Two philosophies
Lead generation is the older model and the one most marketing automation was built for. Content is produced to be exchanged for an email address: the report behind a form, the webinar registration, the lead form inside a LinkedIn ad. Each contact is scored, nurtured by email and passed to sales when the score or the behaviour crosses a threshold. The programme is judged on the volume and cost of leads, and the FunnelModel of the customer journey stages. from visitor to lead to marketing-qualified lead to opportunity is the picture everyone shares. Its strength is accountability: every pound can be traced to a contact, and the CRM lead stages guide describes the stage model it feeds.
Demand generation starts from the observation that most of the category is not in market and will not fill in a form, and that forcing them to does not change that. Content is published where buyers already are, ungated: the report is a web page, the webinar is on the channel, the expert's thinking is in the feed. The aim is to be known, trusted and remembered by the whole category, so that when a buying window opens the vendor is already on the list. Research from LinkedIn's business-to-business research institute supports the premise: buyers who are not in market now will buy later from the vendors they already know. The programme is judged on pipeline and on leading indicators of memory, such as branded search and self-reported attribution, and it hands sales fewer, warmer conversations rather than lists.
Compare table
Lead generation vs Demand generation
| Row label | Lead generation | Demand generation |
|---|---|---|
| What it optimises for | Contacts captured at an acceptable cost per lead | Buyers who know and prefer the brand before they are in market; pipeline later |
| Core assumption | Interest can be captured and nurtured into a purchase | Most of the market is not looking; be remembered for when it is |
| Content | Gated: forms in front of reports, templates, webinars | Ungated: published freely where buyers research |
| Primary channels | Search, lead-form ads, email nurture, events with registration | Organic social, video, podcasts, communities, expert content, reach advertising |
| What sales receives | A list of contacts with a score and a source | Inbound requests from buyers who have already decided to talk |
| Main metrics | Leads, cost per lead, marketing-qualified leads, lead-to-opportunity rate | Pipeline by source, cost per opportunity, branded search, self-reported attribution |
| Attribution | Click-based and form-based; works within the platform reports | Multi-touch and self-reported; much is unattributable by design |
| Time to read results | Weeks to a quarter | Two or more sales cycles |
| Relationship with sales | Hand-off by stage and service level; disputes over lead quality | Fewer hand-offs; disputes over whether marketing did anything |
| Typical failure | Many leads, few opportunities; cost per lead rises as demand is exhausted | Awareness that is never captured; budget cut before it pays back |
| Works best when | Demand already exists and the product is searched for; sales capacity is high | The category is crowded or new; the buyer is risk-averse and shortlists from memory |
| Budget pressure | Easy to defend monthly; easy to over-fund | Hard to defend quarterly; easy to under-fund |
Where lead generation wins
When the demand exists and the product is searched for. A buyer typing the category term wants to be captured; a form, a demo button and a fast follow-up are the right answer. Making them read ungated content first would be a disservice.
When sales capacity is the constraint. A new sales team needs conversations to learn from, and a steady flow of scored contacts gives them something to work while demand creation is still compounding.
When the content genuinely deserves a form. A benchmark report built from original data, a calculator, a tool, an assessment: things buyers will trade an address for, and would expect to. The problem with gating is not the form; it is gating material that nobody would pay for.
When accountability is non-negotiable. A programme that must report cost per lead monthly to a board that will not accept lagged pipeline needs lead generation as its base, with demand generation argued for alongside it. The LinkedIn targeting and lead-form guide covers the most common capture mechanism on the channel most business-to-business teams use. The lead generation objective hub sets out the full KPI set and which channels can carry the objective.
Where demand generation wins
When most buyers are not in market. In any category with long replacement cycles, the share of organisations looking this quarter is small. Lead generation harvests that share quickly and then runs dry; demand generation grows the share that will consider the vendor next year.
When the buyer is risk-averse. Committees shortlist from memory and from what peers say, and they prefer the vendor nobody will question. Being familiar before the window opens is worth more than a well-timed form.
When the category is crowded or new. A new category has no search demand to capture; a crowded one has search demand that costs too much. Both reward the vendor that owns the ideas buyers use to think about the problem.
When sales are drowning in bad leads. A high volume of unqualified contacts is expensive to work and damages trust between the teams. Demand generation produces fewer conversations with people who have already decided to talk, which sales prefer even when the total is lower. The business-to-business content guide covers what to publish for each role so that those conversations are informed.
Running both
The practical answer for most teams is a demand-generation programme that is captured by lead generation, with both measured on pipeline.
Divide the content, not the budget. Gate the few assetsContent a visitor can reach only after giving contact details, usually through a form. that merit it and publish everything else. A report can be a web page and a downloadable summary; the page builds memory, the download captures the few who want it.
Capture intent, not interest. Forms belong on the demo, pricing, trial and contact paths, and on assets with real exchange value. Lead scoringAssigning points to leads for fit and behaviour to decide which are ready for sales. should weight those signals far above downloads, as the lead scoring guide sets out, so that sales receive buyers rather than readers.
Measure on one scoreboard. Pipeline and cost per opportunity by source, with self-reported attribution on every form to catch the demand that the click-based model cannot see. The choice of AttributionAssigning conversion credit to touchpoints. model is a decision in its own right, covered in the attribution approach guide; the point here is that lead volume is not the KPIA metric tied directly to an objective. for either programme. The share of that pipeline whose original source is marketing is the marketing-sourced pipelineValue of sales opportunities whose original source was a marketing activity., read beside the influenced figure.
Protect the creation budget in writing. Agree a share for demand creation for the year, with the leading indicators it will be judged on, before the first monthly cost-per-lead report invites someone to cut it. Customer acquisition cost (CACCost to acquire one new customer.) read by channel will always flatter capture; read by cohort over a year it tells the truth. For software sold by subscription, where the purchase is decided again at every renewal, the SaaS industry guide explains how self-serve and sales-led buying change this balance.
A software vendor rebalances
An HR software vendor runs a lead-generation programme: £40,000 a month across search, LinkedIn lead-form ads and four gated reports, producing 900 leads a month at £44 each. Sales accept 12% as qualified and 2% become opportunities, so each opportunity costs about £2,200 and the sales team spends most of its time disqualifying downloads. The vendor ungates three of the four reports, keeps a form on the one built from original salary data, moves £15,000 a month into an expert video series and a reach campaign, and restricts lead scoring to demo, pricing and trial actions. Six months later lead volume has fallen to 380 a month, but 31% are accepted, 7% become opportunities and cost per opportunity is £1,500. Self-reported attribution shows 44% of new opportunities first heard of the vendor through the video series. Marketing-sourced pipeline is up 35% on the same quarter a year earlier.
To map content to intent across a buying committee, See-Think-Do-Care for lead generation shows how clusters are read per account and why Do means a sales conversation.
Teams running lead programmes can set a spending ceiling with the break-even cost per lead walkthrough, which values each enquiry from deal size and the share that close.
When a programme is said to produce no leads, marketing not generating leads checks whether demand-creation work is simply being judged on a lead window it was never built to meet.
If a demand programme is about to be cut for producing too few leads, the demand generation diagnosis checks the horizon, the missing self-reported sources, the path for ready readers and the breadth of reach first.
The gating argument at the centre of the two philosophies can be settled for one asset with the gated content test, read on qualified leads per visitor rather than form fills.
The practical question of which individual assets to put behind a form is covered in the comparison of gated and ungated content.
Common mistakes
- Calling a download a lead. A person who wanted a template is not a buyer; treating them as one wastes sales time and inflates every rate downstream.
- Gating everything. The form costs more in lost readers than it earns in addresses, and the readers lost are the ones who would have remembered the brand.
- Cutting demand generation because its cost per lead is high. It is not meant to produce leads; judged on cost per opportunity and on pipeline it usually wins.
- Running demand generation with no capture. Awareness that never meets a demo button is a gift to competitors with better forms.
- Judging both on the same window. Lead generation reads in weeks; demand generation reads in sales cycles. Comparing them at one month decides the argument before it starts.
- Ignoring self-reported attribution. It is the cheapest way to see what the click-based model misses, and the field takes a minute to add.
- Letting the argument become tribal. The demand section treats them as two halves of one job, which is what they are.
