TL;DR: The Marketing Qualified Lead was built for a world where buyers filled out forms to access information. That world is gone. Modern B2B buyers complete 70% of their research before ever contacting a vendor, and the MQL — a proxy metric built on form fills and email clicks — is increasingly disconnected from actual buying intent. The replacement is signal-based demand generation: identifying and acting on behavioral signals that indicate real purchase readiness, often weeks before a traditional “lead” would have been created.
The MQL Was Always a Compromise
Let’s be honest about what the MQL actually was: a handshake agreement between marketing and sales that said “if someone does X, Y, and Z actions, they’re ‘qualified’ and sales should follow up.”
It was never about buying intent. It was about creating a shared language between two departments that historically didn’t trust each other. Marketing needed to prove they were generating pipeline. Sales needed a filter so they weren’t chasing every webinar attendee. The MQL was the compromise — and like most compromises, it optimized for agreement rather than accuracy.
of B2B buyer research happens before a vendor is ever contacted. Your MQL was built for the other 30%. Signal-based demand gen — tracking intent signals like product page visits, competitor comparisons, and community engagement — identifies buying intent weeks before a traditional lead would exist. The playbook has changed.
The problems with MQLs are well-documented but worth restating:
They’re gamed within months. The moment a scoring threshold becomes known, marketing optimizes to hit it. More content downloads. More webinar registrations. More “engagement” that looks like intent but is really just content consumption.
They create perverse incentives. When your bonus depends on MQL volume, you naturally build programs that generate MQLs — not programs that generate pipeline. These are not the same thing.
They lag reality by weeks. By the time a buyer fills out a demo request form, they’ve typically completed 70% of their evaluation. Your MQL process just discovered someone who was already in-market. The real question is: who’s entering the research phase right now, and how do you reach them?
They ignore the best signal. The most powerful buying signal in B2B isn’t a form fill. It’s not a content download. It’s not even a pricing page visit — though that’s getting warmer. The best signal is when multiple people from the same company start researching your category simultaneously from different angles. That pattern — the “surge” — is what enterprise sales teams have known about for years but marketing automation never captured.
What Signal-Based Demand Gen Actually Looks Like
Signal-based demand gen flips the model. Instead of waiting for a buyer to raise their hand (form fill → MQL → sales outreach), you’re monitoring for behavioral patterns that indicate an account is entering a buying cycle, then orchestrating a response.
The signals fall into three categories:
Tier 1: First-Party Behavioral Signals (Your Own Data)
These are the signals you already have but probably aren’t aggregating properly:
- Account-level research surges: 3+ unique visitors from the same company accessing product, pricing, or comparison pages within a 14-day window
- Content depth patterns: A single visitor moving from “what is” content → “how to” content → “vs competitor” content → pricing page. This is a research journey, not casual browsing.
- Return frequency acceleration: A visitor who previously came once a month now visiting twice a week. Something changed.
- Multi-channel engagement: Same contact engaging with your emails, visiting your site, AND following your LinkedIn page in the same week
The technology to capture these patterns exists today. Tools like 6sense, Demandbase, and Clearbit (now HubSpot) provide account identification and intent scoring. But most teams haven’t rebuilt their demand gen process around them — they’re using intent data as an overlay on the same MQL model.
Tier 2: Third-Party Intent Signals
These are the signals from outside your ecosystem:
- Review site activity: An account that suddenly starts reading G2/Capterra reviews in your category
- Job change signals: A champion from your existing customer moving to a new company that’s not yet a customer
- Funding events: A company that just raised money is more likely to invest in new tools; the timing matters
- Technology stack changes: A company that just adopted a complementary or competitive tool in your ecosystem
None of these signals are individually conclusive. But when 3–4 of them fire simultaneously for the same account, the pattern becomes statistically significant.
Tier 3: Intent-to-Act Signals
These are the closest thing to the old MQL — but they’re narrower and more meaningful:
- Direct pricing inquiries: Not “viewed pricing page” but specifically requested pricing information
- Competitor comparison downloads: Someone who downloaded your competitor comparison guide is in evaluation mode
- Demo requests from decision-makers: Title matters here — a VP of Sales requesting a demo means something different than an SDR researching tools
- RFI/RFP inbound: Self-explanatory, but worth flagging as a distinct category
How to Implement Signal-Based Demand Gen (Without Ripping Everything Out)
If you’re reading this thinking “great, another framework requiring a complete tech stack rebuild,” that’s not what I’m suggesting. You can layer signal-based demand gen onto your existing infrastructure incrementally.
Phase 1: Enable Account Identification (Month 1)
If you don’t know which companies are visiting your website, start here. Tools like Clearbit Reveal, 6sense, or RB2B can identify 20–40% of your anonymous traffic at the company level. This doesn’t require any CRM changes — just a script on your site.
Once you have company-level identification, set up a simple dashboard showing which accounts visited in the last 7 days, which pages they viewed, and whether they’re in your ICP. This alone will surface accounts you didn’t know were in-market.
Phase 2: Define Your Signal Model (Month 2)
Don’t build a complex scoring model. Start with three simple flags:
- Research Flag: Account viewed 3+ high-intent pages (pricing, product, comparison, case studies) in 14 days
- Surge Flag: 3+ unique visitors from same account in 14 days
- Intent Flag: Any direct pricing inquiry, competitor comparison download, or demo request
When an account triggers any 2 of these 3 flags, it gets routed to your demand gen team — not as an “MQL” but as a “signal-qualified account.” The response isn’t an automated email sequence; it’s a human decision about whether and how to engage.
Phase 3: Build the Orchestration Layer (Month 3)
This is where signal-based demand gen separates from traditional lead scoring. Instead of “MQL score reached → send to sales,” you’re building a playbook:
- Signal-qualified account → Review by demand gen team within 24 hours
- Decision: Direct outreach or nurture? If the account has existing contacts or relationships, direct SDR outreach is appropriate. If not, add to a high-priority nurture track with content and ads targeted to that specific account.
- Feed back into content strategy: The signals you’re detecting should inform what content you create. If you see accounts repeatedly researching a specific topic before engaging sales, that topic needs more and better content.
The Metrics That Actually Matter Now
If you’re not measuring MQLs, what do you measure? Here’s the replacement dashboard:
| Old Metric | New Metric | Why It’s Better |
|---|---|---|
| MQL volume | Signal-qualified accounts (SQAs) | Measures actual buying signals, not content consumption |
| MQL → SQL conversion | SQA → Opportunity conversion | Higher-fidelity pipeline metric; fewer false positives |
| Cost per MQL | Cost per influenced pipeline dollar | Directly ties marketing spend to revenue |
| Marketing-sourced pipeline | Marketing-influenced pipeline (all channels) | Credits marketing for multi-touch influence |
| Lead response time | SQA-to-engagement time | Measures speed of human response to buying signals |
Notice that every replacement metric ties closer to revenue. The MQL model let marketing claim credit for volume without accountability for outcomes. Signal-based demand gen ties marketing performance to what actually matters: pipeline and revenue.
The Organizational Shift Nobody Talks About
The hardest part of killing the MQL isn’t the technology or the metrics. It’s the organizational change.
When you remove MQLs as the handshake between marketing and sales, you need a new handshake. The signal-based model requires:
A shared view of accounts. Marketing and sales need to see the same account list with the same signal data. This means CRM hygiene matters more than ever — and it’s typically terrible.
New SLAs. Instead of “marketing delivers 500 MQLs per month and sales follows up within 24 hours,” the SLA becomes “marketing surfaces signal-qualified accounts within 24 hours of detection and sales determines engagement strategy within 48 hours.” This is a conversation, not a handoff.
Trust in the signals. The first quarter of signal-based demand gen is uncomfortable. You’ll have fewer “leads” in the traditional sense, and some signals won’t convert. The instinct will be to revert to MQLs because the volume number was comforting. Resist it. Signal-based models typically produce fewer but higher-quality opportunities — the pipeline number will look different, but the revenue number will look better.
The Companies Getting This Right
A few examples worth studying:
Gong built their entire demand gen motion around account-level signals years ago and publicly attributed significant pipeline efficiency gains to it. Their model: identify accounts showing research surges → deploy targeted content and ads → engage through SDRs only when signal density crosses a threshold.
6sense (somewhat meta) uses their own intent data to run their demand gen, proving the model works. Their average deal size increased measurably when they shifted from lead-based to account-signal-based targeting.
Atlassian operates a largely product-led and signal-based motion where traditional MQLs barely exist. Their demand gen team monitors product signup patterns, expansion signals, and account research behavior to identify enterprise opportunities — and they’ve scaled to billions in revenue without a traditional MQL model.
The common thread: none of these companies eliminated demand gen. They eliminated the proxy metric and replaced it with something that measures reality more directly.
One Honest Warning
Signal-based demand gen is not easier than MQL-based demand gen. It’s harder. It requires better data infrastructure, tighter sales alignment, and a willingness to make judgment calls instead of following a scoring algorithm.
If your organization isn’t ready for that level of complexity — if sales and marketing are still fighting over lead definitions — fix the fundamentals first. A bad MQL model that people actually follow is better than a sophisticated signal model that nobody trusts.
But if you’re ready to stop measuring proxy metrics and start measuring what actually predicts revenue, the model is here. The tools exist. The case studies are public. All that’s left is the decision to change.
Related: The Full-Funnel Demand Gen Framework for B2B Marketers | Content Mastery Is the Silent Engine of Stellar Lead Generation | Ingenious Tactics for Seamless Lead Generation