Automation is happening more quickly than it has in years for sales teams. Software is progressively replacing humans in prospecting, data input and follow-up tasks. The speed at which that is occurring has forced even seasoned sales leaders to adapt.
However, the teams with the most automation may not be the ones achieving the greatest success in 2026.
They are the ones combining it with better human coaching. That is becoming a much more important dividing line.
The 2026 Shift: AI Agents Are Doing More of the Sales Workflow
AI in sales has moved beyond the “helpful assistant” stage. Companies are now using agents that execute parts of the sales process rather than simply suggesting what someone should do next. You can see that change in how quickly adoption has spread across sales teams this year.
From AI Assistant to AI Agent
The big difference is execution. An assistant makes a suggestion and waits. An agent can research a prospect, update the CRM and draft a follow-up without someone prompting it every time.
Salesforce’s 2026 State of Sales research found that 87% of sales organisations now use some form of AI across the sales cycle. Among sales leaders who have deployed agents specifically, 94% say they are essential to growth.
What Reps Get Back: Time for Higher-Value Work
Automation only matters if people actually get something useful back from it. According to Microsoft’s 2026 Work Trends Index, 58% of AI users are now doing work they wouldn’t have been able to do a year ago.
For salespeople, this means that less time can be spent on analysing data, and more time is left for what machines can’t do yet, which is talking to buyers.
The New Core Question for Sales Leaders
The old question was whether AI would replace salespeople. That is becoming less useful.
The better question now is what salespeople need to get better at once software takes over more of the repetitive work.
The Coaching Gap: Why More AI Data Doesn’t Mean Better Reps
AI agents can give sales managers an enormous amount of information. They can flag missed objections, summarise calls and identify patterns across hundreds of conversations that no manager could realistically review alone.
But having that information does not automatically change anything.
If a rep talks too much during discovery, knowing that is only the starting point. Someone still needs to explain what happened, show them what better sounds like and give them another chance to try it.
That gap between seeing a problem and actually changing behaviour is where coaching comes in. Salesforce’s own research also identifies coaching as an active use case for AI agents, alongside prospecting and admin tasks.
Managers may have more sales data than ever. They do not suddenly have more hours in the day to coach everyone.
Human Oversight Is the Bigger 2026 Story, and Sales Should Pay Attention
The wider AI conversation is also putting more emphasis on human control. That became particularly visible this month, when Microsoft published a draft Humanist AI Code of Conduct on 14 September 2026.
The document says Microsoft’s AI systems should never resist correction or shutdown, and that violating that principle should be treated as a failure.
That is not a sales story.
As software becomes more autonomous, businesses need to be clearer about where human judgement remains. In sales, that does not mean checking every AI-written email. It means deciding who owns the relationship, the tone of a conversation and the decision on a deal.
Two AI Layers Sales Teams Will Need
Most sales organisations probably will not choose between automation and coaching. They will use both, because they solve different problems.
- AI agents that execute the workflow. Research, data entry, follow-ups and other repetitive jobs that used to take up a rep’s day.
- AI coaching that develops the people running it. This looks at conversations at a scale a manager cannot manage alone, then turns those patterns into specific feedback rather than another dashboard.
That second layer is where AI coaching software can be useful.
They are not there as a substitute for the manager’s judgement. The purpose is that the manager has something tangible to deal with ahead of the next call.
Why This Isn’t Another “SaaS Is Dying” Story
There was a point when AI agents looked like they might make traditional sales software irrelevant. That argument has softened.
The Financial Times recently described the shift away from fears of a “SaaSpocalypse” towards a market where software that integrates AI effectively is holding its value, rather than simply being replaced by it.
Sales software is not disappearing because of agents. It is being rebuilt around them. The products that matter now are the ones combining automated execution with a real answer to the coaching gap, instead of treating AI as another feature bolted onto an old system.
What Sales Leaders Should Do Now
For sales leaders, the practical question is fairly simple. Look at where agents are already doing rep work and then be honest about which skills still need deliberate development, especially things like handling objections, reading a buyer and building trust.
The data from AI should feed those coaching conversations, not replace them.
The teams that get the balance right will not necessarily be the ones that automate the most. They will be the ones that use automation to buy back time, then spend that time making their people better at selling.
