Why qualified leads die between marketing and sales
Most leads are not lost in the campaign. They are lost in the handoff, where nobody agreed what qualified means, nobody owns the first response, and the CRM record is too thin to act on.
Contents
Leads die in the handoff, not in the campaign. Three failures cause most of the loss: marketing and sales never agreed in writing what qualified means, nobody owns the first response, and the CRM record is too thin to act on. Fix the definition first, then the response clock, then the data.
Key takeaways
- A 2011 Harvard Business Review audit of 2,241 US companies found 23% never responded to a web lead, and that the average first reply took 42 hours.
- The same study measured contact within an hour as nearly seven times as likely to qualify a lead as waiting slightly longer, and more than 60 times as likely as waiting a day.
- Forrester rebuilt its demand model around buying groups in May 2021, on the finding that 95% of purchases involve three or more people across two or more departments.
- Gartner estimated in 2020 that poor data quality costs organizations at least $12.9 million USD a year on average.
Marketing and sales are not arguing about effort
They're arguing about a threshold nobody wrote down. Marketing counts a form submission. Sales counts a conversation with someone who can sign. Both measure honestly, and each measures a different object.
The repair is a definition made of fields, not adjectives. Decide which values in which CRM fields make a record qualified: role, budget signal, timeline, territory, product fit. A criterion that can't be stored in a field can't be audited, so it gets disputed later.
Forrester's model moves the unit of measurement from the individual name to the buying group, which is closer to how a purchase is actually decided.
Forrester replaced its individual-lead demand model in May 2021, citing research that 95% of purchases involve three or more people across two or more departments. Source: Forrester press release, 4 May 2021.
What the first hour is actually worth
The most-quoted number here deserves reading at its source: The Short Life of Online Sales Leads, by James Oldroyd, Kristina McElheran and David Elkington, in the March 2011 Harvard Business Review. The authors audited 2,241 US companies with test leads sent through their own websites.
Their results: 37% responded within an hour, 16% between one and 24 hours, 24% took longer than a day, and 23% never responded. Among firms that replied within 30 days, the average first response took 42 hours.
Two caveats. The percentages belong to 2011 and to a US sample, so they describe that population, not yours. What survives is the shape of the curve: the penalty for delay is front-loaded, so an hourly rota beats a daily review.
Firms that contacted a web lead within one hour were nearly seven times as likely to qualify it as those that waited slightly longer, and more than 60 times as likely as those that waited a day. Source: Harvard Business Review, March 2011, a study of 2,241 US companies.
Routing rules that survive a busy week
A rule counts only if it works on the worst day of the month. Assign ownership to a named person at creation, not to a queue. Set an explicit response window and an escalation that fires when it passes.
Then write down the boring cases, because those leak. Who takes a lead arriving at 19:00. What happens when the owner is on a flight between Tegucigalpa and Madrid. What a rejection looks like.
Teams that skip the boring cases rediscover them at the quarterly review, with a hole in the pipeline nobody can attribute. It's the same operations problem underneath most transformation programs: the process exists, but not for the exception.
CRM hygiene is a commercial problem
Duplicate records, empty source fields and free-text stage names are not administrative debt. They are why a routing rule fires at the wrong person and why a report can't be defended.
Salesforce's sixth State of Sales report, published on 25 July 2024 from a survey of 5,500 sales professionals in 27 countries, found reps spend 70% of their time on work that isn't selling, and that only 35% completely trust their organization's data. Gartner's 2020 estimate puts the average annual cost of poor data quality at $12.9 million USD per organization.
Automation sharpens this rather than softening it. Before adding assistants or scoring models, check what AI in the CRM does reliably and what is still demo-ware. A model trained on a dirty pipeline reproduces the dirt faster.
What to measure
Four numbers, reported weekly, settle most of the argument.
- Median time to first human contact, by source.
- Share of leads contacted inside the agreed window.
- Share of leads accepted by sales, with rejection reasons.
- Conversion to closed business by source, using an attribution model you can defend.
Cross them with a content view. If one channel produces leads sales accepts and another produces volume that gets rejected, that is a fact about the content that compounds rather than the campaign that spikes.
FAQ
What counts as a qualified lead?
Whatever both teams have agreed in writing and stored in CRM fields. A workable minimum is role, a budget or authority signal, a timeline and a territory. A criterion that can't be recorded as a field value can't be audited, so it gets argued about later.
How fast does the first response need to be?
The 2011 Harvard Business Review study found the sharpest advantage inside the first hour. Treat an hour as the design target during business hours, publish it as a commitment, and escalate automatically when it passes. A slower window that is always met beats an aspirational one.
Who owns a lead that sales rejects?
Marketing, with a reason code attached. A rejection without a recorded reason breaks the feedback loop that improves targeting and makes the next campaign no better than the last. The returned lead re-enters nurture with its history intact, rather than being deleted or quietly reassigned.
Will a new CRM fix this?
No. A CRM records decisions; it doesn't make them. If the definition of qualified, the response window and the routing rules aren't agreed first, a migration moves the same disagreement into a more expensive system and delays the repair by a quarter.
Where to start
Do it in order: definition, response window, routing rules, then data. Reversing that order produces a tidy database nobody uses. Six months of weekly measurement answers what you can't answer today: whether the loss sits in the definition or in the clock.
