Demand-gen vs lead-gen: budgeting for tech advertisers

The demand-gen versus lead-gen debate has been running in B2B marketing circles for several years now, and it has produced a lot of LinkedIn opinions and very little useful budgeting guidance. The framing usually pits the two against each other: demand-gen evangelists arguing that lead-gen is dying, lead-gen pragmatists pointing out that the CFO still wants pipeline this quarter. Both are right and both are missing the point.

For B2B tech firms with a paid budget to allocate, the question is not “demand-gen or lead-gen?” but “what proportion of each, and how do we measure them differently?”. The accounts that work treat the two as complementary, fund them deliberately and judge them on different metrics. Below is how we approach the split.

What each side actually does

A working definition matters here, because the words have been stretched.

Demand generation, as we use the term, is paid spend designed to create or surface demand that does not currently exist as a measurable lead. It includes brand awareness ads on LinkedIn, YouTube pre-roll, programmatic display, podcast sponsorship and content distribution. The conversion event is engagement (a video view, a content read, a recurring site visitor), not a form fill.

Lead generation is paid spend designed to capture demand that already exists, in the form of a measurable identity. Form fills, demo requests, gated content downloads, free trials. The conversion event is the lead.

Both produce pipeline. They produce it on different time horizons, through different channels and with very different cost structures. Treating them with the same metric (cost per lead) is what gets accounts in trouble.

Why most B2B tech accounts over-fund lead-gen

The default state of most B2B tech paid programmes is heavy lead-gen. The reasons are easy to understand. Lead-gen produces a number that goes in a spreadsheet within hours of going live. Demand-gen produces a number that goes in a spreadsheet six to nine months later. Marketing leaders under pressure for quarterly results gravitate to the channel that produces visible output.

The problem is that pure lead-gen targets a small audience: the buyers actively searching today. In categories where buying cycles are long and the pool of in-market buyers is small at any given moment, lead-gen alone produces diminishing returns quickly. CPCs rise, the same competitors are bidding on the same terms, and the cost per opportunity creeps up.

Demand-gen widens the pool. By spending against the much larger group of buyers who will be in-market in the next twelve months, the future lead-gen audience grows. Skip demand-gen for too long and the lead-gen channel slowly dries up.

The split we typically recommend

There is no single correct ratio, but the patterns across our B2B tech client base are reasonably consistent.

The temptation is always to under-invest in demand-gen and over-invest in lead-gen, particularly at the early-stage end. We see this most in newer SaaS firms that have raised on the promise of efficient lead-gen and are now squeezing a small auctioning audience for diminishing returns. The lift, when they finally fund a demand-gen layer, is usually visible six to nine months later in the cost per opportunity on the lead-gen channels.

Channels that tend to favour each side

The channel mix typically falls into three broad camps:

Demand-generation: LinkedIn awareness and document ads, YouTube, programmatic display via DV360 or The Trade Desk, podcast sponsorship, sponsored newsletter placements.

Mixed: LinkedIn consideration and engagement campaigns, Google Demand Gen, Microsoft Audience Network, Meta for B2B (genuinely useful in some categories despite the snobbery) and increasingly Reddit ads for B2B where the audience earns it.

Lead-generation: Google Ads search, Microsoft Ads search, LinkedIn lead-gen forms versus conversation ads, retargeting display, gated content syndication.

We cover the LinkedIn-specific structure in our LinkedIn Ads buyer-journey playbook, the search-side mechanics in Google Ads for SaaS and the often-overlooked Microsoft Ads layer.

How to measure each side without confusing yourself

The mistake is judging both sides on cost per lead. Demand-gen will lose that comparison every time, because its job is not to produce leads.

What we measure on the demand-gen side:

• Reach inside the ICP (not total reach)
• Engaged accounts (companies engaging with multiple touches)
• Branded search lift (a useful proxy when other measurement is hard)
• Recurring site visitor growth from target firmographic segments
• Pipeline-influenced revenue with longer attribution windows (90 to 180 days)

What we measure on the lead-gen side:

• Cost per qualified lead (not raw lead)
• Cost per opportunity, not cost per click
• Lead-to-opportunity rate by channel and campaign
• Pipeline-sourced revenue with shorter attribution windows (30 to 90 days)

Both feed into the same multi-touch attribution view, which we cover in attribution models for tech companies with multi-touch journeys. Without a working multi-touch view, demand-gen always looks worse than it is and lead-gen always looks better.

The retargeting bridge

Retargeting sits between demand-gen and lead-gen, and it is where most of the budget arguments actually get won or lost. Demand-gen creates the audience pool, retargeting harvests it, lead-gen converts it. Cut retargeting too tight and the demand-gen budget never pays back. Run retargeting too aggressively and the brand experience suffers.

We cover the discipline in retargeting tech buyers without burning the brand, but the budget point worth flagging here is that retargeting should be funded out of the lead-gen line, not the demand-gen line. The audience that converts is the audience that demand-gen built.

Common mistakes when shifting the budget

Three patterns we see when clients move from a heavy lead-gen mix to a more balanced one.

First, expecting demand-gen to produce attributable leads in the first quarter. It does not. The honest framing for the finance director is “this is twelve-month spend that should produce visible pipeline contribution from quarter three onwards”.

Second, under-funding demand-gen so it never reaches the frequency required to register. A £2,000 monthly LinkedIn awareness campaign across a 50,000-person ICP will produce noise, not recognition. Demand-gen needs scale or it is a waste of money. Better to fund one channel properly than three poorly.

Third, killing demand-gen the moment lead-gen comes under pressure. Demand-gen is the slow compounding investment that makes lead-gen efficient. Cutting it during a quarterly squeeze is the marketing equivalent of selling at the bottom.

What the conversation looks like with sales

The harder conversation, often, is with sales rather than finance. Sales want leads now. Demand-gen produces “engaged accounts” that look like nothing on a CRM dashboard. We typically run a monthly review with sales showing which target accounts are moving from cold to engaged, with a forecast of when each cohort is likely to convert into outreach-worthy contacts. This keeps demand-gen visible to the team that has to convert its output.

If your paid programme is heavy on lead-gen and the cost per opportunity has been climbing for two or three quarters, the answer is rarely a new lead-gen channel. It is usually a demand-gen rebuild. Working through any of this on your own account? Tell us where you’re stuck. You can also see how we run mixed paid programmes on our paid media service page.