The CMO of a 40-person SaaS company came to us last year with a familiar problem. They had been running LinkedIn lead gen campaigns for two years. Cost per MQL was around $180. Sales qualified about 12% of those leads. Win rate was hovering near 9%. "We think we need demand gen," she said. "Our main competitor keeps showing up in our deals and their content is everywhere. People already like them before the call."
So they switched. Stopped gating content. Started publishing opinionated posts. Pulled back on MQL targets. Three months later, the board wanted to know where the pipeline had gone.
Demand gen got blamed. Budget got cut. They went back to lead gen.
This is the most common way demand gen experiments fail. And it has nothing to do with demand gen being a bad idea.
The problem is that most B2B teams treat demand gen and lead gen as a binary choice when they are actually two parts of the same system that need different tuning depending on where the company is.
What demand generation actually is
Demand generation is the work of making people want what you sell before they go searching for it. It builds awareness, creates preference, and educates your market long before any buying conversation begins.
The key insight: only 5% of your total addressable market is actively looking to buy at any given moment. Demand gen exists to influence the other 95%. If you wait for buyers to find you, you are competing in a small, expensive pool of people who are already comparing vendors and shopping on price.
At Ziel Lab, we count things like this blog, LinkedIn posts, and client case studies in this bucket. The goal is that when someone does start shopping for RevOps or CRM help, they already have a shortlist that includes us because they have been reading our work for months.
What lead generation actually is
Lead gen is the work of capturing people who already have a need. They are actively researching. They will search something specific, click an ad, download a guide, or sign up for a webinar. Your job is to be there, collect their information, and pass them to sales at the right moment.
This includes gated content, paid search, demo request flows, intent-triggered outreach, and any activity that turns an anonymous visitor into a named contact with a buying signal attached. Lead gen works best when the targeting is sharp. If you are capturing people who are not ready to buy, you inflate MQL count with noise and send sales chasing contacts that were never qualified in the first place.
The practical framing: lead gen harvests demand that already exists. Demand gen creates demand that doesn't.
Both matter. Neither replaces the other.
That 79% number deserves a pause. If nearly eight in ten leads you generate never convert, adding more lead gen budget probably will not fix it. You are attracting the wrong people or qualifying them before they have any preference for you. That is a demand generation failure showing up as a lead generation problem.
Why most articles get this wrong
The standard advice is "demand gen creates awareness, lead gen captures it." That is technically accurate and completely useless for a founder or CMO trying to decide where to put their next $50K.
The real question is not which one. It is what ratio makes sense for where your company is right now.
The stage-dependent split nobody talks about
The right demand gen to lead gen ratio changes as your company grows. Most content treats this as static, which is why so many teams apply enterprise-stage thinking to early-stage budgets and wonder why nothing works.
Pre-$1M ARR: you need qualified conversations fast. Lead gen should dominate at roughly 70% of your marketing effort. Demand gen at 30% starts building category awareness, but you cannot afford to wait six months for content to compound. You need pipeline now.
$1M to $10M ARR: this is where most teams get confused. You have enough data to know who buys and why. Lead quality starts to matter more than raw volume because your sales team cannot keep chasing dead leads. A rough 50/50 split makes sense here, shifting toward demand gen as you build content depth, community presence, and organic channels.
$10M ARR and up: your buyers are researching you before they talk to you. Your ICP is defined, your deal size is meaningful, and a single well-qualified account is worth more than 50 low-intent leads. Demand gen at 60 to 70% of your marketing effort, with lead gen focused on high-intent signals like direct demo requests and trial sign-ups.
How to tell if you have a demand problem or a lead problem
This diagnostic usually clears the confusion faster than any framework.
If lead gen is working (volume is there) but win rates are low: demand problem. You are capturing people who found you, but they haven't been pre-sold on your approach. They shop on price, compare you to three alternatives, and take forever to decide. Demand gen fixes this by creating preference before the buying conversation starts.
If win rates are strong but pipeline is thin: lead gen problem. You are not capturing enough of the people who are ready. This calls for more distribution, better intent coverage, or tighter paid channel targeting.
If both win rates and pipeline are poor: start with your ICP. You may be going after the wrong people entirely. No combination of demand gen and lead gen fixes a broken ICP. I have written more on this in our ICP guide here.
I have run this diagnostic across our RevOps and CRM engagements and the pattern holds. Most teams believe they have a lead gen problem when the real issue is demand gen. Adding more lead gen spend to a demand gen problem generates more low-quality conversations that frustrate sales and drive up cost per acquired customer.
RevOps is what connects them
Most articles stop at explaining the two strategies and leave you to figure out how to connect them. RevOps is what makes the connection work in practice.
Three specific things RevOps does that neither marketing nor sales can do alone:
It defines what "qualified" means before a single lead is captured. Not during a post-quarter debrief where sales and marketing argue about who failed. RevOps sets SQL criteria in your CRM ahead of campaign launch: lead score thresholds, ICP fit score, minimum data quality requirements, industry and company size filters. This means demand gen and lead gen are aimed at the same buyer from day one.
It tracks pipeline contribution, not just cost per lead. Demand gen does not show up well in CPL. It shows up in pipeline influenced, deal velocity, and close rates by channel. RevOps builds the attribution reporting that makes this visible to leadership, so demand gen does not get cut the moment MQL volume drops. (More on attribution in our full guide here.)
It closes the feedback loop. When sales disqualifies a lead, that information goes back into targeting. When a deal closes in half the expected time, the signals that predicted it get added to the scoring model. Without this loop, lead gen and demand gen both run on stale assumptions for months.
The tool stack for each approach
This varies by budget and stage. Here is what a typical growth-stage B2B SaaS team uses across these functions.
For lead gen, the automation matters a lot. A lead enters HubSpot, Clay enriches it with firmographics and company data, n8n checks ICP score against your criteria, and HubSpot routes it to the right sequence or rep automatically. This reduces the qualification failure rate because unqualified contacts get filtered or deprioritized before they reach sales. We build these AI automation workflows for clients regularly, and the time-to-qualified metric usually drops by 40 to 60% after the first 30 days.
For demand gen, the stack is simpler. Content distribution, LinkedIn, YouTube, and a way to measure which accounts from your target list show up in your CRM after a content campaign. The hard part is attribution, not tooling.
Your lead gen and demand gen need to share the same definition of a good buyer.
If marketing defines ICP by company size and sales defines it by budget and urgency, you will always have a qualification problem. RevOps exists to make that definition shared, documented, and enforced in the CRM before campaigns launch.
The timeline problem that kills good programs
Cognism, whose case I referenced earlier, saw their pipeline grow from $2M to $13M after correcting their demand gen strategy. Their close rate went from 0.2% to nearly 20% on inbound leads. That did not happen in 90 days.
The realistic timeline for a B2B demand gen program:
Month one to two: strategy is set, new content is live, lead gen continues at current pace. No visible pipeline impact yet.
Month two to three: content starts getting indexed and shared. Early pipeline from accounts that have been consuming content. If you measure success here, it looks flat.
Month four to five: qualified inbound starts increasing. Win rates improve because buyers who found you through content already believe your approach. They close faster and negotiate less.
Month six and beyond: you can see the ROI clearly. Branded search grows. Deal velocity is up. Sourced pipeline from content is measurable.
If your leadership is evaluating demand gen success at month three, you will cut it before it works. This is precisely what RevOps should be communicating upfront: what the success metrics are, when to expect them, and what leading indicators to watch in the interim. Branded search volume, engagement rate from target accounts, time-to-qualified on inbound, and pipeline influenced are the early signals worth tracking.
The CMO I mentioned at the start came back to us four months after that failed experiment. We rebuilt the program with this framing, reset expectations with the board, and swapped MQL count for time-to-qualified as the primary metric. By month five, win rate was up 14 points. The board was satisfied. They had not switched strategies. They had corrected the ratio and measured it properly.
That is usually the fix. Not a complete overhaul. A recalibration with the right metrics and enough patience to see it through.
If you want help running the diagnostic on your own pipeline, you can reach us here.
FAQ
What is the main difference between demand generation and lead generation?
Demand generation builds awareness and preference among people who are not yet actively looking to buy. Lead generation captures people who already have a need and are researching solutions. Demand gen targets the 95% of your market not currently in-market; lead gen captures the 5% who are.
Should B2B companies do demand gen or lead gen first?
For early-stage companies (pre-$1M ARR), lead gen usually makes sense first because you need qualified conversations quickly. As you scale and your ICP becomes clearer, you shift more budget toward demand gen to reduce paid channel dependency and improve lead quality.
How do you measure demand generation success if not by MQL count?
The right metrics for demand gen are pipeline influenced (deals where demand gen content appeared in the buyer's journey), branded search volume growth, time-to-qualified (how quickly inbound leads meet SQL criteria), and close rate by channel. MQL count measures volume; these metrics measure quality and revenue impact.
What role does RevOps play in connecting demand gen and lead gen?
RevOps defines lead qualification criteria before campaigns run, builds the attribution reporting that makes demand gen ROI visible to leadership, and closes the feedback loop from sales disqualification data back to marketing targeting. Without this function, the two strategies run independently on misaligned assumptions.
How do tools like Clay and n8n help with B2B lead generation?
Clay enriches every inbound lead with firmographic and intent data from over 100 data sources, checking ICP fit before the lead reaches sales. n8n automates the orchestration between HubSpot, Clay, and outreach tools, so enrichment, scoring, and routing happen in real time. This combination significantly reduces the percentage of unqualified contacts that reach your sales team.
Not sure which you need more of right now?
Book a free 30-minute session and we'll run the demand vs lead gen diagnostic on your current pipeline data.
Book a free session →