Marketing in the Dark Funnel


Last week, that question surfaced repeatedly when I opened a Section-led event on AI and strategic marketing.

People asked how to tell whether their content is being used inside AI tools, how to measure brand presence when clicks disappear, whether SEO still matters, and how to explain marketing impact when attribution breaks.

This week’s issue answers those questions.

Let’s dive in.

Missed Last Week? Read This First

As a long time marketer, I learned to operate inside the traditional funnel.

You know the one: where we move buyers from awareness to consideration and down into purchase.

Unfortunately buyers are no longer playing that game.

Instead, they’re moving in unpredictable and non-liner ways, bouncing between search, social and increasingly AI tools. Our beloved funnel now looks more like spaghetti on a wall and managing it is just as messy.

Google’s AI Mode is decimating our visits, website traffic (when we get it) no longer drives pipeline reliably, and buyers are showing up to sales calls more informed and harder to please.

What are we to do?

If you’re trying to wrap your head around how buyer behavior is changing, last week’s issue is a good place to start.

You can read it here.

Feel like you’ve got that covered? Read on.

How to Operate in the Dark Funnel

Modern marketing teams are going through an operating shift.

Buyer decisions increasingly form inside the “dark funnel” — forums, private communities, and AI tools like ChatGPT and Google’s AI Mode. A

As a result, large parts of the journey are no longer observable and the old model of tracking every click no longer reflects influence.

From Click Paths to Visibility Proxies

As AI search engines and social platforms pass less referral data, measurement is moving away from traffic and toward presence.

Instead of relying on click-through rate, teams are tracking Share of Model, Generative Position and Query Coverage.

This measures how often a brand is cited when an AI system is asked a category-level question, such as which solution is best for a specific use case.

Search experts measure this by running synthetic queries across AI tools. They ask consistent category questions and track whether their brand appears in the top responses over time using tools like Scrunch and AirOps.

The team at retail platform Tapestry realized their products weren’t appearing in AI shopping results because they were optimizing for too many generic keywords like “handbag”.

They conducted an audit to find the questions shoppers were asking: things like “glamorous adjustable crossbody for luncheon”. They used a large language model to analyze their product catalogue and created detailed tags and descriptions that matched the most popular question categories — which resulted in double digital sales growth.

From Last-Touch Heroics to Decision Signals

Last-touch attribution is giving way to signals that indicate whether a buyer is genuinely progressing toward a decision.

nstead of treating the thank-you page as the primary success metric, teams are tracking consumption velocity. This reflects how quickly and deeply an account or lead engages with high-intent content, including pricing pages, comparison tables, and implementation guides.

Intent data platforms such as 6sense and Demandbase help surface these shifts by detecting spikes in research intensity across the open web, even when no ad click occurs.

Snowflake applies this approach by weighting deep content engagement far more heavily than surface interactions. Sales prioritization is driven by readiness signals rather than individual form fills, so an SDR reaches out when a buyer hits a learning moment and is ready to talk.

From Time to Lead to Time-to-Confidence

In AI-assisted journeys, you’re more likely to win the deal when you help buyers feel confident enough to decide sooner.

Time-to-confidence measures how many interactions it takes for a buyer to stop searching and start evaluating.

It helps to create a “buyer friction index”, which looks at how easily a buyer can access answers without unnecessary gates or delays. If you’re still putting forms in front of your best content or pricing page you’re just slowing them down.

But how do we measure without a form fill?

Windstar Cruises used Experian’s identity graph to link bookings with their digital ad campaigns. They discovered that buyers who engaged with high-confidence content converted at significantly higher rates than those who followed traditional click-based paths, resulting in $20 million of measurable revenue impact tied to exposure.

From ROAS to Sales Conversation Quality

Measurement is shifting from cost per lead to pipeline influence.

One key signal is the qualified conversation rate: the percentage of marketing-sourced leads that enter discovery already understanding the product’s value.

Conversation intelligence platforms such as Gong and Chorus help quantify this by tracking brand mention density and question maturity during calls.

When buyers arrive asking how a product compares to alternatives instead of asking what it does, marketing is doing its job.

Want a quick snapshot of what you should be tracking?

The table below summarizes how legacy metrics map to new buyer journey signals:

Next Week

Before I sign off, I want to say thank you to everyone who joined the Section event last week and popped into the chat to say hi and share some love.

If you need convincing to watch the session on demand Javier P. called it “a killer presentation”.

I also appreciated the shoutouts from Kay B., Colleen K., Christine E., and Monica V. — it was so lovely to see you there!

And to everyone who had questions, I’ll be answering those in next week’s edition, so stay tuned!

Want to Level Up Your AI Game?

If your team is ready for a hands-on AI strategy session, my custom-designed workshops are built to uncover the workflows that can save you hours every week.

Prefer to start small? My YouTube channel is packed with quick, practical “how-to” videos that show you exactly how I use AI tools for marketing, content, and automation.

Planning an event or conference? I deliver high-energy AI sessions that engage audiences and leave them with actionable strategies they’ll talk about long after the event. Book me for your event here.

Did some one forward you this email? You can subscribe here.

2120 Contra Costa Blvd #1059 , Pleasant Hill, CA 94523
Unsubscribe · Preferences

AI at Work

AI at Work is a weekly newsletter on how marketing teams redesign workflows, roles, and systems with AI. Real examples, practical frameworks, and repeatable processes operators can use immediately. Join thousands of successful marketing leaders by subscribing below!

Read more from AI at Work
Getting AI Approval

Hello Reader There’s a dynamic I hear from marketing leaders right now, and it usually sounds like this. “My CEO keeps pushing us to drive results with AI, but when I put together a proposal, with budget, headcount, and a timeline, I struggle to get it approved.” If that feels familiar, you’re not alone, and the problem isn’t your proposal. It’s that most AI programs get stuck in what I call pilot purgatory: lots of small experiments, zero coordinated investment, and a CEO who’s enthusiastic...

Demand Generation

Hello Reader When I was the director of marketing for an events management company, we ran demand programs the way you pack for a trip at 5am, throwing things in and hoping you didn't forget something important. Every Monday we'd look at the pipeline number, panic, and launch something. By Friday we'd already moved on to the next thing without checking if the last one worked. If we missed our number, budget was on the line (and so was my job). And that was years ago. Now the pressure is even...

For most of my career, the stack was something I learned to live with. Five-year CRM contracts, clunky workflows and bolt-on tools the IT team never wanted to approve. We worked around limitations and stitched together processes. Proving attribution was near impossible. Then AI arrived—promising to finally simplify our lives. But has it? You’ll have to read on to find out 👇 You Stack Has a Context Problem Your team is running campaigns across a dozen tools. Some overlap, some were added to...