations, and recognising them early saves money. If average contract value is under about $5,000 a year and the cycle needs more than one call, the arithmetic rarely closes. If your buyer genuinely cannot be reached by phone, no amount of dialing fixes it. If the product needs a demonstration before anyone understands the value, price the channel as appointment setting and nothing more.
If you have no differentiated point of view about the prospect's problem, outbound exposes that faster than any other channel, because nothing about a cold call lets you hide behind design or brand. The cheaper path there is to build demand that arrives on its own, which is the argument for pairing outbound with content-led lead generation.
The strongest programmes run both. Content and search build a base of people who already recognise the problem; the outbound call center reaches the ones who have it but have not started looking. Teams that treat the two as competitors for budget underperform the ones that treat them as a sequence.
FAQ
What is the difference between an inbound and an outbound call center? An inbound call center answers calls that customers initiate, so its core challenge is capacity planning against unpredictable volume. An outbound call center initiates calls to a target list, so its core challenge is list quality and agent morale. They need different metrics, different compensation, and different people.
How much does an outbound call center cost? Outsourced agents run roughly $9 to $16 per hour offshore, $16 to $28 nearshore, and $28 to $50 for US-based teams. Per-appointment pricing typically falls between $150 and $500 per qualified meeting. In-house, budget $75,000 to $110,000 fully loaded per SDR plus tooling and a two to three month ramp.
What connect rate should an outbound programme expect? On verified direct dials in B2B, 8% to 15% is a reasonable working range, with speed-to-lead follow-up on fresh inbound leads running considerably higher. Sustained rates under 5% almost always point to a data problem or a flagged caller ID rather than a script problem.
Is cold calling still legal for B2B in the US? Yes, with conditions. Business-to-business calls carry narrower obligations than consumer calls, but do-not-call requests, calling-hour limits, recording-consent laws, and the rules around automated dialing of mobile numbers still apply. Configure those constraints into the dialer and have a qualified adviser review your programme rather than assuming B2B is exempt.