AI agents are now mainstream in sales, not a future bet. Salesforce's 2026 State of Sales report found 54% of sales teams use agents today, and nine in 10 use them or expect to within two years. But most teams still spend more than half their week on non-selling work, so the time AI agents save so far is concentrated in a few tasks, mainly prospecting, not the whole job.
Adoption has jumped since 2022
Salesforce's State of Sales report, 2026 edition, surveyed 4,050 sales professionals across 22 countries in August and September 2025 (1,417 sales leaders, 1,032 reps, 1,022 sales ops professionals, 579 sales support staff). It found: "Nine in 10 sales teams use agents today or expect to within two years." Broken down, 54% use agents now, 34% expect to within two years, 8% within five years, 3% don't expect to, and 1% don't know.
Compare that with the 5th edition of the same report, surveyed in 2022 (7,775 responses across 38 countries, fielded August 24 to September 30): "Only one-third of sales organizations currently use artificial intelligence (AI)," with an additional 20% of organizations planning to within two years.
Reported AI usage in sales roughly jumped from a third of organizations to over half of teams using agents specifically in about three years. The two surveys differ in sample and in asking about AI broadly versus agents specifically, so treat this as a directional trend rather than a precise before and after of the same panel.
Where the time actually goes
Adoption is not the same as time saved. The 2026 report found reps "spend more than half of their time on nonselling work like data entry and prospecting," with 60% of time on non-selling work and 40% on selling, and reps spending almost a full day of their workweek on prospecting alone.
The 2022 edition found reps spent "only 28% of their week actually selling," with manual data entry and administrative tasks each eating up 8.8% of the week.
| Metric | 2022 (5th edition) | 2026 edition |
|---|---|---|
| Organizations using AI / agents | 33% | 54% use agents now (90% now or within 2 years) |
| Share of week spent selling | 28% | 40% |
| Top non-selling task named | Manual data entry (8.8% of week) | Prospecting (about 1 day per week) |
Read together, both studies agree on one thing: most of a rep's week has never gone to selling, before or after agent adoption. What has changed is less the overall split than where agents are being pointed.
Agents are concentrated on prospecting, not full-cycle selling
The 2026 report found that among sales teams with AI agents, 34% use them specifically for prospecting, and 92% of sales professionals with agents say AI benefits their prospecting work. High performers are 1.7 times more likely than underperformers to use prospecting agents, and 94% of sales leaders with agents call them critical for meeting business demands.
The 2022 edition found a similar pattern at the team level: eight in 10 leaders and sales ops professionals with AI say it has improved use of reps' time at least moderately, and listed improved lead prioritization among the reported benefits, without a specific lead-scoring accuracy figure.
What this means if you're planning a rollout
- Expect the first measurable win in prospecting and qualification, not a rewrite of the whole week. The data from both reports points there first.
- Set a time baseline before you start. If you don't know today's split between selling and non-selling time, you can't credit an agent with changing it.
- Treat "high performers use agents more" as correlation, not proof. High performers may simply adopt more tools generally, not just agents.
- Scope narrowly. A single-job agent, one task, one system, is easier to measure and trust than an agent meant to help with sales in general.
How Lean Discovery Group helps
Lean Discovery Group builds custom AI agents for sales and lead qualification work through a five-step process: define the job (document inputs, outputs, success criteria and prohibited actions), design the tools and permissions the agent actually needs, build and evaluate it against real examples before it touches live data, deploy it with approval queues and logging for sensitive steps, then monitor and improve it over time. You own the code, workflows and data, and a free automation audit gives you a fixed scope and price before any build starts. Narrow, single-job agents can typically be piloted in weeks.
Prospecting and qualification agents usually connect into CRM work too. Our AI CRM automation service logs activity, updates deal stages and flags stalled deals automatically, and our AI lead response automation replies to new inquiries within seconds and books qualified meetings straight onto a rep's calendar. Book a discovery call to talk through where an agent would save your team the most time.
Frequently asked questions
How many sales teams actually use AI agents?
Salesforce's 2026 State of Sales report found 54% of sales teams use AI agents today, and nine in 10 use them now or expect to within two years, based on a survey of 4,050 sales professionals in 22 countries.
Has AI actually freed up reps' selling time?
Partly. Salesforce's data shows the share of the week spent selling moved from 28% in 2022 to 40% in 2026, but reps still spend more than half their week on non-selling work like data entry and prospecting.
What do sales AI agents mostly get used for today?
Prospecting. Among teams with AI agents, 34% use them for prospecting specifically, and 92% of sales professionals with agents say AI benefits that work, per Salesforce's 2026 report.
Should I roll AI agents out across the whole sales process at once?
The data suggests starting narrower. Both the 2022 and 2026 Salesforce reports show the clearest gains concentrated in prospecting and qualification rather than the full sales cycle.
How fast can a sales AI agent be piloted?
A narrow, single-job agent, such as one focused on lead qualification or prospecting, can typically be piloted within weeks when it is scoped to one task and one system.
