New research shows that a sales task that once required roughly 15 minutes of research and writing can now take less than two. According to Cubeo, AI-assisted email personalization can reduce the time required for the task by 87% while maintaining response rates.
Good news! Your sales rep just got 13 minutes back. Now what? Thirteen minutes doesn’t sound revolutionary. Multiply it across 20 emails. Then 50 reps. Then an entire quarter. Suddenly, we aren’t talking about writing emails faster. We’re talking about sales capacity. And that raises a much more important question for RevOps leaders: What should you do with all that time?
KEY TAKEAWAYS
How much time can AI save on sales email personalization?
Cubeo reports that AI can reduce a roughly 15-minute personalization task to under two minutes—an 87% reduction. The bigger opportunity is deciding how to reinvest that recovered seller capacity.
Does more AI-generated outreach automatically improve sales productivity?
No. Higher email volume is an activity gain, not necessarily a revenue gain. RevOps should measure meetings, pipeline, conversions and account engagement alongside activity.
Does the type of email personalization matter?
Yes. Gong’s analysis of more than 30,000 sales emails found substantial differences depending on the buyer and personalization strategy, including 3.3x reply rates for individual personalization with non-managers and 3x reply rates for company personalization with executives.
How could AI change sales capacity planning?
As AI reduces time spent on prospect research, email drafting and administrative work, RevOps teams can reconsider account loads, territory capacity, headcount requirements and where sellers spend their time.
How Much Time Do Sales Reps Spend Personalizing Emails?
Personalization has always presented sales organizations with an uncomfortable tradeoff.
Buyers want relevant outreach. Sellers know relevance matters. But producing it takes time.
Gong surveyed more than 600 sales professionals and analyzed more than 30,000 sales emails. Its research found that reps spend an average of 12 hours per week personalizing emails—5.9 hours customizing existing templates and another 6.2 hours writing messages from scratch.
That’s a day and a half of a 40-hour workweek.
And sellers aren’t doing it because they particularly enjoy spending Tuesday afternoon rewriting email templates. They believe personalization produces results.
Gong found that only 3% of sellers were happy with their email reply rates, while better personalization was the No. 1 way respondents believed they could improve them.
The problem has never been convincing sellers that personalization matters. The primary problem, here, is making good personalization economically scalable.
What Happens When 15 Minutes Becomes 2?
This is where the math gets interesting.
Cubeo estimates that AI can reduce an email personalization task from approximately 15 minutes to less than two—an 87% reduction in time.
Consider a rep preparing 20 personalized messages.
Old model:
20 emails × 15 minutes = 300 minutes
That’s five hours.
AI-assisted model:
20 emails × 2 minutes = 40 minutes
That’s 4 hours and 20 minutes returned to the rep.
Now scale the example across a 10-person sales team.
That’s potentially more than 43 hours of capacity returned to the organization for that batch of outreach alone.
Of course, that’s illustrative math, not a promise of 43 newly productive hours. Some of that time will disappear into meetings, CRM work, Slack messages and everything else competing for a seller’s attention.
But the larger point remains. When a meaningful sales activity suddenly requires a fraction of the time, RevOps has a capacity-planning question on its hands.
Myth: AI Productivity Means Reps Should Send More Emails
The easiest response to greater efficiency is greater volume. We can write emails faster? Great. Send more emails. That logic works beautifully in a spreadsheet. Buyers may feel differently.
But the goal of sales personalization was never to produce personalized emails. The goal was to create relevant conversations with the right buyers.
Gong’s analysis of more than 30,000 prospecting emails demonstrates why that distinction matters. Different forms of personalization produced very different results depending on the audience.
Individual-based personalization generated a 3.3x reply rate with non-managers. Company-based personalization produced a 3x reply rate with executives. Activity-based personalization generated a 3x reply rate and booked meetings. Industry-based personalization delivered a 1.8x reply rate across personas.
In other words, personalization isn’t a binary choice between “personalized” and “not personalized.”
What you personalize—and for whom—matters.
That changes the productivity conversation.
AI doesn’t simply make it possible to create more personalized outreach. It makes it possible for sales organizations to become much more deliberate about where human selling time produces the greatest return.
Reality: RevOps Just Got a New Capacity Lever
For years, sales capacity planning has centered largely on headcount. How many reps do we need? How many accounts can each rep cover? How much pipeline should each rep create? What quota can the territory reasonably support? AI adds another variable:
How much productive selling capacity can we create from the team we already have?
Salesforce’s 2026 State of Sales research suggests sellers themselves expect that equation to change. Sellers surveyed by Salesforce expect AI agents, once fully implemented, to reduce prospect research time by 34% and email drafting time by 36%. Top-performing sellers are also 1.7 times more likely to use prospecting AI agents for outreach than underperformers, according to Salesforce.
And there’s another important clue in Salesforce’s research: 80% of sales reps using AI say it’s easy to get the customer insights they need to close deals, compared with 54% of reps without AI. The advantage isn’t simply generating words faster. It’s getting the right information in front of the seller faster. That matters to RevOps because capacity isn’t just a headcount number anymore. It’s also a function of how much friction exists between a rep and the work that actually generates revenue.
Evidence: Personalization Still Needs Judgment
There’s a danger hiding inside all this efficiency. Bad personalization can now be created faster, too.
Give an AI tool weak CRM data, irrelevant signals or a poor understanding of the ICP and it can efficiently produce a lot of outreach that looks personalized without being particularly useful. “Congratulations on your recent LinkedIn post” at industrial scale isn’t much of a GTM strategy.
Gong’s research makes this particularly clear. Its analysis found that different buyers respond to different forms of personalization. An executive may respond to company-level context. Another stakeholder may respond more strongly to something relevant to their individual role. An existing account with recent engagement may require an entirely different approach. That’s why the most valuable question for RevOps isn’t:
How quickly can AI write the email?
It’s:
Do we have enough information to know what the email should be about?
AI can dramatically reduce the cost of execution. It doesn’t eliminate the need for good account data, territory design, ICP definition, engagement signals or sales judgment. In fact, faster execution may make those things more important.
Action: Decide Where the 13 Minutes Should Go
Suppose AI really does give a rep 13 minutes back on a personalization task. Where should those minutes go? Not automatically into another 10 emails. Maybe that time belongs with the five accounts showing the strongest buying signals. Maybe it belongs in deeper account research. Maybe it belongs in follow-up on stalled opportunities. Maybe it belongs in multithreading an account where the rep has only one relationship. Maybe it belongs in pipeline creation. Maybe it belongs in an actual conversation with a customer. This is where AI productivity becomes a GTM design problem rather than a software feature. The question isn’t how much activity can we generate? It’s where should the additional capacity go?
Stop Measuring the Productivity Gain in Emails Sent
This may require changing the scorecard. If AI dramatically reduces the time required to produce outreach, measuring success primarily by activity volume risks rewarding the least interesting benefit of the technology.
Instead, RevOps teams should consider what happens downstream.
Track measures such as:
- Positive reply rate
- Meetings created
- Pipeline created per rep
- Pipeline created per account
- Conversion by personalization strategy
- Opportunities created from target accounts
- Seller time per opportunity created
- Engagement across multiple stakeholders
- Pipeline progression after engagement
Imagine two reps. Rep A uses AI to send 300 additional emails. Rep B uses the same productivity gain to engage 30 carefully selected accounts and creates three qualified opportunities. Which rep became more productive? Activity tells you what the seller did. Revenue outcomes tell you whether the activity mattered.
The Bigger Question: Does AI Change Sales Capacity Planning?
This is where the implications extend beyond email. Cubeo’s examples include other formerly time-intensive activities becoming significantly faster with AI: competitive research, prospect research and proposal generation.
Salesforce’s 2026 research similarly points toward reductions in prospect research and email drafting. Stack enough of those efficiencies together and the traditional assumptions behind sales capacity begin to change.
A territory that once required 100% of a rep’s available capacity may no longer require it. A rep may be able to manage more accounts without sacrificing relevance. A smaller team may be able to cover a market that previously required additional headcount. Or—and this is important—the organization may decide not to increase account loads at all. It may choose to reinvest the capacity in better account penetration, stronger relationships and higher-quality pipeline. Those are fundamentally different GTM strategies. And choosing between them requires more than an AI tool. It requires RevOps.
The Real AI Productivity Question
There will be plenty of attention paid to how much time AI saves sellers. That’s understandable. The numbers are compelling. But “we made email 87% faster” isn’t much of a revenue strategy. The more consequential question comes next: What are you going to do with the other 13 minutes? That’s the opportunity for RevOps. AI may give sales teams more time. RevOps decides where that time should go.