What did the Richardson sales leader survey find?
On November 17, 2026, Richardson Sales Performance published The Revenue Edge, a survey of 1,105 senior sales leaders. 70% said humans outperform AI on the activities that matter most in complex deals. Only 4% said AI does. 88% expect the skills AI can't copy to separate their best sellers in 2027.
The respondents are big-company people. Every one holds a sales, general management or enablement role at vice president level or above, at an organization doing at least $750 million a year, in North America or Europe. Richardson fielded the survey from October 12 to 20, 2026, and the report page says the sample covers 14 industries. Nearly half the respondents are C-suite or senior vice presidents.
These are the people buying AI sales tools at scale, and 4 in 100 of them think the software beats their reps where the deal gets decided.
Why do some sales teams get more out of AI than others?
Teams that hit quota report much better AI returns. 62% of leaders whose organizations met or beat quota say AI paid back financially, against 39% of leaders below 90% of quota, according to the release. Same tools, very different results.
The gap shows up everywhere else too:
- 55% of top performers expect their forecast to land within 5%, against 27% of laggards.
- 45% of top performers coach at least every two weeks, against 27%.
- 52% of laggards kept an underperforming seller longer than they wanted, against 36% of top performers.
- 57% of top performers say their last training investment was worth more than it cost, against 33%.
Richardson flags one limit. Quota attainment is self-reported, so every comparison is correlational. Nobody proved that AI made the winners win. My read after running ads since 2013 is simpler. The teams with a tight process (they coach, they forecast honestly, they cut people who don't perform) get more out of AI because they already know which job they want it to do. Put AI on a messy sales process and you get the same mess, only faster.
Does a survey of big companies apply to a small business?
Not as a statistic, but the split maps cleanly. A small business has complex deals too: a $15,000 roof, a $6,000 dental case, a monthly retainer. The human part is the conversation where the buyer decides to trust you. Everything before that conversation is speed and consistency, and software is better at both.
Most owners I talk to have this backwards. They spend their own time chasing leads, calling back three hours late and sending reminder texts by hand, then show up tired to the one call that makes money. The survey's answer, read for a 5-person company, says to hand the front end to software and protect the human hours for the close.
Which parts of the sale should AI handle and which should stay human?
Give AI the work that rewards speed and repetition: answering, first questions, booking and reminders. Keep humans on discovery, reading hesitation, the price talk and the close. That is roughly where the 70% of sales leaders in the Richardson survey drew the line.
| Stage | Who handles it | Why |
|---|---|---|
| Answering calls and texts, nights and weekends | AI agent | It picks up in seconds and never lets a lead sit until Monday |
| First qualifying questions (timing, budget, location) | AI agent | The same questions get asked every time, and the answers land in the CRM |
| Booking and reminders | AI agent plus automations | Fast booking is what gets a buyer to show up |
| Discovery on the call | Human | Buyers open up to a person who listens and asks the next question |
| Price and the close | Human | Trust gets built here, and the survey says this is where people win |
| Follow-up after a no or a ghost | Automation, then a human on any reply | Software keeps the thread alive and a person takes it back when it warms up |
Our own calendar data backs up the booking row. Over 45 days, calls our AI agents booked went 10 showed and 14 no-shows. The people who showed were mostly booked within 24 hours. The no-shows were mostly booked 2 to 5 days out. A fast AI booking puts a human closer in front of a buyer while they still care. If your leads wait, read why leads go cold before you buy anything.
This is how we set up KenjiAI. Voice AI agents and Conversation AI answer calls and texts, ask the qualifying questions and book the appointment. The CRM pipeline hands your closer the notes before the call. Your people spend their hours on the conversation that the Richardson survey says humans still win. If you want to see it on your own leads, book a call with us.
What should a business owner do with this news this week?
Find out where your deals actually die, then put AI or coaching on that spot. If leads die before anyone talks to them, fix speed with automation. If they die on the call, fix the human with coaching. These steps take about an hour.
- Pull every lead from the last 30 days and mark where each one stopped: never reached, booked but no-show, or talked and didn't buy.
- Count each bucket. Never reached and no-show are speed problems, and they belong to software.
- Talked and didn't buy is a people problem. Copy the top performers in the survey and review one recorded call per seller every two weeks.
- Decide how much rep time a deal deserves by knowing what a customer is worth. Our free LTV calculator at kenjiai.com/tools/ltv-calculator gives you that number in a minute.
Start with step one today. The bucket with the most dead leads tells you whether your next dollar goes to AI or to coaching.