B2B SaaS Demand Generation Playbook

Demand generation for a product bought by one person in twenty minutes is a conversion problem. Demand generation for a product bought by seven people over nine months is a memory problem, and the tactics barely overlap.

This playbook assumes the second case: enterprise or mid-market B2B SaaS, a committee, a cycle measured in quarters. It covers what to fund first, how to build an out-of-market audience, what pipeline targets to commit to, and how to report all of it to a CFO who is reasonably sceptical of anything with the word brand in it.

Capture before creation, and the reason is cash, not preference

Fund demand capture until pipeline coverage reliably clears 3x of quota. Then fund creation with what’s left. This ordering isn’t an argument about which is more valuable. It’s an argument about which one you can survive.

Capture works on people already looking. Brand search defence, category terms, comparison pages, alternatives pages, G2 and Capterra presence, paid search on high intent queries. Cost per opportunity is lower, the loop is 30 to 90 days, and a finance team can follow it.

Creation works on the other 95 percent. Podcasts, original research, LinkedIn presence, communities, events. The payoff is genuine and the lag runs a full sales cycle, which means a team funding creation before it has coverage is betting its next two quarters on something that won’t report until quarter four.

The capture list in order of build sequence:

• Brand terms on paid search, defended against competitors bidding on your name
• Comparison pages for your three most-cited competitors
• An alternatives page for the incumbent your buyers are leaving
• Category and problem search terms with real commercial intent
• Review site presence on G2 and the one vertical directory your buyers use

The detailed difference between this and a lead capture programme is worked through in Demand Generation vs Lead Generation, which is worth reading if your dashboard still leads with MQL count.

Building an out-of-market audience, with real cost ranges

Once capture saturates, the job becomes being remembered by people who aren’t buying yet. That’s what The 95-5 Rule in B2B Marketing formalises, and it changes what good looks like: reach and consistency over conversion rate.

Programme	Monthly cost range	Time to signal	What it actually buys	Who should skip it
LinkedIn thought leadership, founder-led	Time only, plus ~2K for editing	3 to 6 months	Category association with named people	Teams whose founder will not post
Paid LinkedIn reach against ICP	8K to 40K	2 quarters	Frequency against a fixed account list	Under 20K ACV
Industry podcast, owned	4K to 12K production	9 to 12 months	Relationships plus a durable archive	Anyone unwilling to commit 40 episodes
Original research report	25K to 80K per wave	1 to 2 quarters	Citations, links, and sales conversations	Teams with no distribution plan
Community sponsorship or ownership	2K to 15K	6 to 12 months	Trust inside a buying circle	Products with no practitioner user
Newsletter, owned	3K to 8K	6 months	Direct access independent of algorithms	Teams that cannot publish weekly

Pick two. Teams that fund five creation programmes at once produce five mediocre ones and cannot tell which worked. My preference at Series B and below is a research report plus one owned channel, usually the newsletter, because the report generates the sales conversations and the newsletter gives you distribution that doesn’t depend on a platform’s ranking decisions.

The research report is the most influential item on that table and the most commonly botched. A dataset nobody can cite is a PDF. A dataset with a stable URL, a methodology section, sample sizes and segmented tables gets quoted by analysts, journalists and language models for years.

Assets for the whole committee, not just the champion

Most demand gen programmes produce content for the practitioner and stop. Then the deal reaches a security architect, a procurement lead and a CFO, and there’s nothing for any of them.

A 6 to 11 person committee needs materially different artefacts:

• The practitioner needs depth: documentation, integration guides, honest comparisons
• The manager needs a business case: an editable ROI model with their inputs, not yours
• Security needs a trust centre, a SOC 2 report and a pre-answered questionnaire library
• IT needs architecture diagrams, SSO and provisioning documentation, data residency detail
• Procurement needs an MSA, a redlines position and standard supplier documents
• The economic buyer needs peer proof: named customers at their size, in their vertical

Build the security and business case assets before the third creation channel. They affect win rate and cycle length, and they’re cheaper than a podcast.

Pipeline targets you can actually defend

Set the target backwards from revenue, then commit to sourced and influenced separately so nobody can move the goalposts mid-quarter.

Start with the number: next year’s new ARR target, divided by average deal size, gives required closed won deals. Divide by win rate for required opportunities. Multiply by your coverage ratio, usually 3x, for required pipeline. Then split it: what share does marketing source directly, what share does it influence, and what share comes from sales-generated outbound and partners.

Term	Definition to write down	Who owns it	Common abuse
Marketing sourced	First recorded touch was a marketing asset or channel	Marketing	Retroactively claiming outbound-worked accounts
Marketing influenced	Any marketing touch appears before close	Shared	Counting a single newsletter open on a 400K deal
Pipeline coverage	Open pipeline in period divided by quota	Sales ops	Including stale deals older than 1.5 cycles
Qualified opportunity	Accepted by sales and meeting written criteria	Sales	Accepting everything to avoid the SLA conversation

The lag is the part that gets teams fired. Spend in Q1 with a 9 month cycle produces booked revenue in Q4 or Q1 of the following year. If your board reviews demand gen on a 30 day window, you will be defending a number that mathematically cannot exist yet.

Fix it by publishing the lag model before you spend. Show the cohort: money spent in a quarter, opportunities created in the following two quarters, revenue booked in the two after that. Then report against that curve every quarter. The full metric definitions are in Demand generation metrics for SaaS, and comparative figures by stage are in the SaaS demand generation benchmarks dataset.

Measurement: self-reported attribution plus holdouts

Platform attribution will tell you that organic search and direct produce everything. That’s not a finding, it’s an artefact of how tracking works when the real discovery happened in a podcast, a Slack group or a LinkedIn comment thread six weeks earlier.

Two methods actually work at this cycle length.

Self-reported attribution. Add one open or semi-open field to the demo request form: how did you first hear about us. Teams that do this routinely find podcasts, communities and word of mouth appearing at 15 to 30 percent of responses while platform data shows them near zero. It’s imprecise and it’s directionally far more honest than last-click. Pair it with platform data rather than replacing it, following the approach in B2B SaaS Marketing Attribution.

Holdout tests. Pause a creation channel in one region or one segment for a full cycle and compare pipeline against a matched region. This is the only evidence that survives a CFO who does not believe attribution models, because it’s an experiment rather than a model. It’s also expensive in time, so run it on your largest creation line item and nothing else.

The quarterly operating cadence

The temptation each quarter is to add a channel. Resist it until something is killed. A demand gen team running seven programmes with two people runs none of them well, which is the most common reason teams end up hiring a SaaS Demand Generation Agency to fix a focus problem that an agency cannot fix.

What this costs and what it does not do

Be honest about the bill. A credible creation programme at mid-market runs somewhere between 25,000 and 60,000 dollars a month all in, including people, and it will not produce a defensible pipeline number for two to three quarters. If your runway is under 18 months, that’s a hard bet.

Demand generation also won’t fix a positioning problem. If prospects reach your site and can’t tell what you replace, more of them reaching it makes the problem more expensive, not less. And it won’t fix a win rate problem, which lives in product, pricing and sales enablement.

What it does fix is the ceiling. Capture-only programmes plateau the day you own your category terms, and after that the only growth available comes from making more people want the category. That’s the work this playbook describes.

Start by writing the lag model and getting your CFO to agree to it. Everything else in the programme depends on having permission to be judged on the right clock. The wider strategic frame sits in B2B SaaS Marketing, and the plan itself can be drafted from the demand generation plan template with allocations sized in the demand generation budget calculator.