Almost every marketing-to-sales handoff I have audited has the same three failures: leads sit in a queue too long, the rep cannot tell which marketing action triggered the assignment, and roughly 15% of records get duplicated because the handoff lives in a spreadsheet that nobody syncs back to the CRM.
These are not fancy problems. They have boring fixes. Here are five rules that get a handoff working inside a quarter.
1. Define MQL with a written, falsifiable test
If you cannot write a one-line query that returns "this person is an MQL," your definition is decorative.
A good MQL definition reads like:
Contact is in our ICP (employee count 10-500, industry in our priority list), has filled a form on a high-intent page (pricing, demo, contact), and has opened at least one nurturing email in the last 14 days.
Three concrete conditions. A query that returns a row count. If sales pushes back on lead quality, you can show them the query and decide together which condition is wrong, not argue about feelings.
If your CRM cannot return that query in under 30 seconds, you have a CRM problem, not a marketing problem.
2. Hand off within 15 minutes of qualification
The data on speed-to-lead is one of the few sales metrics that is not controversial. Conversion drops roughly 80% between minute 5 and hour 24 for inbound demo requests. For form-fill MQLs that are not immediate demo requests, the cliff is gentler but still real.
In practice this means routing happens in the CRM or marketing tool automatically. Not via a Slack message that someone has to read. Not via an email to a shared inbox. A round-robin or territory rule that fires within 15 minutes, with a Slack/email notification to the assigned rep as a secondary alert.
If you cannot automate the routing yet, do not let that be an excuse to do nothing. Assign a single human owner of the queue and a 15-minute SLA. Imperfect routing now beats perfect routing in Q3.
3. Pass context, not just the contact
A bare contact record with no context is worse than no lead at all because the rep wastes time figuring out why the contact exists.
The handoff must include: which form they filled (with the actual form name, not "web form"), which page they were on, the campaign source if known, and any field values they submitted. This goes into a "lead context" note on the contact record so the rep sees it the moment they open the deal.
A useful test: can a rep cold to the deal explain what triggered the handoff in 10 seconds of looking at the record? If not, the handoff is missing context.
4. Acknowledge inside 24 hours or auto-reroute
Reps drop balls. Not maliciously. They get busy, the lead sits, marketing assumes it was worked, and three weeks later the prospect surfaces on a competitor's product.
Build an acknowledgement: the rep marks the lead as "contacted" (with a real activity attached, not a checkbox) within 24 working hours. If no acknowledgement, the lead auto-reroutes to a second rep or to a sales manager for triage. The point is not to punish the first rep. It is to make sure the lead does not vanish.
Track the SLA breach rate weekly. It should be under 5%. If it is over 20%, your routing is sending leads to the wrong people, your reps are over-capacity, or your MQL definition is too broad.
5. Send rejected leads back, with a reason
When a rep disqualifies an MQL, the disqualification reason is a structured field, not a free-text note. Three to six options: not in ICP, wrong job title, no budget, no timeline, already a customer, bad data.
This is the only way marketing learns. Without it, you get the perpetual "lead quality is bad" complaint with no path to fixing it. With it, you can look at a quarterly report and see that 40% of rejections are "wrong job title" and adjust the qualifying form.
The rejected leads themselves should drop back into a nurture sequence, not the trash. About 8-15% of disqualified MQLs become qualified again within 6 months as their situation changes.
What to measure
One leading metric (time to first touch, target <2 hours median) and one lagging metric (MQL-to-SQL conversion rate, expect 25-40% depending on your funnel definition). If either trends the wrong way for two weeks, audit the routing rules and the rejection reasons before changing anything else.
Most handoff problems are operational, not strategic. Fix the plumbing first.