# The marketing to sales handoff

> A handoff that does not leak: qualification thresholds, routing rules, response time SLAs, rejection reason codes and the recycle loop that returns leads to nurture.

Source: https://saas-marketing.net/playbooks/marketing-to-sales-handoff/
Topic: SaaS Sales
Type: playbook
Published: 2026-09-11
Last updated: 2026-09-11
Publisher: SaaS Marketing (saas-marketing.net)
License: CC BY 4.0. Quote or republish with attribution and a link to https://saas-marketing.net/playbooks/marketing-to-sales-handoff/

## Short answer

A working marketing to sales handoff has four parts: a qualification threshold combining fit score and an intent signal, routing logic that assigns an owner within seconds, a response time SLA with escalation, and a rejection path with fixed reason codes that recycles leads back to nurture. Sales acceptance rates of 20 to 35 percent are normal. Most handoff failures are latency and routing failures, not lead quality failures.

## Key takeaways

- Instrument timestamps at form submit, assignment and first touch before you redesign lead scoring. Latency is usually the real problem.
- Responding inside five minutes rather than an hour changes qualification odds by roughly an order of magnitude in the classic Lead Response Management study.
- Sales acceptance rate between 20 and 35 percent is normal. Above 80 percent usually means the threshold is set too high.
- Rejection without a fixed reason code is a lost lead and a lost diagnostic. Use six to eight codes, no free text.
- Recycled leads convert at a lower but non-trivial rate, so the recycle path is worth building before the scoring model is perfect.
- Sample 20 rejected leads a month with both teams in the room. The argument is the point of the meeting.

---

Every quarter, somewhere, a VP of Sales says the leads are bad and a VP of Marketing says the leads are not being worked. Both are usually wrong about the cause. The handoff is a system with four moving parts, and when it leaks, the leak is nearly always in the plumbing rather than in the scoring model everyone wants to argue about.

Start with timestamps. If you cannot say what your median time from form submit to first human touch is, you are not ready to redesign anything.

## Instrument latency before you touch lead scoring

Three timestamps, stored on the lead record, no exceptions: form submit, owner assignment, first outbound attempt. Add a fourth for first connect if your dialer writes it back.

Pull the median and the 90th percentile for each gap over the last 90 days. In most SaaS teams I have seen this done, the assignment gap alone runs 10 to 40 minutes because of a batched workflow, an enrichment step that waits on a vendor callback, or a routing rule that fires on a scheduled job rather than on record creation.

The Lead Response Management study found that contacting a lead within five minutes rather than thirty raised the odds of qualifying that lead by roughly 21 times. That finding is old now and has been replicated in various forms since, and it still describes the largest single lever in most handoff systems.

If your 90th percentile assignment gap is over five minutes, fix that before you spend a quarter rebuilding scoring. It costs an engineer a day and it is the only handoff change that reliably moves conversion without anyone having to agree on definitions.

## The qualification threshold: fit score plus an intent signal

A single numeric score is the wrong shape. Fit and intent are different questions and they fail in different ways, so keep them as two axes and define the threshold as a region rather than a line.

Here is a worked model that is deliberately simple enough to maintain.

| Component | Signal | Points |
| --- | --- | --- |
| Fit | Employee count in ICP band | 25 |
| Fit | Industry in target vertical | 15 |
| Fit | Job title in buying committee | 20 |
| Fit | Uses a required integration, from enrichment | 10 |
| Intent | Demo or pricing form submitted | 40 |
| Intent | Three or more sessions in 14 days | 15 |
| Intent | Viewed pricing or a comparison page | 15 |
| Intent | Product trial started | 30 |
| Negative | Personal email domain | minus 20 |
| Negative | Student, job seeker or competitor title | minus 40 |

Threshold: fit at or above 45 and intent at or above 30 becomes an MQL routed to sales. High fit with low intent goes to nurture. High intent with low fit goes to self-serve or a low-touch queue rather than to a rep, because an enthusiastic prospect who cannot afford you wastes more rep time than a cold one.

The definitions matter more than the numbers. Write down what an [MQL](/glossary/marketing-qualified-lead/) is and what a [sales accepted lead](/glossary/sales-accepted-lead/) is, get both VPs to sign the page, and keep the signed version somewhere both teams can see. Half the friction in handoffs is two teams using one word for two things.

## Routing: territory, round robin and named account overrides

Routing runs in a fixed precedence order, and the order is where most implementations go wrong.

**Routing precedence**

That last step deserves emphasis. If a prospect can book a slot on the confirmation screen, your speed-to-lead problem for demo requests disappears entirely. It is the single highest-return routing change available and it takes a week.

## The response SLA and what escalation actually looks like

An SLA without escalation is a suggestion. Write the tiers, write the escalation, and put the breach report in a channel both leaders read.

Two things make SLAs hold. First, the breach alert goes to the rep's manager, not to marketing, because marketing complaining about sales response time is a political act and a manager seeing their own team's number is a management act. Second, reassignment after a hard breach. A lead that sits unworked for an hour after a demo request should move to whoever will call it.

The full agreement wants to be written down properly. Use the [sales and marketing SLA template](/templates/sales-marketing-sla-template/) for the commitment both ways, and the [demand generation SLA variant](/templates/sales-marketing-sla-template/) if your top of funnel volume commitments need their own section. The wider relationship is covered in [sales and marketing alignment for SaaS](/guides/sales-and-marketing-alignment-slas/).

## Rejection reason codes and the recycle loop

A rejection with free text is a rejection you cannot analyse. Use six to eight fixed codes, make the field required to reject, and never let a rep pick a generic one.

- Not a fit, company too small
- Not a fit, wrong industry or use case
- Not a decision maker and no path to one
- Timing, evaluating in more than two quarters
- Already a customer or in an open opportunity
- Competitor, student or job seeker
- Bad contact data, could not reach
- Duplicate of an existing record

Each code has a destination. Timing rejections go to a long nurture with a re-entry trigger set to fire when the account shows fresh intent. Wrong-size companies go to the self-serve track. Bad data goes back to marketing ops for a source-level audit, because bad data clusters by source and one broken form or one bad list buy generates most of it.

Do not add 'low quality'. It absorbs every other code within two months because it is the fastest option in the dropdown, and then your reason code data tells you nothing. If a rep genuinely cannot fit a lead into the eight codes, that is a conversation, not a ninth code.

Recycled leads convert at a lower rate than fresh MQLs, and that is fine. The comparison is not against fresh leads, it is against zero, which is what an unrecycled rejected lead is worth.

## What must travel with the record

A rep should understand the lead in one screen. Every field they have to go hunting for is a field that will not be used.

| Field group | Specifics | Why it changes the call |
| --- | --- | --- |
| Source | Channel, campaign, first and last touch URL | Tells the rep what the person was reading |
| Self-reported | The 'how did you hear about us' free text answer | Often the truest signal, and it names podcasts and communities no tracker sees |
| Behaviour | Last five pages, session count, pricing page visits | A pricing page visit changes the opening line |
| Product | Trial status, key activation events, seats invited | Lets the rep open with what the prospect already did |
| Firmographic | Employee count, industry, funding stage, tech stack | Sets the fit narrative and the likely budget |
| Scores | Fit score, intent score, and their components | Component visibility stops reps dismissing the score as a black box |

The self-reported field is the one most teams skip and the one reps quote back most often. It costs one form field.

## The monthly audit that keeps it honest

Once a month, 45 minutes, both teams, twenty randomly sampled rejected leads on screen. Read each one, argue about whether the rejection was right, and record the verdict.

You are looking for three patterns: rejections that were wrong, which means the threshold or the rep needs work; rejections that were right for a reason the scoring model should have caught, which is a scoring fix; and rejections clustered in one source, which is a channel fix or a channel to cut.

This meeting is uncomfortable for about two months and then becomes the most useful recurring meeting either team has. Run it with the [handoff audit checklist](/checklists/marketing-to-sales-handoff-audit/) so it does not drift into a general complaint session.

## Start here

Pull your three timestamps this week. If median time to first touch on demo requests is over five minutes, fix routing and add instant booking before you touch anything else. Then write the reason codes, make the field required, and give it 90 days before you draw conclusions from the distribution.

Scoring models are the last thing to rebuild, not the first. They are also the only part everyone enjoys arguing about, which is exactly why they absorb attention that belongs on latency. The wider commercial picture, including what happens after acceptance, sits in [SaaS sales strategies](/saas-sales/), and if your definitions need tightening, the [MQL metrics reference](/glossary/marketing-qualified-lead/) and the [marketing qualified lead definition for marketers](/glossary/marketing-qualified-lead/) are the two pages to align on first.

## Frequently asked questions

### What is a good sales acceptance rate for MQLs?

Twenty to thirty five percent is the normal band for B2B SaaS. Below 15 percent, your threshold is too loose and sales has already stopped trusting the queue. Above 80 percent, the threshold is so tight that marketing is holding back leads that would have converted, which shows up as pipeline shortfall rather than as a quality complaint.

### How fast should sales respond to an inbound lead?

Inside five minutes for demo requests and high-intent forms. The Lead Response Management study found contact odds fall sharply after the first five minutes and qualification odds drop by roughly 21 times between five minutes and thirty. Content downloads can tolerate a longer window, usually 24 hours, because intent is lower and an instant call feels intrusive.

### What fields should travel with a lead from marketing to sales?

Source and campaign, the self-reported attribution answer, the last five pages viewed, form fill content, any product usage if a trial exists, firmographics such as employee count and industry, and the fit and intent scores with their component values. A rep who has to open three systems to understand the lead will open none of them.

### What is lead recycling and why does it matter?

Recycling returns a rejected or unworked lead to a marketing nurture track rather than letting it sit dead in the CRM. It matters because the majority of leads rejected for timing reasons are genuinely good leads on a longer clock. A recycle path with a re-entry trigger typically recovers a meaningful share of them within two quarters.

### Should lead routing be round robin or territory based?

Territory based with named account overrides for anything above your enterprise threshold, and round robin only inside a territory to balance load. Pure round robin breaks account ownership and produces the situation where two reps call the same company in one week, which buyers notice and comment on.

### How do you fix a handoff that sales says is broken?

Measure before you argue. Pull median time from form submit to first touch, sales acceptance rate by source, and the reason code distribution for the last 90 days. In most cases the data shows a routing delay or a source-specific quality issue rather than a broad quality problem, and that turns a political argument into a fixable one.

### Should marketing be measured on marketing-sourced or marketing-influenced pipeline?

Report both and make sourced the accountable number. Sourced is narrow but hard to game, which is why finance trusts it. Influenced is broader and closer to the truth of how buying actually happens, but it inflates easily. Publishing both with the definitions written down prevents the quarterly argument about which one counts.