Salesforce State of Sales Report 2026: What It Found and What to Automate First

The report at a glance

I set up outbound for B2B teams, so I read a report like this for one thing: what it changes about the first month of work. Here is what Salesforce surveyed, from its announcement and the full report, both checked on 15 October 2026.

The report is free on Salesforce's State of Sales page.

Key findings, in Salesforce's numbers

Adoption

• 87% of sales organizations use some form of AI, for tasks like prospecting, forecasting, lead scoring or drafting emails.
• 54% of sellers say they've used agents, and nearly 9 in 10 plan to by 2027. In the report's own chart, 54% use agents now and 34% expect to within two years.
• 94% of sales leaders with agents say they're critical for meeting business demands.

Time and prospecting

• The average seller spends 40% of their time selling. The rest goes to work like prospecting, data entry, quotes and planning.
• Reps spend nearly one full day a week on prospecting, and 47% say their team lacks the bandwidth to do cold outreach (the announcement rounds this to 48%).
• Once agents are fully implemented, sellers expect them to cut prospect research time by 34% and email drafting by 36%.
• High performers are 1.7 times more likely than underperformers to use agents for prospecting, and 92% of sales pros with agents say AI benefits their prospecting.

Data and tools

• 51% of sales leaders with AI say disconnected systems slow down or limit their AI initiatives.
• 46% of sales pros with agents say data quality issues hurt their sales. The top data problems: manual errors, duplicate data, security concerns, incomplete data and corrupt data.
• Only about a third of sales teams run on one platform. The rest use a mix of standalone tools, eight per team on average. 42% of reps say they're overwhelmed by too many tools.
• 74% of sales teams with AI are prioritizing data hygiene: 79% of high performers against 54% of underperformers.

Salesforce's own example

Adam Alfano, Salesforce's EVP of Sales, writes that its agents now work the leads its sales team used to leave untouched. In four months they contacted 130,000 leads and created 3,200 opportunities. That's about 2.5 opportunities per 100 leads contacted, reported by a vendor about its own product, with nothing on how many turned into revenue.

Three things to keep in mind before you quote it

1. It's a vendor survey. Salesforce sells Agentforce, its agent product, and the report closes with four steps to get started with agents. The numbers describe what respondents say and expect, not measured performance.
2. The 34% and 36% are expectations. Salesforce's wording is that sellers "expect" these cuts once agents are fully implemented. Don't put them in a business case as savings you'll get.
3. Most respondents don't look like a B2B outbound team. Retail and financial services are the biggest industries in the sample, and the top agent use cases are fulfilling orders, tracking product usage, creating quotes and managing commissions. Prospecting comes fifth, and 34% of sales teams with agents use them for it.

What the report says to fix before you automate

The most useful part of the report isn't the 34% or the 36%. It's chapter two, "AI Agent Adoption Requires Better Data and Fewer Tools". An agent that reads a CRM full of duplicates and stale job titles will research the wrong person faster. Alfano says it himself in the announcement: agents without the full customer context "tend to fail".

For a B2B sales team, that means a week of plumbing before any agent:

1. Count your tools. List every tool that holds contact or account data. If the same person lives in four places, decide which one is the record.
2. Fix the top two data errors. The report's list is a fair starting point: manual errors and duplicates. Merge the duplicates and make the fields your reps skip mandatory.
3. Write down your baseline. Hours per rep per week on research and on writing emails, before you change anything. Without it, you can't tell whether you got 34%, 10% or nothing.

What to automate first, in the order the data points to

The report's numbers line up with how I'd sequence it anyway: start where a mistake is cheap and move closer to the customer as trust builds.

Research goes first because a wrong research note costs a rep a minute, while a wrong email costs a reply. If I were you, I would first map out the companies, then move to people, with an agent doing the reading in both steps.

This list is short on purpose and tied to the report. For the full ranked list, with CRM logging, call notes, scheduling and follow-ups, see 8 sales workflows worth automating first. For where to stop, see what GTM tasks you should never automate.

How to know it worked

Automate the work, not the thinking. AI owns the speed, you own the judgment. To see whether the speed showed up, track four things against the baseline you wrote down:

• Hours: research and writing time per rep per week.
• Edit rate: the share of AI drafts a rep sends with little or no change. If reps rewrite most drafts, the 36% isn't happening.
• Where the time went: the report has sellers selling 40% of their week. If the freed hours don't show up as more conversations, the saving stayed on paper.
• Results: replies and meetings from accounts with automated research against similar accounts researched by hand.

Once a task holds up on those numbers for a few weeks, give the agent the next one. I lay out the gates for that in my framework for adding AI agents to your GTM stack.

FAQ