Scroll through any GTM newsletter and you'll see a stat claiming 40%+ of enterprise sales teams have already deployed AI SDRs in production. It's a useful number if you're selling AI SDR software. It's also hard to square with the most methodologically solid benchmark data available: The Bridge Group's 2025 SDR Models & Metrics Report, based on 351 B2B companies (78% North America, 83% SaaS), found that 2025 was the first year "AI SDRs" even appeared as a distinct category in its survey, and adoption sat at just 1% of respondents. That's not a rounding-down of a bigger number. That's the honest state of adoption among the companies actually reporting real SDR metrics, not the companies with a product to sell around the hype.

Worth remembering next time a vendor deck cites a huge adoption percentage without naming a checkable primary source: ask what the denominator is and who ran the survey.

What the real numbers say about SDR economics

Setting the AI SDR noise aside, Bridge Group's report gives a grounded picture of what running a human SDR function actually costs and looks like operationally in 2025:

  • Median SDR on-target earnings sit at $80,000, roughly a 68:32 base-to-variable split.
  • Ramp time to full quota productivity is down to 3.0 months, the fastest ramp time recorded in the report's 15-year history.
  • Average SDR tenure is 1.9 years.
  • Annual SDR turnover runs at 40%.

That last two figures matter more than they get credit for. A 40% annual turnover rate against a sub-two-year average tenure means a meaningful share of any SDR team's capacity in a given year is going toward ramping replacements rather than running at full productivity. Faster ramp times are a genuine bright spot, the fastest in 15 years of Bridge Group data, but they're offset by a workforce that keeps churning before it stays ramped for long.

On cost-per-meeting: be suspicious of any single number

You'll find plenty of tables online promising a precise cost-per-meeting-booked figure for cold email versus cold calling versus LinkedIn versus paid ads, often broken down to the dollar. Treat these skeptically. Independent 2026 vendor write-ups on this topic disagree with each other by an order of magnitude, some pricing optimized cold email meetings in the $40-$150 range, others pricing unoptimized cold calling as high as $400-$1,200 per meeting depending entirely on tooling, list quality, and how "fully loaded cost" gets calculated. The spread itself is the finding: cost-per-meeting is too dependent on your specific stack, target market, and campaign quality to be usefully summarized as one number you can copy from a blog post. If a source hands you a clean, precise per-channel cost table without disclosing its methodology or sample, that's a reason for caution, not confidence.

What does hold up: channels work better together than alone

One finding that shows up consistently across independent 2026 benchmark sources, even though the exact size of the lift varies by source and none of the specific multipliers could be pinned to a single fully verifiable report, is that combining email, phone, and LinkedIn into one cadence reliably outperforms running any single channel alone, generally landing in the range of roughly 2x or more booked meetings per prospect pool compared to cold email by itself. That directional finding is consistent enough across multiple independent write-ups to be a reasonable planning assumption, even without a single authoritative number to cite.

This tracks with the operational reality most SDR leaders already sense: a prospect who's seen your name on LinkedIn and gotten a call attempt is a different read on a cold email than someone hitting your message with zero prior context. Sequencing across channels doesn't just add volume, it changes how the prospect interprets each individual touch.

None of this requires exotic tooling to act on. It requires treating email, call, and social touches inside the same cadence as a single coordinated sequence rather than three disconnected activity quotas tracked in separate spreadsheets. Most of the ROI gap between teams isn't which channels they use, it's whether the channels are actually talking to each other around the same target account.

What this means for planning 2026 outbound spend

A few conclusions that are actually defensible from the data above, as opposed to the versions floating around in vendor marketing:

  • Don't budget around a 40%+ AI SDR adoption assumption. The most rigorous available survey data puts real adoption at 1%. If competitors are ahead of you here, it's a small minority, not the market.
  • Do budget for turnover. At 40% annual turnover and under two years of average tenure, your SDR function's effective capacity is meaningfully lower than headcount times months employed. Build hiring and ramp pipeline accordingly.
  • Don't anchor to a single cost-per-meeting figure from any one vendor report. Build your own baseline from your actual costs and treat published numbers as context, not targets.
  • Do invest in cross-channel sequencing rather than pure-play cold email or pure-play cold calling. The multi-channel advantage is one of the more consistently observed findings across independent sources, even without one canonical multiplier attached to it.

Whatever mix of channels you run, the email touches inside that sequence still carry a disproportionate share of first impressions, and with SDR capacity this expensive and this hard to ramp before it churns, it's worth making sure every one of those emails is actually earning its send before it goes out, not after the reply rate comes back low.