How to Set Sales Quotas When AI Does Prospecting

Most sales teams still measure their reps the way they did a decade ago: calls dialed, emails sent, meetings booked. That math made sense when a human did all the prospecting by hand. It falls apart the moment an AI agent takes over the volume.

Here is the part worth sitting with. Salesforce found that reps spend less than 30% of their day actually selling; the rest disappears into research, list-building and data entry. And the people stuck doing that low-value work keep leaving: the Bridge Group puts median annual SDR turnover at 32%. So we hire someone to sell, bury them in admin, then act surprised when they burn out and walk. Pushing harder on the same activity numbers will not fix that.

This guide lays out how to set quotas for a team where AI runs the prospecting grind and your salespeople own the conversations that close. It is built on Pair Selling, AvairAI's model for splitting the work between AI and humans, and you can put these structures to work this quarter.

The short version

• When AI owns the volume, measure reps on what they control: pipeline created, conversion and closed revenue, not dials and sends.
• Expect more pipeline. Move your coverage target from the old 3x rule of thumb toward 4x or 5x, because AI-sourced interest needs more human qualification.
• Recalibrate every quarter. AI output climbs over time, so reset the human targets instead of punishing your team for the new baseline.
• The payoff is documented: Gartner found sellers who partner with AI are 3.7 times more likely to meet quota.

Why activity quotas break when AI does the prospecting

Old quotas reward motion. A hundred dials, two hundred emails, fifteen meetings a month. Those targets were a decent proxy for effort back when effort was the bottleneck. The bottleneck moved. Manual prospecting now eats the majority of a rep's week, and an AI agent can run that outreach at a scale and consistency no human team can match. Teams adopting it are already seeing it land: Salesforce reports that sales teams using AI are 1.3x more likely to grow revenue.

So when you track "calls made" on a team where AI makes the first touch, what are you actually measuring? Usually, your best closer gaming a number. They hit the dial count and miss the revenue target. They fire off emails nobody opens to keep a dashboard green.

The real division of labor is simpler than the metrics suggest. AI is good at volume, coverage and never having an off day. People are good at judgment, trust and reading a room. Measure a salesperson on the machine's strengths, and you demoralize them while leaving their actual talent on the table.

Split the scoreboard: who owns which number

Pair Selling gives you a cleaner way to assign accountability: let each side own the metrics it can actually move. Think of a relay. AI runs the opening leg, the prospecting, and your reps run the anchor leg, the relationship and the close. You do not grade the sprinter on a handoff they never touched.

What the AI should own

These are the prospecting numbers, and they should be steady:

• Outreach volume and account coverage: how many of the right contacts get reached across your target accounts.
• Engagement: opens, clicks, replies and connect rates.
• Interested leads delivered: the count of prospects who respond with genuine interest, your MQLs.

If these slip, that is a targeting or messaging problem to fix in the system, not a performance issue to pin on a person. For the full metric set, see how to measure AI SDR performance.

What your reps should own

Here the scoreboard shifts to revenue and relationships:

• Lead-to-opportunity conversion: how many of those interested leads a rep qualifies into real, working opportunities.
• Meetings booked and held: the rep does the booking and the discovery off AI-sourced interest.
• Deal velocity: how fast an opportunity moves through the pipeline.
• Multi-threading and close rate: champions built, executives engaged, deals won.

Notice what moved. Reps are no longer paid to manufacture activity. They are paid to turn AI-sourced interest into revenue, which lines incentives up with the business instead of the dashboard.

Recalibrate the ratios for a hybrid team

With AI feeding the top of the funnel, your old benchmarks need a reset.

The familiar rule of thumb is 3x pipeline coverage: a rep carrying a $100,000 quota needs about $300,000 in open pipeline to land it. When AI prospects around the clock and works every target account with the same diligence, you tend to generate more pipeline, so push the target higher, toward 4x or 5x. This is not padding for its own sake. AI-sourced interest is more abundant, and some of it needs more human qualification before it is real, so the wider coverage absorbs that variance and still leaves each rep enough genuine opportunity to hit the number.

Add one metric: lead-to-opportunity conversion

Most leaders trip on the same assumption: AI surfaces more interested leads, so reps should close proportionally more deals. Not all interest is equal, and a rep's hours are still finite.

So watch lead-to-opportunity conversion, the share of AI-delivered interested leads your team turns into qualified opportunities. It tells you where a problem actually lives:

• Under 25%: the targeting or messaging needs work. Fix the system before you coach the rep.
• 25% to 40%: a healthy range for most B2B teams.
• Above 40%: your reps are converting well, and you can likely widen AI outreach to feed them more.

Read it before you assign blame. Low conversion points upstream, to who AI is reaching and what it is saying. High conversion paired with weak close rates points to coaching.

A quick worked example

Say a rep used to self-source 12 meetings a month and close 3, a roughly 1-in-4 win rate. Move prospecting to AI and that rep now receives 40 interested leads a month. The tempting reaction is to multiply the old close number and hand them a quota of 10. That ignores reality: a rep can only run so many real conversations, and a chunk of those 40 leads will need more qualification before they earn a calendar slot.

Set the quota off conversion instead. If 30% of 40 interested leads become opportunities, that is 12 working deals, and at the rep's historical 1-in-4 close rate you would expect about 3 wins, not 10. Start there, watch the actual numbers for a quarter, then raise the target as the rep and the AI both find their rhythm. Quotas built on conversion stretch a team; quotas built on wishful multiplication break it.

Pay people for what they control

A quota structure only holds if pay follows it. The old per-meeting bonus, say $200 a demo booked, actively backfires now: it pays reps to chase volume right when AI already supplies the volume.

Tie variable pay to outcomes a human genuinely controls: qualified opportunities created rather than raw meetings, pipeline advanced from stage to stage, a slice of closed revenue, and a multiplier for larger or strategic deals. That rewards the work AI cannot do, the judgment about which opportunities deserve a rep's hours and the relationships that actually close. For a deeper build, see our framework for redesigning SDR compensation on a hybrid team.

One workable model for an AI-augmented team:

• Base salary at 60% to 70% of total comp, higher than the traditional 50% to 60%, because the activity grind that used to justify a thin base is gone.
• Variable pay split roughly 40% on qualified opportunities created, 30% on pipeline value generated, 20% on lead-to-opportunity conversion and 10% on team revenue influence.

Top performers still earn well above the pack. They just earn it on quality of judgment rather than quantity of dials.

Roll it out without spooking the team

Change a comp plan mid-quarter and every rep assumes they are being set up to fail. How you communicate the shift matters as much as the metrics themselves.

Frame it around what reps gain. The grind they hate, the list-building and the cold first touches, moves to AI. Their day shifts to conversations and relationships, the part most salespeople actually like and are good at, and the new bonuses reward exactly that. Be honest that the change will not suit everyone equally: reps who lived on activity may wobble at first, while strong closers who hated prospecting will finally get to run.

Do not flip the switch and walk away. Give it about 90 days:

1. Weeks 1 to 4: measure the new metrics with no pay attached. Learn where AI output and human output land versus your guesses.
2. Weeks 5 to 8: calibrate. Set targets off real AI output, stretchy but reachable.
3. Weeks 9 to 12: attach compensation and keep a weekly review to catch problems early.

The trap to avoid is the demoralization spiral. If AI suddenly delivers far more interested leads than your team has ever handled and you set quotas as if every one converts, people miss, morale drops, and good reps leave, which only feeds the SDR churn AI was supposed to relieve. Start conservative and raise targets as you learn what is genuinely achievable. For more on leading a blended team day to day, see our playbook for managing a hybrid human-AI sales team.

The bottom line

Setting quotas in an AI-prospecting world comes down to one move: measure each side on what it controls. AI owns the activity, the volume, the coverage and the response rates. Your reps own the outcomes, the conversion, the deal progression and the close. Run them as partners, the way Pair Selling treats AI and humans, and the team reaches numbers neither would hit alone.

The leaders who get this right will not just stop their reps from gaming a dashboard; they will keep their reps. And the upside is on the record: Gartner found that sellers who partner with AI are 3.7 times more likely to meet quota than those who do not.

If you want to see how the math changes when AI runs the prospecting from just your website, start with AvairAI and put your team on the metrics that actually move revenue.