The Marketing Operator Podcast with Fexingo · 2026-06-30 · 12 min
Key moments - from our scoring
Substance score
80 / 100
Five dimensions, 20 points each
B2B companies are systematically leaving revenue on the table by ignoring the behavioral signals of anonymous website visitors. The episode walks through how form-first scoring models - the default in Marketo, HubSpot, Pardot, and Eloqua - create a fundamental blind spot: a decision-maker who visits your pricing page four times and spends 12 minutes on your product tour receives a score of zero if they never fill out a form. Lucas breaks down a real case study from ManuSoft, a mid-market manufacturing software company using Marketo, that recovered 15% of previously invisible traffic (1,100 leads) by building a parallel intent model. This model assigned points to behavioral signals: pricing page visits (15 points), time on product tour (10 points), multi-page sessions (5 points), and return visits within 7 days (20 points). At a 40-point threshold, the system automatically created lead records. The payoff: 18% of closed-won revenue in the following two quarters came from intent-scored leads, and these leads actually converted at 6.2% versus 4.8% for form-scored leads. The technical lift is minimal - just configuration work - but requires cross-functional alignment between marketing ops, sales, and analytics on which signals matter and what threshold triggers a lead.
Only about 12% of visitors receive a lead score in most marketing automation platforms because the default model is form-first; the other 88% remain completely unscored until they submit a form.
Build a parallel intent-based scoring model that assigns points to trackable behaviors like time on pricing page (15 points), time on product tour (10 points), returning within 7 days (20 points), and multi-page sessions (5 points), then automatically create a lead record when the score crosses a threshold (typically 40 points).
Yes - in the ManuSoft case study, intent-scored leads converted at 6.2% versus 4.8% for form-scored leads, likely because anonymous high-intent visitors are further along in their buying journey.
Most modern marketing automation platforms (Marketo, HubSpot, Pardot, Eloqua) have built-in web tracking via JavaScript snippets that capture page views, time on site, scroll depth, and referrer source; the key is ensuring tracking cookies persist across sessions for B2B work computers.
Implement a decay factor that cuts the score in half after 30 days of inactivity, add IP exclusion lists for known competitors, and set up nurture campaigns that require intent-scored leads to take action (like clicking an email CTA) before sales engagement.
Our reviewer’s read on each dimension, with quotes from the episode.
The episode is packed with specific, actionable insights that challenge conventional lead scoring wisdom. The core claim - that 88% of B2B site traffic goes unscored because platforms default to form-first models - is substantive and immediately useful. The ManuSoft case study provides concrete mechanisms (intent model thresholds, decay factors, signal weighting) and quantified outcomes (18% of closed-won revenue, 6.2% vs 4.8% conversion lift) that a marketing ops practitioner could operationalize. There is minimal filler; nearly every exchange advances the technical or strategic argument.
most B2B marketing automation platforms only assign a lead score to about 12 percent of the people who visit your website
18 percent of their total closed-won revenue in the next two quarters came from leads that originated from that intent model
The inversion of form-based vs. behavior-based scoring is contrarian within the typical lead-gen conversation; most platforms and practitioners still treat form completion as the primary signal, making this a genuine counterintuitive reframe. The insight that pricing-page visits may outweigh demo requests as intent signals, and that form-first models filter out self-paced buyers, challenges standard playbooks. However, anonymous behavior tracking and intent-based models are not entirely novel in martech; platforms like Marketo and HubSpot have supported this for years. The originality lies in the explicit critique and the rebalancing logic, not in the mechanics themselves.
Form-first scoring prioritizes people who are willing to give up their contact info early. But signal-based scoring captures people who are gathering information, building a case internally, and then buying when they're ready.
someone requesting a demo could just be kicking tires, while someone repeatedly checking pricing is probably comparing vendors
Lucas presents as a practitioner with hands-on martech operations experience, evidenced by the detailed ManuSoft case study and granular knowledge of platform capabilities (Marketo, HubSpot, Pardot, Eloqua). He demonstrates operational credibility through discussing decay factors, CRM field mappings, and cross-functional implementation challenges. However, his identity and seniority are ambiguous - he is not introduced with a title, company, or track record of scale. Luna appears to be a co-host rather than a distinct guest. The conversation is sophisticated but lacks the gravitas of a VP of Marketing or founder who has shipped this at enterprise scale.
I was talking to the VP of marketing at a manufacturing software company - let's call them ManuSoft - mid-market, about 200 employees, using Marketo
ManuSoft created a custom field called 'First Intent Signal' with a timestamp and the signal type
The episode excels in specificity, grounding every major claim in named metrics and concrete examples. The ManuSoft case study includes precise numbers: 8,000 unique visitors/month, 800 form-based leads, 7,200 unscored, 40% from relevant IPs, 1,100 intent-model leads (15%), 18% revenue contribution, 6.2% vs 4.8% conversion rates, 3% false positive rate by Q2, 23% nurture conversion to known lead. Scoring thresholds (40 points), signal point values (15 for pricing page, 10 for product tour, 20 for 7-day return), and platform specifics (JavaScript snippet, cookie persistence, decay logic) are all named. The host requests a specific action item (filter leads with zero score but 3+ page views in 30 days), demonstrating rigor.
They set a threshold of 40 points. Once an anonymous visitor hit 40, the system would automatically create a lead record in their CRM with a source of 'anonymous intent - high.'
18 percent of their total closed-won revenue in the next two quarters came from leads that originated from that intent model
Luna asks clarifying and probing questions that drive the conversation forward ('So what's the fix?', 'Did they see any downside?', 'What do you put in the CRM fields?'), demonstrating genuine curiosity rather than softball setup. She pushes back gently with counterintuitive observations ('I'd think someone who fills out a form is more qualified'), which Lucas uses to deepen the insight. However, the conversation lacks adversarial edge or deep follow-ups on limitations. Neither host challenges the methodology (e.g., survivorship bias in the ManuSoft example, whether 18% generalizes, or the assumption that high-intent anonymous visitors are actually prospects vs. price-shopping competitors). The aside about ad-free podcasting disrupts momentum and feels like a tangent rather than a hard question. Overall, competent interviewing but not masterful.
Good question. They did have some false positives. About 8 percent of the intent-scored leads turned out to be students, competitors, or people who just clicked around a lot.
That's counterintuitive. You'd think someone who fills out a form is more qualified. Luna: You'd think. But the data suggests...
Computed from the transcript - who did the talking, and the words that came up most.
Episode 82 of The Marketing Operator Podcast tackles a hidden drain on B2B revenue: the leads that never get a score at all. Lucas and Luna break down why most marketing automation platforms only score leads that fill out a form, ignoring the 60-plus behavioral signals that happen before a prospect ever touches your database. They walk through a real example from a manufacturing software company that recovered 18 percent of its pipeline simply by scoring anonymous website engagement - time on page, scroll depth, repeat visits - and letting those signals trigger a lead record. The episode contrasts 'form-first' scoring with a signal-based model that captures intent earlier, and shows how one mid-market SaaS brand turned 43 percent of its previously unscored traffic into qualified leads within two quarters. If your CRM shows most site visitors as 'unknown,' this episode will change how you think about marketing automation.
Transcribed and scored by The B2B Podcast Index.
Lucas: So here's a number that stopped me this week: most B2B marketing automation platforms only assign a lead score to about 12 percent of the people who visit your website. Luna: Twelve percent? That means 88 percent of site traffic is just... invisible to the scoring engine?
Lucas: Exactly. And it's not because the tech can't handle it. It's because the default scoring model is form-first - you only get a score once you fill out a gated piece of content. Everything before that is a ghost.
Luna: Right, so if a decision-maker visits your pricing page four times, reads three case studies, spends 12 minutes on your product tour - but never submits a form - they get a score of zero. Lucas: Zero. And that's a leak. Because those are your highest-intent visitors.
They're doing research, comparing options. And your system treats them the same as someone who bounced after two seconds. Luna: So what's the fix? Do you have to scrap the whole scoring model and rebuild from scratch?
Lucas: Not necessarily. But you do need to add a layer of signal-based scoring that runs on top of your form-based score. Think of it as two tracks. Track one is the traditional form-submission scoring - someone downloads a whitepaper, gets 20 points.
Luna: And track two? Lucas: Track two scores anonymous behavior. Time on page. Scroll depth.
Number of visits in a rolling 30-day window. Pages per session. And the kicker - you can assign points to those signals and then, at the point where the score crosses a threshold, you create a known contact record. Luna: So you're essentially minting a lead out of thin air based on intent data alone.
Lucas: Thin air with a paper trail. Let me give you a real example. I was talking to the VP of marketing at a manufacturing software company - let's call them ManuSoft - mid-market, about 200 employees, using Marketo. Luna: ManuSoft, got it.
Lucas: They had a standard scoring model. Whitepaper download: 25 points. Demo request: 50 points. Webinar attendance: 30 points.
About 800 leads a month coming in through forms. But their site was getting 8,000 unique visitors a month. So 7,200 were completely unscored. Luna: That's a lot of gray area.
Lucas: And here's the thing - we know from their analytics that about 40 percent of those anonymous visitors were from IPs associated with manufacturing companies. So they were relevant. They just hadn't filled out a form. Luna: So what did they do?
Lucas: They built a parallel scoring model - we called it the intent model - that assigned points for: visiting the pricing page, 15 points. Spending more than 3 minutes on the product tour page, 10 points. Visiting more than 3 pages in a session, 5 points. Returning within 7 days, 20 points.
Luna: And when did they create the contact record? Lucas: They set a threshold of 40 points. Once an anonymous visitor hit 40, the system would automatically create a lead record in their CRM with a source of 'anonymous intent - high.' No form required.
Then the lead was routed to SDRs with a note saying what signals triggered it. Luna: How many of those 7,200 crossed the threshold? Lucas: In the first quarter, about 1,100. That's 15 percent of their previously invisible traffic, now sitting in their CRM as scored leads.
And here's the real number - 18 percent of their total closed-won revenue in the next two quarters came from leads that originated from that intent model. Luna: Eighteen percent. That's almost a fifth of their pipeline they were just leaving on the table. Lucas: And that's what I mean by a leak.
It's not a data-quality issue. It's a model-design issue. Most marketing automation platforms come with out-of-the-box scoring that's entirely form-dependent. You have to deliberately turn on anonymous behavior tracking and build the logic.
Luna: And that's the part that a lot of teams skip because it takes time to set up, and it's not as straightforward as 'download this gated asset equals a lead.' Lucas: Exactly. It requires a different mental model. You're not waiting for the prospect to raise their hand.
You're reading their body language. And that takes a bit more sophistication in your automation setup. Luna: You know, speaking of sophistication, this is exactly the kind of conversation that makes me glad we don't run ads on this show. Because if we did, we'd have to squeeze a commercial break right in the middle of this scoring model walkthrough.
Lucas: Yeah, we've always kept The Marketing Operator ad-free. It's a choice. If these conversations have sparked something you've actually used, and you want to support that choice, the link is buy me a coffee dot com slash fexingo. Luna: No pressure - just if it's valuable to you.
Anyway, back to the scoring model. You said they hit 18 percent of closed-won from intent leads. That's impressive, but I'm wondering - did they see any downside? False positives?
Lucas: Good question. They did have some false positives. About 8 percent of the intent-scored leads turned out to be students, competitors, or people who just clicked around a lot. But they built in a decay factor - if a lead didn't engage for 30 days, the score dropped by half.
And they added a 'competitor IP' exclusion list. Luna: So the model got better over time. Lucas: Much better. By the end of the second quarter, the false positive rate was down to 3 percent.
And the conversion rate for intent-scored leads was actually higher than form-scored leads - 6.2 percent versus 4.8 percent. Luna: That's counterintuitive.
You'd think someone who fills out a form is more qualified. Lucas: You'd think. But the data suggests that anonymous high-intent visitors are often further along in their buying journey. They've done their research.
They're just not ready to talk to sales yet. And by the time they do reach out - or by the time they cross your intent threshold - they're more likely to buy. Luna: So the form is actually a friction point that filters out people who want to move at their own pace. Lucas: Exactly.
And that's the core insight. Form-first scoring prioritizes people who are willing to give up their contact info early. But signal-based scoring captures people who are gathering information, building a case internally, and then buying when they're ready. Luna: So what about the technical implementation?
What signals are actually trackable without a form? Lucas: Most modern marketing automation platforms - Marketo, HubSpot, Pardot, Eloqua - they all have web tracking capabilities. You place a JavaScript snippet on your site, and it captures page views, time on site, pages per session, referrer source, even scroll depth if you configure it. The key is to map those to point values in your scoring model.
Luna: And you need to make sure the tracking cookie persists across sessions. Lucas: Right. That's where it gets tricky. If a visitor clears cookies or uses a different device, you lose the thread.
But for B2B, where most research happens on a work computer during business hours, the cross-device issue is less severe than in B2C. Luna: So the low-hanging fruit is really just turning on anonymous behavior scoring and setting a reasonable threshold. Lucas: That's step one. But I'd argue step one point five is to review your existing form-based scoring and see if you're over-indexing on gated content.
A lot of teams give 50 points for a demo request and 10 for a pricing page visit. But the pricing page visit might actually be a stronger signal of purchase intent. Luna: Yeah, because someone requesting a demo could just be kicking tires, while someone repeatedly checking pricing is probably comparing vendors. Lucas: Exactly.
So you might want to rebalance your points. Give more weight to deep-content consumption and return visits. And less weight to one-off form fills. Luna: Let's talk about the data structure.
When you create a lead from anonymous behavior, what do you put in the CRM fields? Lucas: That's a great operational question. You have to be intentional. ManuSoft created a custom field called 'First Intent Signal' with a timestamp and the signal type.
They also populated the 'Lead Source' as 'Web - Anonymous Intent.' And they left the name field blank - because you don't have a name yet. That freaks out some sales teams. Luna: Right, because sales wants to call someone by name.
Lucas: Yes. So you need to set expectations. These are not hand-raisers. They are warm leads you can nurture with targeted content until they identify themselves.
ManuSoft set up a drip campaign specifically for intent-scored leads - no phone call, just educational emails with a 'talk to an expert' CTA. Once they clicked that, they got a name and the SDR stepped in. Luna: So the nurture was essentially the bridge from anonymous to known. Lucas: Exactly.
And that nurture campaign had a 23 percent conversion rate to known lead within 90 days. Which is pretty solid. Luna: I want to zoom out a bit. Across the B2B landscape, what percentage of companies actually do any form of anonymous behavior scoring?
Lucas: Based on what I've seen from industry surveys and platform usage data, probably around 30 to 35 percent. Most are still purely form-based. And of those that do anonymous scoring, many only track page views without assigning points. So they have the data but they're not operationalizing it.
Luna: That's a huge opportunity gap. Especially because the cost to implement is basically just configuration time. Lucas: It's not even a dollar cost. It's a thinking cost.
You have to decide what signals matter and what threshold means 'this person is worth following up with.' And that requires a cross-functional conversation between marketing ops, sales, and analytics. Luna: Which is exactly the kind of conversation that marketing ops should be leading. Lucas: Absolutely.
And once you have that model in place, you can keep iterating. ManuSoft eventually added firmographic scoring - matching IPs to company names and assigning bonus points if the company was in their target industry. That pushed their pipeline recovery even higher. Luna: So the playbook is: audit your current scoring model, identify which anonymous signals you're already tracking but not scoring, set up a parallel intent model with a threshold, and create a nurture path for those unknown leads.
Lucas: That's it. And the first step is honestly just looking at your analytics and asking: how many of my site visitors are scoring zero? If it's more than 70 percent, you have a leak. And you can plug it with a few weeks of configuration work.
Luna: Love it. Alright, let's leave listeners with one action item. Lucas: Go into your marketing automation platform this week and look at the scoring logs. Filter for leads with a score of zero but more than three page views in the last 30 days.
Export that list. That is your hidden pipeline. Then start building the signals. Luna: Consider it homework.
Lucas: Good homework. That's it for this episode of The Marketing Operator. We'll be back next week with another angle on the marketing tech stack.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.
Why Your Marketing Automation Ignores SMS85 / 100How B2B Brands Wreck Pipeline with Unsyncroned CRM Data92 / 100How B2B Brands Leak Revenue via Unchecked Data Decay76 / 100Why B2B Brands Fail at Account Based Marketing Attribution88 / 100How B2B Brands Leak Revenue Through Incomplete Lead Data86 / 100
