The one outbound number most sales leaders cannot answer
Ask a founder what their cost per meeting is and you will usually get one of three answers: a blank look, a number that only counts agency retainer, or a number that quietly counts meetings that were booked but never happened. All three are wrong, and all three are expensive.
Cost per meeting is the most useful outbound metric in 2026 because it is the only one that survives the current mess. Reply rates are falling for structural reasons nobody controls. Connect rates are falling. Deliverability rules got stricter. LinkedIn enforcement got sharper. In that environment, "our reply rate is 3%" tells you almost nothing about whether your outbound function is worth running. Cost per held, qualified meeting tells you everything.
The good news is that 2026 gave us unusually clean public benchmarks to compare against. The ORRJO State of B2B Outbound 2026 report, published in May, cross-referenced its own data across 10,000+ booked meetings against Bridge Group, Pavilion, Martal and Validity figures. Outbound Sales Pro's 2026 pricing guide, updated this summer, published a full in-house cost stack. Between them you can build a defensible model in about an hour.
This post covers what those benchmarks actually say, the two accounting errors that make most internal numbers useless, and the four levers that move cost per meeting more than any tool purchase will.
What a meeting actually costs in 2026
Start with the build-versus-buy picture, because the gap is wider than most people assume.
The in-house stack
Outbound Sales Pro's 2026 breakdown puts a productive US SDR at $9,800 to $14,200 per month fully loaded, once you stop pretending base salary is the whole cost. The components:
- Base plus variable comp (OTE): $6,500–$9,500/month, with US OTE typically landing in the $75K–$100K band
- Employer taxes and benefits: $1,300–$2,000/month, roughly a 20–25% burden
- Sales engagement, dialer, data and enrichment, inbox infrastructure: $200–$600/month
- Management, QA, coaching and enablement overhead: $800–$1,800/month
Against Bridge Group's widely cited median of roughly 14.6 meetings per SDR per month (top quartile 20–24), that produces a cost per meeting of $821 to $1,150. That is the honest in-house number, and it is materially higher than the "$75K salary divided by meetings" arithmetic most boards see.
The outsourced stack
Agency pricing in 2026 clusters into three models:
| Model | Entry-level | Mid-market | Enterprise |
|---|---|---|---|
| Monthly retainer | $2,500–$4,000/mo | $4,000–$7,500/mo | $7,500–$15,000/mo |
| Pay-per-meeting | $125–$250 | $250–$450 | $450–$800+ |
| Expected meetings/month | 8–12 | 12–20 | 20–35+ |
Source: Outbound Sales Pro, Outsourced SDR Pricing 2026
A $5,000/month retainer delivering 10–14 qualified meetings implies a cost per meeting of $357–$500. Pay-per-meeting sits at $150–$600 for mainstream B2B ICPs, with enterprise and multi-region targets pushing past $900.
The UK market prices differently and slightly higher. ORRJO's figures put average UK cost per booked meeting at £1,200–£3,500, with well-run programmes at £700–£1,200 and the strongest at £500–£700. Their in-house comparison is brutal: a London SDR at £130K–£170K fully loaded in year one versus £40K–£90K for an agency programme at the same volume, with the crossover point around 12 meetings per month.
The benchmark table to compare yourself against
| Metric | Average | Good | Strong |
|---|---|---|---|
| Cold email reply rate | 1.9–3.4% | 4–6% | 7%+ |
| Meeting booked rate per outreach | 0.5–1.2% | 1.5–2.5% | 3%+ |
| Meeting attendance rate | 70–80% | 85–90% | 90%+ |
| Meeting → qualified opportunity | 30–40% | 45–55% | 60%+ |
| Opportunity → closed-won | 15–22% | 25–30% | 35%+ |
| Time to first qualified meeting (in-house) | 5–7 months | 4–5 months | 3–4 months |
Source: ORRJO State of B2B Outbound 2026, drawing on Bridge Group and Pavilion data
Note the reply-rate trend line underneath all of this. Martal Group's cold email data tracks the B2B average from 8.5% in 2019 to 5% in 2025 to 3.43% in 2026. That is not a messaging problem you can copywrite your way out of. It is inbox saturation, and it means every year the same volume of activity buys you fewer meetings at a higher cost.
Error one: you are counting booked meetings, not held ones
This is the single most common way teams flatter their own cost per meeting, and it is worth being blunt about.
Industry meeting attendance runs 70–80% across high-volume outbound. Well-run programmes hit 85–90%. The best hold 90%+.
Do the arithmetic. If you are paying a $5,000 retainer for 12 booked meetings, your headline cost per meeting is $417. At 72% attendance you actually held 8.6 conversations, and your real cost per held meeting is $581 — a 39% increase over the number in your dashboard. At 92% attendance the same spend produces 11 held meetings at $455.
ORRJO makes the point sharply: an agency reporting 20 meetings booked at 60% attendance delivered 12 actual conversations. An agency reporting 12 meetings at 95% attendance delivered 11. The first looks dramatically better on a slide and is dramatically worse as a programme.
Three questions to ask before you sign anything
If you are evaluating an outbound function, internal or external, ask these and insist on numbers rather than adjectives:
- What was the attendance rate over the last 90 days, calculated as meetings held divided by meetings booked?
- What is the qualification standard before a meeting reaches an AE's calendar, written down?
- Who reviews each booked meeting before it lands, and what is the replacement policy when one no-shows?
Programmes that cannot answer all three usually have a 70–75% attendance rate they would rather not publish. That is not necessarily malice. It is often just that nobody built the reporting, because booked meetings are easier to count.
Error two: cost per meeting is a checkpoint, not a destination
Cost per meeting is where you should start, but it is not where the money is decided. Cost per qualified opportunity is.
Run it through the full funnel. Take a mid-market retainer at $6,000/month producing 15 booked meetings:
- 15 booked × 78% attendance = 11.7 held meetings
- 11.7 held × 38% meeting-to-qualified-opportunity = 4.4 opportunities
- $6,000 ÷ 4.4 = $1,364 cost per qualified opportunity
Now run the same spend through a better-qualified programme:
- 15 booked × 90% attendance = 13.5 held
- 13.5 × 52% meeting-to-opportunity = 7.0 opportunities
- $6,000 ÷ 7.0 = $857 cost per qualified opportunity
Same budget, same booked-meeting count, 37% cheaper pipeline. Nothing in that second scenario required a new tool. It required better targeting, better qualification, and better pre-meeting discipline.
Conversion varies enormously by industry
The other reason cost per opportunity beats cost per meeting: downstream conversion is not uniform, and benchmarking against a generic SaaS number will mislead you badly if you sell into a regulated market.
| Industry | Meeting → opportunity | Opportunity → closed-won |
|---|---|---|
| SaaS (transactional, sub-3-month cycle) | 50–60% | 25–32% |
| SaaS (enterprise, 6+ month cycle) | 40–50% | 18–25% |
| Cybersecurity | 40–50% | 20–28% |
| Fintech / regulated | 35–45% | 15–22% |
| Professional services | 50–60% | 25–35% |
| Healthcare / MedTech | 35–45% | 15–22% |
| Manufacturing / industrial | 45–55% | 22–30% |
Source: ORRJO observed ranges, State of B2B Outbound 2026
A fintech team benchmarking its meeting-to-opportunity rate against transactional SaaS will spend two quarters concluding its SDRs are underperforming when the funnel maths were simply set wrong.
Why more volume now raises your cost per meeting
The traditional lever for a bad cost per meeting was to send more. In 2026 that lever is broken, and in several places it actively runs backwards.
Deliverability enforcement moved from filtering to rejection
Google, Yahoo and Microsoft have all moved their bulk sender requirements from recommended to enforced. The current bar: spam complaints under 0.3%, bounces under 2%, SPF and DKIM and DMARC all passing and aligned, DMARC at p=quarantine or p=reject rather than p=none, and one-click List-Unsubscribe per RFC 8058 honoured within 48 hours.
Two things changed the economics. First, the penalty for failing these checks is increasingly outright rejection rather than the spam folder — mail that fails does not arrive anywhere, and you do not get a bounce you can learn from. Second, the practical volume threshold has eroded. Filtering now applies prospecting-pattern scrutiny to lower-volume senders, so a poorly authenticated domain sending 200 cold emails a day can trigger the same treatment as a large newsletter.
The cost implication: domain and inbox infrastructure, warmup, rotation, seed testing and complaint remediation are no longer optional line items. They belong in your cost-per-meeting denominator, and they are part of why a cheap-looking DIY stack often produces a worse number than a retainer.
LinkedIn enforcement got behavioural
LinkedIn published no formal policy change, but enforcement against cloud-proxy automation visibly intensified through 2026. The most-discussed event was LinkedIn removing HeyReach's company page and banning its founder's personal profile in March, an action widely read as targeting the server-side architecture rather than one vendor's conduct. Northlight's Q1 analysis estimated that close to 40% of accounts running non-compliant automation tools picked up some form of restriction between January and March. LinkedIn's own March 2026 transparency reporting cited 78.2 million fake accounts blocked and 23.5 million automated sessions flagged in a single quarter.
The relevant point for unit economics: enforcement is behavioural, not categorical. LinkedIn does not restrict activity because it is automated. It restricts activity that looks like a bot — implausible action volumes, server-side sessions, no human rhythm. A restricted sender account costs you the account, the warm network attached to it, and every meeting it would have produced. That is the most expensive possible outcome per meeting, and it does not show up in any dashboard until it has already happened.
Autonomous AI SDRs post the worst numbers in the dataset
The AI SDR category was the loudest bet of 2024–2025 and the 2026 report card is unflattering. ORRJO ranks pure autonomous AI SDR outreach last on observed cost per qualified opportunity, behind even undifferentiated volume cold email. The supporting detail matters more than the ranking: Validity data cited in the report shows median sender reputation dropping 38 points within 90 days of scaling to agentic volumes, with reply rates from those accounts decaying more than 60% within 18 months. SaaStr's AI SDR reality check and an Indie Hackers review of eight AI SDR products describe the same pattern from the practitioner side.
The nuance is important, because "AI SDRs do not work" is too crude. What the data supports is narrower and more useful: AI used to research, monitor signals, build and score lists, enrich records and draft variations is consistently positive. AI used to autonomously write and send outreach at volume has not matched human-quality reply rates, and it damages the sending assets it runs on. The category's own marketing has quietly shifted from replacement to augmentation between 2024 and 2026, which tells you where the evidence landed.
The four levers that actually cut cost per meeting
None of these are tools. All of them are upstream of tools.
1. Shrink the target list
Programmes targeting 200–500 named accounts with documented ICP criteria consistently beat programmes targeting 5,000+ accounts filtered by industry and headcount. The volume programmes report higher absolute meeting numbers at two to three times the cost per meeting. That trade is almost never worth it once you price in AE time spent on unqualified conversations and the domain reputation burned reaching people who were never going to buy.
The uncomfortable version of this advice: if you cannot name the 300 accounts, you do not have an ICP, you have a filter.
2. Trigger on signals rather than on calendar
Programmes triggered on real events — recent funding, leadership change, public job postings, technology adoption, a competitor complaint in public — reply at roughly 2x the rate of generic campaigns running against the same accounts. Signal data costs money. In ORRJO's dataset the reply-rate lift consistently more than covers it.
This is the cheapest structural change available to most teams, because it does not require more headcount or more sending. It requires sequencing the same accounts in a different order. A prospect who viewed your profile last Tuesday, or whose company just posted three roles in the function you sell to, is the same person you were going to email in November. Reaching them this week costs the same and converts at twice the rate.
Capturing those signals reliably is the part teams underestimate — it means monitoring profile views, post engagers, competitor mentions, hiring pages and pain-point posts across LinkedIn, Reddit and X continuously rather than pulling a list once a quarter. This is exactly the problem Updately was built for: watching for warm intent signals, scoring the people behind them against your ICP, and putting them in front of a human with the research already done.
3. Warm the audience before the cold touch
The largest single lever in the ORRJO dataset is whether a prospect has any prior brand exposure before outreach arrives.
| Metric | Cold-only | Demand-warmed | Lift |
|---|---|---|---|
| Reply rate | 2–4% | 5–8% | 2–3x |
| Meeting attendance | 75–85% | 88–95% | +5–10pp |
| Meeting → qualified opportunity | 30–40% | 45–55% | +10–15pp |
| Cost per qualified opportunity | £3,000–£5,000 | £1,500–£3,500 | −40 to −50% |
The maths runs the opposite way to how most leaders think about it. Demand generation does not add cost on top of outbound. It reduces the per-opportunity cost of the outbound function, frequently by more than the demand spend itself. Cold outreach to a buyer who has never heard of you is cold twice over: unfamiliar message, unfamiliar brand. Founder-led content and employee advocacy remove one of those layers for a fraction of what another SDR costs.
This compounds with the buying-behaviour trend. Buying groups now routinely run to six or more stakeholders, and a large majority of the evaluation happens before anyone talks to a seller. If your first appearance is a cold email to one person, you are arriving late to a conversation that has already been happening.
4. Run coordinated multi-channel, not parallel single-channel
Coordinated sequences combining email, LinkedIn and phone against a tight ICP consistently produce the lowest observed cost per qualified opportunity. Teams running genuinely coordinated multi-channel book meetings at roughly half the activity-to-meeting ratio of single-channel teams.
"Coordinated" is doing real work in that sentence. Three channels firing the same generic message on independent schedules is not multi-channel; it is the same campaign annoying someone three times. Coordinated means the LinkedIn touch references what the email raised, the timing responds to what the prospect actually did, and the whole sequence reads like one person paying attention.
Build your own model this week
You can do this in a spreadsheet in under an hour. The point is not precision, it is honesty.
- Total the real monthly cost. Fully loaded comp for anyone touching top-of-funnel, plus tools, data, enrichment, inbox and domain infrastructure, plus a fair share of management time. Include ramp: an SDR drawing salary through a four-month ramp against zero pipeline is a real cost that most models omit entirely.
- Count held meetings, not booked ones. Pull the last 90 days. Divide meetings actually held by meetings booked and write the attendance rate on the same page as the cost.
- Divide. Cost ÷ held meetings = your true cost per meeting. Compare against $821–$1,150 in-house, $357–$600 outsourced.
- Go one stage further. Multiply held meetings by your meeting-to-qualified-opportunity rate, using your own industry's band rather than a generic SaaS average. Cost ÷ opportunities = cost per qualified opportunity. This is your real number.
- Sanity-check against ACV. If cost per qualified opportunity is more than roughly 10% of your average contract value and your win rate is average, outbound is not paying for itself yet. Fix targeting before you add volume.
- Track it monthly, alongside attendance. A cost per meeting that improves while attendance falls is not an improvement, it is a qualification standard quietly loosening.
Takeaways
- In-house cost per meeting runs $821–$1,150; outsourced runs $357–$600. If your internal number looks dramatically better than that, you are almost certainly counting booked meetings or excluding overhead.
- Attendance is where cost per meeting is really decided. The gap between a 72% and a 92% show rate is a 25–30% swing in your true unit cost, and it is fixable with qualification discipline rather than spend.
- Cost per qualified opportunity is the metric that should reach your board. Cost per meeting is a checkpoint on the way there, and it varies enormously by industry.
- Volume is no longer a lever. Deliverability enforcement, LinkedIn's behavioural detection, and the decay pattern in autonomous agentic sending all mean more sending buys fewer meetings at a higher risk-adjusted cost.
- The four levers that work are all upstream: a smaller named-account list, signal-based timing, a warmed audience, and genuinely coordinated multi-channel sequencing. Together they routinely halve cost per qualified opportunity without changing budget.
The teams that win on outbound economics in 2027 will not be the ones that sent the most. They will be the ones who knew, to the pound or dollar, what a held and qualified conversation cost them — and spent every improvement on relevance rather than reach.
Sources: ORRJO, State of B2B Outbound 2026 · Outbound Sales Pro, Outsourced SDR Pricing 2026 · Martal Group, B2B Cold Email Statistics 2026 · Bridge Group, SDR Metrics · SaaStr, The AI SDR Reality Check · Northlight, LinkedIn Automation Rules 2026