Top of Funnel Marketing: The Simple Guide to More Leads

You rewrote the landing page for the fourth time. Conversion rate crawled from 2.1% to 2.4%, cost per lead is still $180, and pipeline has been flat for three months.

Here is the uncomfortable version: you are optimizing the wrong people. Top of funnel marketing is not the brand exercise you get to later. It decides whether anything downstream can work at all.

A 50% lift in conversion rate on a small audience is a rounding error. Doubling the right reach at the same conversion rate is not. Systems beat hacks, and the top of the funnel is where the system starts.

What Top of Funnel Actually Is (And What It Isn't)

TOFU is the "See" and "Think" stages in the See-Think-Do-Care framework, which sorts audiences by intent instead of funnel position. See and Think is about people who have the problem but no open RFP, no budget line and no vendor shortlist.

The Ehrenberg-Bass Institute put a number on that. Their 95-5 rule says that at any given moment, about 95% of B2B category buyers are out of market.

Only roughly 5% are actively buying. In consumer categories it is closer to 96/4. The LinkedIn B2B Institute popularized the framing, and it is the most useful planning number in B2B.

That means the 5% you are bidding on right now is the smallest and most expensive slice of your market. Everyone else is bidding there too.

It gets sharper. Widely cited Gartner and CEB research, plus Forrester's later framing, shows buyers finish most of the decision before they ever talk to a rep.

Gartner's data puts supplier meetings at about 17% of total buying time, spread across a buying committee of 6 to 10 people. Your shortlist is largely set before your first discovery call.

So the honest test of TOFU is not reach. It is this: when the category need hits, can a buyer name your brand?

Awareness without association is just spend. That raises a question nobody asks early enough: can you even afford the 5% you are fighting over?

Step 1: Do the Arithmetic Before You Touch an Ad Account

One spreadsheet row. Work backwards from the target.

200 demo requests per quarter ÷ 3% demo-request rate ÷ 40% qualified share of traffic = 16,667 visitors needed.

Now push it up a stage. At a 0.5% top-of-funnel click-through rate, 16,667 visitors requires roughly 3.33 million impressions. Spending $18,000 on TOFU implies a $5.40 CPM (cost per thousand impressions).

That single number changes the plan. LinkedIn in North America commonly runs $30 to $50 CPMs for senior B2B targeting. You cannot buy 3.33 million LinkedIn impressions for $18,000.

The math just killed the LinkedIn-only strategy before anyone opened Ads Manager. That is the point of doing it first.

Then split the budget with the 60:40 rule from Binet and Field's IPA research: the strongest campaigns put roughly 60% into brand building and 40% into activation, shifting toward 40:60 only when short-term sales are the entire objective. Your top of funnel marketing budget is not a cost center. It is the compounding asset.

That is why last-click attribution hates it. TOFU is almost always the first touch, so it looks worthless in the report. Cut it, and pipeline sags two to four quarters later while everyone blames sales.

Step 2: Build TOFU Content Around Buying Triggers, Not Personas

Category Entry Points, another Ehrenberg-Bass idea, are the situations and cues that make someone think of your category at all. Build assets around those cues, not around personas.

Personas say "Marketing Manager, 35 to 44, interested in HR technology." That is useless for content. Cues look like this:

• "Our HRIS can't handle multi-country payroll"
• "We just raised a Series B and headcount doubled in a quarter"
• "We have a compliance audit in Q3"

Now the hard part. Ahrefs analyzed roughly 2 million pages and found about 90.6% got zero organic search traffic. Publishing more blog posts is not a strategy. One original data study ("we analyzed 400 mid-market payroll migrations, here is what broke") outperforms fifty listicles.

Informational content is also being absorbed by AI answers. Pew Research measured in July 2025 that users clicked a source link on about 8% of visits when an AI summary appeared, versus about 15% on traditional results pages. Write to be the cited source, not just a blue link.

Ungate two of your three flagship assets. Gating inflates lead counts and selects for people who want the PDF, not the product. And when a strong asset points at a mismatched offer, you get a message match problem instead of leads.

Step 3: Rent Reach. Match the Paid Channel to the Job

LinkedIn buys precision at a price. Expect roughly $8 to $15+ per click and $30 to $50 CPMs in North America for senior targeting. The strongest formats are Thought Leader Ads (budget behind a founder's personal post) and Document Ads (carousels of a gated PDF).

Meta buys volume cheaply. Roughly $0.50 to $2.00 per click at $10 to $25 CPMs. Google Ads Demand Gen spreads cheap reach across YouTube, Discover, Gmail and Shorts.

Search runs about $1 to $10+ per click by industry, and paid search costs have been climbing high single to low double digits year over year. Demand capture is inflating faster than demand creation, which is the whole argument for TOFU in one line.

Targeting is no longer the lever. Meta's Advantage+ suite and its Andromeda retrieval overhaul pushed performance toward creative diversity. Broad targeting plus creative volume now beats narrow ICP slicing, because signal quality has degraded anyway.

Practical number: ship 5 to 10 creative variants per month, per platform, and kill losers at a threshold you set in advance. If you scale production with AI, watch for AI ad variation sprawl before it drowns your data.

Realistic plan: pick two of the three (LinkedIn, Meta, YouTube plus Demand Gen), then add one newsletter sponsorship in a niche publication for attention you partly own.

Step 4: Measure It Honestly When Last-Click Lies to You

Last-click will make TOFU look like waste, so stop using it as the scoreboard.

First, add a required open-text "How did you hear about us?" field to your demo form and treat the answers as primary data. GA4's direct and unassigned bucket will never explain dark social, podcasts or word of mouth.

Second, run a geo holdout. Pick two matched metros, run TOFU in only one for eight weeks, then compare branded search volume and demo requests against the control. This is the practical replacement for broken attribution, especially once consent mode errors start clipping your data.

Third, stop reporting email opens. Apple's Mail Privacy Protection has inflated "opens" for a large share of Apple Mail users since 2021. Track clicks, replies and downstream pipeline instead.

Review at 30, 60 and 90 days on five things: branded search volume, direct plus dark traffic, AI citation presence, self-reported attribution mix, and assisted pipeline. Tools like Semrush's AI toolkit, Ahrefs Brand Radar, Profound and Otterly track whether assistants name you in category answers. That is the new leading indicator of brand salience.

One reality check while you are here: organic reach for brand pages on social sits around 1% to 6% of followers. Treat it as a rounding error, not a channel.

Step 5: Where This Goes Wrong, And Whether to Build It Yourself

Four mistakes show up again and again.

1. Sending cold TOFU traffic to a "Book a demo" page. The offer must match the awareness stage. This is the cheapest fix and the most common mistake, and it is usually a demo no-show problem in disguise.
2. Optimizing to MQLs. MQL models reward harvesting existing demand and produce leads sales quietly rejects.
3. Over-targeting. Narrow ICP filters shrink audiences, raise CPMs, and are less accurate than they were.
4. "We'll do brand later." Brand effects take 6 to 24 months to reach pipeline. Teams that wait never build the asset.

DIY is viable, but the currency is attention, not skill: 5 to 10 creative variants per platform per month, UTM discipline, a holdout design, and a weekly kill-or-scale call. Budget one to two days a week of someone's job, indefinitely, plus the media spend.

Hand it to a top of funnel marketing agency or contractor when the budget exists but nobody owns it. Keep three decisions in-house regardless: your ICP, your Category Entry Points, and your offer.

Those are the business. The rest is execution.

Where to Go Next

• Fill in the Step 1 spreadsheet row with your own numbers today. If the implied CPM is under $10, you know you need a broader channel mix.
• List three buying triggers you hear on sales calls, then brief one flagship asset per trigger.
• Add the open-text attribution field before you release the 60% brand budget. Measurement wired in after launch is measurement you never get back.

A lot of teams can run this themselves with the steps above, and they should. If you would rather have the reach, tracking and follow-up handled as one system instead of five disconnected projects, our managed growth and scale retainers start at $2,500 a month, and we will tell you straight if you are too early for one.

Cover photo by Javier Miranda on Unsplash.