What is the go-to-market playbook for marketing-led demand-gen growth in 2027?

What changes by company stage

The single biggest mistake teams make when reading a demand-gen playbook is treating it as stage-agnostic. A $3M ARR company and a $150M ARR company are running fundamentally different machines under the same label, and copying the wrong one burns twelve months of runway.

Pre-$5M ARR (founder-led demand). At this stage there is no orchestration layer, because there is nothing to orchestrate. You have one or two channels that work and a founder whose personal credibility is the highest-converting asset in the company. The "playbook" is: pick one ICP slice narrow enough that you can name 300 target accounts, produce content that only that slice cares about, and let the founder distribute it personally. Marketing spend should be under 15% of revenue and heavily weighted toward content production and one paid channel used purely for testing. Intent-data platforms are actively harmful here — you'll spend $50K/yr on signals for accounts you have no capacity to work, and the platform's minimum contract will outlive your ICP hypothesis. The metric that matters is *qualified conversations per week*, tracked in a spreadsheet, not a dashboard.

$5M–$20M ARR (repeatability). This is where the playbook starts to look like a playbook. You now have enough closed-won data to reverse-engineer what a good-fit account looks like, and enough traffic that anonymous behavior carries information. You hire your first demand-gen manager and your first marketing ops person — and the ops hire is the one people skip, which is why so many companies at this stage have three tools that disagree about what a lead is. Budget moves to 20–30% of new ARR target. You add a lightweight scoring model, but keep it simple: five to eight signals, not fifty. The failure mode here is premature sophistication. Teams buy an ABM platform before they have a functioning lifecycle in the CRM, then spend a year building attribution on top of dirty data.

$20M–$100M ARR (orchestration). Now the ABM and intent stack earns its keep. You have multiple segments, multiple products or tiers, and a sales org large enough that routing decisions have real cost. Demand gen splits into functions — paid, content/SEO, lifecycle, events, ABM — and someone owns the seams between them. Spend typically lands at 30–40% of the GTM budget. This is where the signal-scoring model in the raw playbook actually applies, and where a CDP starts to justify itself, because identity resolution across anonymous web, product usage, and CRM records becomes a real engineering problem rather than a vendor pitch.

$100M+ ARR (portfolio management). Demand gen becomes a portfolio: mature channels you optimize, emerging channels you incubate, and a constant experiment budget. The interesting problem shifts from "generate more pipeline" to "generate pipeline in the segments where we have margin and win rate," which means marketing gets pulled into pricing, packaging, and expansion motions. Net revenue retention starts competing with new pipeline for marketing attention, and the smartest teams give expansion and cross-sell a named owner inside demand gen rather than tossing it to customer success.

The adjacent read: motion type matters as much as stage. A PLG company at $20M ARR runs a different demand-gen playbook than a sales-led one at the same revenue, because product usage is its highest-quality intent signal and no third-party vendor can sell it to them. A vertical SaaS company selling to hospital systems may skip paid social almost entirely, because the buying committee lives in association conferences and peer referral networks. Read the stage guidance below as a default, then adjust for how your buyers actually buy.

Stage-by-stage playbook

Here's what the actual quarter-by-quarter work looks like when you build the machine in the right order.

Foundation (before any spend scales). Get three things right or nothing downstream works. First, a single definition of an account, a lead, and a qualified opportunity, written down and enforced in the CRM with validation rules. Second, UTM discipline — one naming convention, applied by a tool rather than by humans typing into a spreadsheet, because manual UTM entry is where attribution goes to die. Third, a working lifecycle model with clear entry and exit criteria for each stage, and a report that shows conversion rate and time-in-stage between each pair. Most demand-gen "strategy" problems are actually operations problems wearing a costume.

Layer one: capture. Before you generate new demand, capture the demand that already exists. That means bottom-of-funnel SEO for category and comparison terms, branded paid search defense, a pricing page that doesn't hide pricing behind a form, and review-site presence on G2 or TrustRadius. Capture is unglamorous and has the best cost-per-pipeline-dollar in the entire mix, usually by a wide margin. Teams skip it because it doesn't scale infinitely — you can only capture as much demand as exists — but capturing it first gives you a baseline conversion rate to measure everything else against.

Layer two: creation. Now generate demand that didn't exist. This is content, events, community, partnerships, podcast and newsletter sponsorships, and paid social run for reach rather than direct response. The measurement problem is real: creation shows up as an increase in branded search and direct traffic six to twelve weeks later, not as a form fill with a UTM on it. Budget it as a portfolio with a longer payback assumption, and defend it with cohort analysis rather than last-touch attribution.

Layer three: orchestration. With capture and creation both running, orchestration decides who gets touched, by whom, with what, and when. Signals get weighted — a pricing-page visit from a known target account is worth far more than an ebook download from a student — and crossing a threshold triggers a play. The critical design decision is *play capacity*: never build a scoring model that generates more triggered accounts than your SDR team can actually work in a week. An orchestration engine that fires 400 plays into a team with capacity for 120 doesn't produce pipeline, it produces a backlog and a credibility problem.

Layer four: expansion and recycling. The most underrated pipeline source is accounts you already touched. Closed-lost from 9–18 months ago, churned customers whose blocker has since been fixed, opportunities that stalled at a champion departure. Build a recycling motion with its own triggers — new funding round, new executive hire in the buying role, competitor contract renewal window — and it will frequently out-convert net-new outbound at a fraction of the cost.

Sequencing note. Teams routinely try to start at layer three because it's the most interesting. The result is an expensive orchestration platform sitting on top of a broken lifecycle, firing plays based on scores nobody trusts, producing meetings sales complains about. Build in order.

Numbers that matter at each stage

Benchmarks are directional, not prescriptive — they vary enormously by ACV, sales cycle, and category maturity. Use these as sanity checks on your own trend line, not as targets handed down from a board deck.

Pipeline coverage. The common working target is 3x coverage against quota for the current quarter, but the correct number is derived, not inherited: it's roughly 1 divided by your historical win rate, adjusted for how much of your pipeline is created and closed inside the same quarter. A team with a 33% win rate needs about 3x. A team with a 20% win rate needs 5x and is deceiving itself running at 3x. Recalculate this every two quarters, because win rate drifts as you move upmarket.

Cost per pipeline dollar. Total demand-gen spend divided by pipeline created in the same period. The useful version of this metric is segmented by channel and by stage — a blended number hides the fact that branded search is producing pipeline at a fraction of the cost of paid social while the paid social budget grows. Watch the trend across four quarters rather than the absolute value, and pair it with cost per *closed-won* dollar, since a channel that produces cheap pipeline that never closes is worse than an expensive channel that converts.

Pipeline velocity. Opportunities × average deal size × win rate, divided by sales cycle length in days. This is the only demand-gen metric that captures all four levers at once, which makes it the right number for a weekly executive review. When it drops, decompose it: fewer opportunities is a top-of-funnel problem, longer cycle is usually a qualification or champion problem, lower win rate points at competitive positioning or ICP drift.

Stage conversion rates. Track conversion and median time-in-stage for every transition in your lifecycle. The value isn't the absolute rates — it's finding the one transition where the rate collapsed this quarter. That's your actual problem, and it's usually invisible in aggregate top-line numbers.

Marketing-sourced versus marketing-influenced. Report both, and be honest about the difference. Sourced means marketing created the opportunity; influenced means marketing touched it at some point. Influenced numbers approach 100% in any company with a website, which makes them nearly useless for budget decisions but genuinely useful for arguing against cutting brand spend. Pick one as your primary and don't switch it mid-year to make a quarter look better.

Payback period. CAC payback in months, calculated on fully-loaded sales and marketing cost including salaries, not just program spend. This is the number your board actually cares about, and it's the one that connects demand-gen decisions to the financing conversation. A channel that looks efficient on cost-per-pipeline-dollar but drags payback past 24 months is a growth trap.

A note on early-stage measurement. Below roughly $5M ARR, most of these metrics are statistically meaningless — you don't have enough closed deals to compute a stable win rate, and a single large deal distorts every ratio. Track absolute counts of qualified conversations and opportunities, watch the trend, and resist building dashboards that imply precision you don't have.

Decision framework

When demand isn't hitting plan, the instinct is to add channels or add budget. Both are usually wrong, and the diagnostic order below saves quarters.

Start with the demand type question. Are you in a category buyers already search for, or one you have to teach? If prospects type your category into a search engine, capture is your dominant motion and your problem is probably conversion or coverage, not awareness. If nobody searches for what you sell, no amount of bottom-of-funnel optimization saves you — you need demand creation, and you need to budget for a longer payback.

Then check the constraint. Demand-gen underperformance has four common root causes, and they need opposite responses. *Not enough traffic or reach* means the top of the funnel is the constraint — spend on capture and creation. *Enough traffic, bad conversion* means the problem is the offer, the page, or the ICP fit of who you're attracting — fix targeting or the conversion path before adding spend. *Good conversion, sales rejects the leads* means a definition and routing problem — get marketing and sales in a room with fifty specific rejected records and force agreement on why each one failed. *Sales accepts them but they don't close* means qualification is too loose or the positioning oversells; that's a product-marketing problem, not a demand-gen one.

Match the play to the account tier. Not every account deserves the same investment. A workable default: top-tier accounts (usually 50–150) get 1:1 treatment with custom research and executive engagement; mid-tier gets 1:few by segment or use case; the long tail gets programmatic — SEO, self-serve, lifecycle email, and broad paid. Trying to run 1:1 across the whole target list is the fastest way to burn a demand-gen team out and produce nothing at any tier.

Decide what to kill, not just what to add. Most demand-gen budgets accumulate channels the way a garage accumulates boxes. Run a quarterly review where every channel must re-justify its allocation against cost per closed-won dollar and payback, and kill the bottom decile rather than shaving 10% off everything. Uniform cuts protect underperformers and starve the channels that were about to compound.

Related questions

Should marketing own an SDR team?

It depends on where the handoff friction is. Marketing-owned SDRs tend to be better at following up on inbound signals and worse at cold prospecting. Sales-owned SDRs invert that. Many companies split it — inbound response under marketing, outbound under sales.

How long before demand creation shows results?

Typically two to four quarters before it shows up cleanly in pipeline, longer in categories with 9+ month sales cycles. Measure leading indicators in the meantime: branded search volume, direct traffic, and inbound requests that mention your content unprompted.

Is ABM a replacement for demand gen?

No — ABM is a targeting and prioritization discipline layered on top of demand gen, not a substitute for it. It concentrates existing capabilities on named accounts. If your underlying content, conversion, and follow-up are weak, ABM concentrates the weakness.

What's the right marketing-to-sales headcount ratio?

Highly variable, but in sales-led B2B a common range is one demand-gen marketer per five to eight quota-carrying reps, with marketing ops added around the point where you have three or more marketers. PLG companies skew far more marketing-heavy.

Do I need a CDP or is my CRM enough?

Most companies under $20M ARR do not need a CDP. The problem it solves — stitching anonymous behavior, product usage, and CRM records into one identity — only becomes expensive enough to buy a tool for once you have multiple data sources and real traffic volume.

FAQ

How much of the budget should go to brand versus performance?

There's no universal split, but the common failure is running 100% performance and wondering why cost per acquisition climbs every quarter. Performance harvests demand; brand creates it. A reasonable starting posture for a growth-stage B2B company is to protect a meaningful minority of the budget for creation work with a longer payback, then defend it with branded-search and direct-traffic trends rather than last-touch reports it will always lose.

What should we do if sales says our leads are bad?

Stop arguing in the abstract and pull fifty specific rejected records. Categorize each rejection: wrong ICP, wrong timing, wrong persona, no real interest, or reject-by-reflex. The distribution tells you whether it's a targeting problem, a definition problem, or a trust problem, and each has a different fix. This meeting is worth more than a quarter of dashboard work.

Is outbound dead in this playbook?

No, but untargeted outbound is. Outbound works when it's triggered by a real signal — a hiring pattern, a funding event, a technology change, a champion moving to a new company — and fails when it's volume-based spray. The shift is from outbound as a separate silo to outbound as one execution channel inside a signal-driven orchestration layer.

How do we handle attribution when buying committees have eight people?

Move from lead-level to account-level measurement. Track how many contacts within a target account have engaged, from which functions, and how that engagement breadth correlates with win rate. Single-threaded opportunities close at meaningfully lower rates in most sales orgs, which makes buying-committee coverage a leading indicator worth reporting weekly.

What's the first hire for a demand-gen function?

A generalist who can write, run paid, and read data — not a specialist. The second hire, sooner than most teams think, is marketing operations. Ops is the hire that determines whether everything after it is measurable, and companies that defer it end up rebuilding their entire lifecycle two years later.

Should we build the tech stack before or after the process?

After. Tools encode process; if the process is undefined, the tool encodes chaos and makes it expensive to change. Run the motion manually until it's painful, document exactly where the pain is, then buy the tool that removes that specific pain. This sequence also gives you a far stronger negotiating position, because you know precisely which features you need.

Sources

• Gartner — B2B Buying Journey research
• Forrester — B2B Marketing research
• HubSpot Research — marketing and sales benchmark reports
• OpenView Partners — SaaS benchmarks
• Bessemer Venture Partners — State of the Cloud
• Google — Search Central documentation
• G2 — software review and buyer intent platform
• Salesforce — CRM platform documentation

Related on PULSE

• [Inbound demand-capture GTM playbook in 2027](/knowledge/gp0511)
• [Sales-assisted PLG for mid-market in 2027](/knowledge/gp0510)
• [Reseller and VAR channel GTM playbook in 2027](/knowledge/gp0509)
• [International and geo-expansion GTM playbook in 2027](/knowledge/gp0508)
• [Vertical SaaS go-to-market playbook for healthcare in 2027](/knowledge/gp0507)