TL;DR: The MQL fails because it scores activity (opens, downloads, webinar signups) as a proxy for intent and the two are only weakly correlated. In an analysis of 300 consecutive MQLs, only about 32% were genuinely worth a rep's time; the rest were in-category-but-not-ready, out-of-category, or junk form fills. Track these instead: MQL-to-opportunity conversion rate, MQL-to-revenue by source, and time from MQL to opportunity. Then shift from activity-based scoring to intent-based routing. Here's the full case.

‍

Why are MQLs unreliable?

MQLs are unreliable because most lead-scoring models measure marketing engagement email opens, content downloads, webinar attendance and treat it as a signal of buying intent, when in reality activity and intent are only weakly correlated. Someone can download an ebook for research, attend a free webinar out of curiosity, or visit a site because a colleague mentioned it, all while having zero purchase intent and still cross the MQL threshold. Meanwhile, a genuine buyer who visits the pricing page three times but never fills out a form may score nothing. The detailed breakdown and the metrics that actually predict revenue follow below.

‍

The MQL Was Supposed to Fix Everything

‍

When the Marketing Qualified Lead became the standard handoff metric between marketing and sales in B2B SaaS, it felt like a solution to a real and longstanding problem. Sales teams were frustrated with unqualified leads. Marketing teams were frustrated with having no visibility into what happened after leads were generated. The MQL was the compromise: marketing would score contacts based on behavior, hand off the ones above a threshold, and both teams would have a shared definition of what constituted a lead worth pursuing.

‍

In practice, the MQL has become one of the most reliable sources of misalignment in B2B go-to-market. Marketing optimizes for generating more MQLs, frequently at the expense of MQL quality. Sales teams develop skepticism about MQL quality to the point where many teams quietly deprioritize them in favor of their own outbound pipeline. And somewhere between the marketing dashboard and the sales CRM, real buyers who are genuinely ready to purchase fall through the cracks while both teams argue about whose fault it is.

‍

This post is about why the MQL model fails so consistently and what to do about it.

‍

The Fundamental Flaw: Activity Is Not Intent

‍

The root problem with almost every MQL model in production today is that they score activity as a proxy for intent. A contact opens an email: plus five points. Visits the website twice: plus ten points. Downloads an ebook: plus fifteen points. Attends a webinar: plus twenty points. Cross a threshold and you're an MQL, regardless of whether any of that activity reflects genuine purchase consideration.

‍

The problem is that activity and intent are only weakly correlated. A contact downloads your ebook because they're doing research for a blog post they're writing about your category. They attend your webinar because the topic is interesting and it's free. They visit your website because a colleague mentioned your company in a meeting and they were curious. None of these activities represent purchase intent. All of them score points in the typical MQL model.

‍

Meanwhile, a genuine buyer who visits your pricing page three times in a week, downloads your customer case studies, and then calls your sales line directly might score zero points in your MQL model if they didn't open any emails, register for any webinars, or download any gated assets. They expressed unambiguous purchase intent. Your model didn't catch it because they didn't take the specific paths the model was designed to detect.

‍

Activity-based lead scoring rewards the paths you designed, not the intent buyers actually express. That's a fundamental mismatch that no amount of scoring optimization can fully resolve.

‍

The reason this flaw is so durable is that activity is easy to measure and intent is hard to measure. Email opens, page views, and downloads are captured automatically by every marketing automation platform. Genuine buying intent is messier, more contextual, and harder to reduce to a point value. So teams measure what's easy rather than what matters, and then build their entire qualification model on the convenient proxy instead of the inconvenient truth. The model isn't wrong because anyone designed it badly; it's wrong because it optimized for measurability over meaning.

‍

Who's Actually in Your MQL Queue

‍

We ran an analysis of 300 consecutive MQLs from a typical B2B SaaS demand gen program to understand who was actually in the queue. The breakdown was illuminating and sobering.

‍

Roughly 32% were genuinely good fits with demonstrated interest people who were actually in some phase of research or evaluation in our category. About 23% were in-category contacts with no current buying urgency practitioners who consumed content about the space and engaged with educational material but were not actively evaluating solutions. About 18% were out-of-category entirely: students, consultants who work in the space but don't buy solutions, job seekers, researchers, and competitors doing competitive intelligence. And approximately 27% were low-quality form fills bad email addresses, obviously incomplete names, contacts at companies that didn't match any reasonable definition of the ICP.

‍

That means roughly one in three MQLs in this program was worth a sales rep's time. The other two were either unqualified or not ready to buy and the program was treating all three equally, routing them to the same sequences, generating the same follow-up tasks, and demanding the same time from the sales team.

‍

This is the hidden cost of the activity-based model, and it compounds in a way that's easy to miss. Every junk or not-ready MQL that gets routed to sales as if it were a genuine opportunity costs a rep time and erodes their trust in the entire MQL system. After enough bad handoffs, reps stop working MQLs seriously at all which means that when a genuinely good MQL does come through, it gets the same skeptical, deprioritized treatment as the noise around it. The two-thirds that are noise don't just waste time; they poison the well for the one-third that's signal.

‍

The Metrics That Actually Predict Revenue

‍

If MQL volume is a misleading metric, what should replace it? Several metrics are more predictive of revenue and more useful for making decisions about demand gen investment.

‍

MQL to Opportunity Conversion Rate. Not how many MQLs you generated, but what percentage converted into genuine pipeline opportunities with identified next steps and qualified need. If fewer than 15% of your MQLs are becoming opportunities, your scoring model is producing more noise than signal. This metric also creates healthy accountability: when marketing is measured on the quality of MQLs rather than just the volume, the incentive to optimize for form fills over genuine intent disappears.

‍

MQL to Closed Revenue by Source. Breaking down conversion rates by the source that generated the MQL almost always reveals massive variation that aggregate metrics hide. A benchmark study, a targeted webinar, and a broad content download might all generate 100 MQLs each but the benchmark study's MQLs might convert to revenue at 4x the rate of the broad content download's MQLs. If you're not tracking at this level of granularity, you're making investment decisions based on volume data that hides the signal you actually need.

‍

Time from MQL to Opportunity. A lead that becomes an opportunity within 48 hours of MQL creation is expressing different intent than one that converts after three weeks of nurture. Fast conversion indicates high current intent. Slow conversion often means the prospect wasn't actually ready when the MQL was created and only became ready after time passed and more buying criteria were met. Tracking this metric reveals whether your scoring model is catching buyers at the right moment or generating work for the sales team weeks before those contacts are actually ready for a sales conversation.

‍

The common thread across all three metrics is that they measure outcomes rather than inputs. MQL volume is an input it tells you how much activity your marketing generated, which feels like progress but says nothing about whether that activity produced revenue. Conversion rate, revenue by source, and speed to opportunity all measure what actually happened downstream. The shift from input metrics to outcome metrics is the single most important change a demand gen team can make, because input metrics can always be gamed by generating more activity, while outcome metrics can only be improved by generating better-qualified buyers.

‍

What to Build Instead

‍

The most effective demand gen teams we've observed are moving away from activity-based lead scoring toward intent-based lead routing. The shift sounds subtle but requires rethinking the fundamentals of how you identify and act on buyer interest.

‍

Intent-based routing doesn't ask "how much has this contact engaged with our marketing content?" It asks "what is this contact doing that signals they might be in an active buying process?" The signals are different: repeated visits to the pricing page, consumption of competitive comparison content, direct inbound inquiry through the contact form, or use of high-specificity search terms that indicate active evaluation. These behaviors are harder to game, harder to generate through broad content programs, and much more strongly correlated with actual purchase intent.

‍

When you identify a contact generating these signals, the right response isn't to enroll them in a nurture sequence designed for the top of the funnel. It's to route them immediately to a sales rep for a high-priority, personalized follow-up. The window of peak buying intent is short. A contact who's actively evaluating you today may have made their decision in three days. A nurture sequence that drips emails over three weeks doesn't serve them.

‍

The distinction between the two models comes down to what kind of behavior each one rewards. Activity-based scoring rewards consumption the more of your content someone consumes, the higher they score, regardless of why they're consuming it. Intent-based routing rewards evaluation behavior the specific actions people take when they're genuinely comparing solutions and moving toward a purchase. Consumption is easy to generate and weakly predictive. Evaluation behavior is hard to fake and strongly predictive. Building your system around the second category, even though it produces fewer "leads," produces dramatically better pipeline because every signal it surfaces means something.

‍

This reorientation requires marketing and sales to agree on what high-intent behavior looks like, build the infrastructure to detect it in real time, and commit to speed-to-follow-up as a key operational metric. It's more demanding than generating form fills. It's also the approach that produces the kind of pipeline quality that makes sales teams trust marketing which, in the long run, is the alignment that the MQL was always trying to create.

‍

How to Make the Shift

‍

If you want to move from activity-based MQLs to intent-based routing, start without rebuilding everything at once. First, audit your own MQL queue the way described above: pull 200 to 300 recent MQLs and categorize each as genuinely qualified, in-category but not ready, out-of-category, or junk. The resulting percentages will tell you how much of your current model is noise and build the internal case for change. Second, define jointly between sales and marketing the three to five behaviors that genuinely indicate active evaluation for your specific buyer, things like repeated pricing-page visits or competitive-comparison content consumption. Third, build a fast lane: when one of those high-intent signals fires, route the contact to a rep immediately rather than into a nurture sequence, and measure how quickly the rep follows up. Keep your existing nurture for the genuinely early-stage contacts, but stop treating them as sales-ready. The goal isn't to generate fewer leads; it's to make sure the leads you call sales-ready actually are.

‍

Frequently Asked Questions

‍

What is the difference between activity-based and intent-based lead scoring?
Activity-based scoring assigns points for marketing engagement email opens, content downloads, webinar attendance and treats accumulated activity as a signal of buying readiness. Intent-based routing instead looks for behaviors that specifically indicate active evaluation, like repeated pricing-page visits, competitive-comparison content, or direct inbound inquiries. Intent signals are harder to game and far more strongly correlated with actual purchase intent.

‍

What percentage of MQLs are actually qualified?
It varies by program, but in an analysis of 300 consecutive MQLs from a typical B2B SaaS demand gen program, only about 32% were genuinely good-fit contacts with demonstrated interest. Roughly 23% were in-category but not actively buying, 18% were out-of-category entirely, and 27% were low-quality form fills meaning about two-thirds weren't worth a rep's time as sales-ready leads.

‍

What is a good MQL-to-opportunity conversion rate?
A useful benchmark is 15% or higher. If fewer than 15% of your MQLs convert into genuine pipeline opportunities with qualified need and identified next steps, your scoring model is likely producing more noise than signal. Tracking this rate rather than raw MQL volume also removes marketing's incentive to optimize for form fills over genuine intent.

‍

Why do sales teams ignore marketing leads?
Because too many MQLs are unqualified or not ready to buy. When most handed-off leads turn out to be junk, not-ready, or out-of-category, reps stop trusting the MQL system and deprioritize all of it including the genuinely good leads buried in the noise. The fix is qualifying on intent so that "sales-ready" actually means ready.

‍

What metrics should replace MQL volume?
Three outcome metrics are far more predictive of revenue: MQL-to-opportunity conversion rate (quality of leads, not quantity), MQL-to-revenue by source (which channels produce real pipeline, since conversion can vary 4x by source), and time from MQL to opportunity (fast conversion signals high current intent). All three measure outcomes rather than activity, which can be gamed.