Is the SDR role shrinking? AI and SDR hiring in 2026

Yes, the SDR role is shrinking in headcount, but it is not disappearing. In SDR hiring for 2026, the pattern is fewer, more senior sales development reps who use AI for research and first-touch volume, and who are judged on qualified pipeline rather than on activity counts.

That is the short answer. The longer one matters more if you are a founder deciding whether to hire two SDRs or buy an "AI SDR" licence, or a rep wondering whether sales development is still a sensible first job in software sales. This article sets out what the data shows up to the start of 2026, what AI does well and badly in outbound, how team design is changing, and how to hire and pay the SDRs you still need.

What does the data say about SDR headcount?

The clearest signal comes from venture-backed software companies. Emergence Capital's "Beyond Benchmarks" survey of more than 560 B2B software companies, run in April 2025 and reported by SaaStr, found that 36% had reduced SDR/BDR headcount over the previous 12 months and only 19% had increased it. Sales development had the highest rate of cuts and the lowest rate of growth of any sales role in the survey. By contrast, account executive headcount was broadly flat (25% cut, 28% grew) and professional services grew (34% increased).

Two other numbers explain why boards have been asking hard questions. The Bridge Group's 2025 SDR Models, Motions & Metrics report, published in February 2025 from 351 B2B companies, found that only 60% of SDRs were at quota, the lowest figure in the study's history. Median SDR OTE had sat at $80,000 since 2022. Flat pay plus falling attainment is what a squeezed role looks like.

Yet the same report showed something that gets lost in the "SDRs are dead" posts: ramp time fell to an average of 3.0 months, the lowest since 2010, and each SDR was still associated with a median of $3.78 million in raw pipeline. Sales development still produces pipeline at a cost most companies cannot get any other way. The question is how many people you need to do it, and what kind.

Why is AI shrinking SDR teams?

Look at what an SDR's week used to contain: building lists, finding emails and phone numbers, researching accounts, writing first-touch emails, sequencing follow-ups, logging activity, and then a smaller slice of live conversations and qualification. AI tools now do most of the first five reasonably well and very cheaply. A large share of the job that justified a 1:2 SDR to AE ratio was list-building and copy-writing, and that share has collapsed.

The buyer side is pushing in the same direction. Gartner's survey of 632 B2B buyers, published in June 2025, found that 61% prefer a rep-free buying experience and 73% actively avoid suppliers who send irrelevant outreach. 6sense's 2025 Buyer Experience Report, published in November 2025, found that 94% of buying groups had ranked their preferred vendors before first contact with a seller, and the vendor favoured before contact won 80% of the time. If buyers decide most of the shortlist before they speak to anyone, more generic outbound volume does not help. It hurts.

So the economics have flipped. When outreach was expensive to produce, volume was a moat. Now that anyone can send 5,000 "personalised" emails a day, volume is noise, inboxes are filtered harder, and the scarce thing is relevance: the right account, the right person, the right reason, at the right time.

What do AI SDR tools do well, and where do they break?

"AI SDR" covers a spread of products, from research assistants that summarise an account to fully autonomous agents that build lists, write sequences and book meetings. The Bridge Group's 2025 survey was the first to record "AI SDRs" as a category, and only 1% of respondents reported using them as part of their sales development model. Adoption of AI-assisted work by human SDRs is far wider than adoption of AI that replaces them.

The failure modes are predictable. Fully autonomous outbound tends to burn domains and brand reputation when it sends at volume to poorly fitted lists. Agents miss context a junior human would catch (a recent layoff, a competitor just signed, a champion who has left). And meetings booked by bots convert to opportunities at lower rates if the qualification is shallow, which pushes cost onto your AEs' calendars.

Our view, from the hiring side: use AI to remove the low-value hours from every SDR's week, then hire fewer, better SDRs to spend those hours on conversations and qualification. Teams that tried to replace humans entirely tend to come back to the market for senior SDRs within a couple of quarters.

How is SDR team design changing in 2026?

Four changes show up repeatedly in the briefs we take and the plans founders share with us.

• Ratios are widening. With AI doing research and sequencing, one SDR can support more AEs. The Bridge Group's average of 1 SDR to 2.4 AEs is a starting point, not a target; we see more teams planning on 1:3 or 1:4 for outbound and pairing that with automated inbound handling.
• Fewer "pure" entry-level seats. The first job in sales development increasingly asks for some evidence of prior commercial work, research skill or domain knowledge, because the easiest tasks have been automated away.
• More senior SDR and team lead roles. Companies keep the people who can run account-based plays and coach others, so demand, and pay, shifts up the ladder.
• Metrics move down the funnel. Meetings booked is giving way to opportunities accepted by the AE, pipeline value created and, in more mature teams, pipeline that converts to revenue.

Reporting lines are moving too. The Bridge Group found 60% of SDR teams reported into sales, 26% into marketing and 10% into a C-level leader. As demand generation teams take more ownership of signals and intent data, we expect more inbound SDR functions to sit closer to marketing, with outbound staying under sales. Either works if one leader owns the pipeline target end to end.

What skills should an SDR have in 2026?

The profile has shifted from "high activity, thick skin" to "high judgement, thick skin". Resilience still matters; cold calling has not become pleasant. But the differentiators are now:

• Research and prioritisation: choosing which 30 accounts deserve real effort this week, using intent signals, hiring data and triggers, and explaining why.
• Writing that reads as human: short, specific messages that reference something real about the buyer's situation.
• Live conversation: confident cold calls and discovery, because the phone is where AI is weakest and where good SDRs now stand out.
• Qualification discipline: knowing a framework (BANT, MEDDICC or the company's own) well enough to disqualify, not just book.
• Tool fluency: working with AI research and sequencing tools, checking their output, and fixing it when it is wrong. RevOps-minded SDRs who can build a list in Clay or Apollo and spot data quality problems are in demand.
• Commercial context: enough understanding of the product and buyer to hold a credible first conversation with a director-level buyer.

How should SDRs be paid in 2026?

Pay has been flat at entry level and is rising for senior SDRs and leaders, which matches the shift in demand. Our 2026 salary guide sets out indicative ranges for software companies:

These are indicative ranges from the TRG GTM Salary Guide 2026, built from published sources including The Bridge Group and RepVue. The usual split is around 70/30 base to variable in the UK and 65/35 to 70/30 in the US (The Bridge Group's median is 68:32). Lower-ACV businesses typically pay less on OTE; see the full salary guide for location adjustments.

Plan design matters more than the headline. If AI is producing meetings, paying SDRs on meetings booked encourages them to pass anything through. Better plans pay on meetings held and accepted by the AE, or on qualified pipeline value, with a small kicker when that pipeline closes. That ties the SDR to quality without making their pay depend on a six-month sales cycle they cannot control.

Should you hire an SDR or buy an AI SDR?

For most Series A and B companies, the answer is "both, in sequence". Use this decision path before you open a requisition:

A rule of thumb we use with founders: if your average contract value is low and your buyers are happy to self-serve, lean heavily on automation and product-led motions. If deals are mid-market or enterprise, involve several stakeholders and need a real conversation to start, a small team of strong SDRs remains one of the best investments you can make.

How do you interview SDRs when AI writes the emails?

If every candidate can produce a polished sample email with an AI tool in two minutes, the written exercise has to change. We recommend testing the parts AI cannot fake:

What does this mean for SDRs and people wanting the job?

The entry door is narrower, but the role on the other side of it is better. Fewer seats means more competition for first jobs, so candidates who show research skill, a few real cold-call reps and fluency with the tools stand out. Our advice to anyone in or targeting sales development:

• Become excellent on the phone. It is the skill AI has not taken, and hiring managers notice.
• Learn the tooling deeply, including how to check and correct AI output. Being the person on the team who builds the best lists is a real advantage.
• Track your numbers in pipeline and revenue terms, not activity. "I sourced £1.2m of qualified pipeline, of which £300k closed" beats "I made 80 calls a day".
• Plan your next step. With average SDR tenure around 1.9 years (The Bridge Group), know whether you are aiming for an AE seat, a team lead role or a move into RevOps or marketing.

If you are weighing your next move, our SDR and BDR specialism page explains what companies are hiring for, and you can register your CV to hear about roles before they are advertised.

So, will the SDR role exist in five years?

We think the job title will survive, but the work will look more like a junior account-based seller or "pipeline analyst who calls" than the volume-dialler of the 2010s. The companies that hire well in 2026 will run smaller teams of stronger SDRs, with AI handling volume and humans handling judgement and conversation. The companies that get it wrong will either keep a large, low-attainment team doing work a tool now does, or cut too deep and find their AEs spending half their week prospecting.