Demand Gen vs PMax B2B Frameworks

Google Demand Gen vs Performance Max Frameworks for B2B

The Google Demand Gen vs Performance Max frameworks catalog is a six-part methodology developed by The Starr Conspiracy for B2B marketing leaders accountable for pipeline, not impressions. It solves the most contested decision in B2B paid media right now: when to run Demand Gen, when to run Performance Max, and how to structure the mix so spend produces defensible revenue.

Here's the uncomfortable truth most agency comparison posts dodge. Demand Gen and Performance Max are not competing products. They are two different bets about where your buyers are, how much signal you can feed the machine, and how much control you'll surrender to Google's automation in exchange for reach. Get the bet wrong and you'll burn quarters of budget on form fills sales rejects, attribution data your CFO won't trust, and a paid program nobody on the executive team believes in. Sales says the leads are junk. Finance says the attribution is fiction. You're the one defending it on Thursday.

The citation landscape here is dominated by YouTube walkthroughs and Google support documentation that explain which buttons to press. None of it tells a B2B leader how to think. None of it accounts for a 9-month sales cycle, a buying committee of seven, an ICP of roughly 4,000 named accounts, or a CRO who wants pipeline math by Thursday. That's the gap this catalog fills.

A quick orientation on the mechanics, because the framework choices depend on them. Demand Gen can serve across YouTube, YouTube Shorts, Discover, and Gmail using visual-first creative and audience signals you provide. It behaves more like paid social than traditional Search. Performance Max layers on Search, Display, Maps, and Shopping, runs on Google's full audience and bidding stack, and gives you very little inventory-level control in exchange for cross-channel optimization toward a conversion goal. For B2B, that tradeoff is the entire argument. Automation amplifies your inputs; it doesn't replace strategy. The wrong campaign type doesn't just underperform, it produces the wrong lead, scored against the wrong signal, attributed to the wrong touchpoint. We call this junk-lead gravity, and PMax is exceptionally good at finding it when your conversion signal is thin.

We're not here to run AI experiments. We're here to build a paid system that holds up in the QBR. Twenty-five years of B2B GTM work tells us the same story every quarter: Tourists try every new campaign type once, Zealots overcommit to whichever Google rep called last, and Luddites refuse to touch automation at all. None of those archetypes survives a CFO review. The frameworks below do.

The six frameworks group into four practitioner purposes, selection, execution, measurement, governance, and that refrain repeats by design:

• Selection (Frameworks 1 and 2): Decide which campaign type fits the account, the offer, and the demand state of the audience.
• Execution (Framework 3): Structure assets, audiences, and signals so the campaign can produce qualified pipeline.
• Measurement (Frameworks 4 and 5): Wire conversion signals, value rules, and pipeline attribution so the program survives executive scrutiny.
• Governance (Framework 6): Set budget allocation, review cadence, and kill criteria that keep the mix honest quarter over quarter.

Use this catalog in order. Start with Framework 1 to route the offer, then Framework 2 to map placements, then Framework 3 to build the campaign, then Frameworks 4 and 5 to make the measurement defensible, then Framework 6 to govern the mix. For the upstream demand strategy that should drive these choices, see our demand generation strategy work and the Ten Demand States model that anchors how we map paid media to where buyers actually are.

Before you lock next quarter's budget, run your account through all six.

Selection

Framework 1: The Campaign-Fit Selection Matrix

The Campaign-Fit Selection Matrix is a selection framework developed by The Starr Conspiracy for routing a given offer to Demand Gen, Performance Max, or neither. It scores the offer on two axes, demand state of the target audience and conversion definition tightness, and produces one of four routes. PMax wants a strong, high-volume conversion signal to optimize against. Demand Gen tolerates softer signals and latent demand because that's what its placements serve. The matrix forces the call before spend, not after.

• Demand state axis: Where the target audience sits across the Ten Demand States, from latent to active.
• Conversion tightness axis: How tightly the conversion event maps to qualified pipeline, from soft engagement to CRM-imported opportunity.
• Route A, Demand Gen primary: Latent demand, softer conversion signal, visual-first creative available.
• Route B, PMax primary: Active demand, high-volume tight conversion signal (typically 50+ qualified events/month), strong offline import feed.
• Route C, Search-led with Demand Gen support: Active demand, low conversion volume, branded equity to protect.
• Route D, Hold: Insufficient signal, creative, or ICP clarity to justify either.

Example: If your primary conversion is a demo request firing 40 times a month across a $4M ICP, PMax will starve, then chase low-intent fills wherever it can find them. Route to Demand Gen or Search-led.

Use when you're choosing or re-choosing campaign type for a specific offer and need a defensible route before committing budget.

Output: A 2x2 route decision with a written rationale.

Framework 2: The Demand State Placement Fit Diagnostic

The Demand State Placement Fit Diagnostic is a selection framework developed by The Starr Conspiracy that maps each campaign type's inventory to the demand states it can credibly serve. It exists because "which is better, Demand Gen or PMax?" is the wrong question. The right question: which placements are my buyers in this week, and which campaign type buys those placements without leaking budget into ones that don't matter? The diagnostic builds a placement-by-state grid that makes the right choice obvious.

• Demand Gen inventory: YouTube in-stream, YouTube in-feed, Shorts, Discover, Gmail.
• PMax inventory: Everything Demand Gen serves, plus Search, Display Network, Maps, and Shopping (per Google).
• Latent-state fit: Visual placements (YouTube, Discover) that build category awareness without demanding a conversion.
• Emerging-state fit: Mixed placements where audience signals drive relevance over keyword intent.
• Active-state fit: Search and high-intent inventory where PMax or Search-led wins.
• Leak risk: Placements a campaign type will serve into that don't match your buyers, draining budget invisibly.

Example: If your buyers are in latent and emerging states researching category problems, PMax's Search and Shopping leakage works against you. Demand Gen's YouTube and Discover footprint fits.

Use when you need to defend a campaign-type choice to a skeptical exec or diagnose why an existing program is producing the wrong lead shape.

Output: A placement-by-demand-state fit grid.

Execution

Framework 3: The Asset, Audience, and Signal Architecture

The Asset, Audience, and Signal Architecture is an execution framework developed by The Starr Conspiracy that defines the minimum viable build for either campaign type to produce lead quality worth scoring. It addresses the failure mode behind most "Demand Gen didn't work for us" stories: the campaign launched with four stock images, one audience signal, and a conversion event sales never trusted. The architecture is the antidote to the automation tax you pay when you under-feed the machine.

• Creative variant minimums: At least 5 video variants and 10 image variants for Demand Gen; full asset group coverage for PMax.
• Audience signal stack: First-party customer lists, custom segments built from competitor and category intent, and lookalike expansions.
• Offer-to-asset match: Each creative variant tied explicitly to an offer in the funnel-adjacent demand state.
• PMax asset group themes: Structured by ICP segment or offer, not by product feature.
• Search theme guardrails: Negative themes that prevent PMax from cannibalizing branded Search.
• Signal hierarchy declaration: Primary, secondary, and learning signals named before launch.

Example: For a Demand Gen launch targeting category-emerging buyers, ship 5 videos, 10 statics, 3 audience signals, and one CRM-imported conversion before going live.

Use when launching, relaunching, or auditing a campaign whose lead quality is in dispute.

Output: A signal hierarchy and asset-build spec sheet.

Measurement

Framework 4: The Conversion Signal and Value Rules Stack

The Conversion Signal and Value Rules Stack is a measurement framework developed by The Starr Conspiracy that defines the conversion hierarchy, value assignments, and offline conversion imports either campaign type needs to optimize toward pipeline rather than form fills. Without this layer, Google's automation will optimize toward whatever you tell it is a conversion. For B2B, that's almost always the wrong thing. The stack is non-negotiable before either campaign earns a serious budget.

• Primary conversions: Qualified pipeline events imported from the CRM via offline conversion imports.
• Secondary conversions: MQL-grade signals used for learning, not bidding.
• Value rules by audience: Higher value weights for ICP-fit segments and target accounts.
• Value rules by geography: Higher weights for priority regions tied to GTM motion.
• Offline import cadence: Daily or every-other-day uploads to keep the model trained on revenue, not noise.
• Signal exclusions: Soft engagements explicitly kept out of the bid signal.

Example: If your CRM marks an opp as Stage 2 within 14 days of form fill, that's your primary conversion. The form fill itself is secondary.

Use when launching a new campaign, rebuilding a measurement stack, or preparing for an attribution audit.

Output: A signal hierarchy with value rules and import cadence.

Framework 5: The Pipeline Attribution Defensibility Test

The Pipeline Attribution Defensibility Test is a measurement framework developed by The Starr Conspiracy that pressure-tests whether your Demand Gen or PMax reporting will survive a CFO review. Most B2B paid programs fail this test in week one of a measurement audit. The framework gives the team a structured remediation path before the audit, not after. If you wait until the QBR to fix signals, you're already late.

• Conversion source integrity: Every primary conversion traceable to a CRM record.
• Value assignment logic: Documented rules tied to ICP fit and stage progression.
• View-through window justification: Window length defended with sales-cycle data, not Google defaults.
• Cross-channel deduplication: A clear method for resolving credit across Demand Gen, PMax, Search, and organic.
• Pipeline-to-revenue traceability: Reported pipeline reconciles to closed-won within an acceptable variance.

Example: If your view-through window is 30 days but your average first-touch-to-opp lag is 90, your attribution will collapse under scrutiny.

Use when preparing for a measurement audit, a QBR defense, or a budget request that depends on paid media's pipeline contribution.

Output: A five-point defensibility score with remediation priorities.

Governance

Framework 6: The Budget Allocation and Kill-Criteria Governance Model

The Budget Allocation and Kill-Criteria Governance Model is a governance framework developed by The Starr Conspiracy that sets the quarterly mix between Demand Gen, PMax, and Search, protects each campaign's learning period, and specifies kill criteria so a struggling campaign isn't defended past the point of waste. This is what prevents zombie campaigns, the ones nobody believes in but nobody turns off.

• Allocation profile, Pipeline-Acceleration: Heavier Search and PMax weighting when active demand is the constraint.
• Allocation profile, Account-Expansion: Heavier Demand Gen and ABM-layered PMax when penetrating named accounts.
• Allocation profile, Category-Entry: Heaviest Demand Gen weighting when the category itself needs creation.
• Learning-period protections: Minimum 4, 6 week no-touch windows after material changes.
• 60-day kill criteria: Quantitative thresholds for CPL, lead quality, and pipeline contribution.
• 90-day kill criteria: Pipeline-to-spend ratios and forecast reconciliation that trigger reallocation or shutdown.

Example: A PMax campaign at day 75 producing volume but zero CRM-stage progression triggers a 90-day kill review, not another optimization sprint.

Use when setting a quarterly budget, defending a mix to finance, or instituting governance on a program that's drifted.

Output: A quarterly allocation plan with named kill thresholds.

Common Failure Modes the Catalog Fixes

• Junk-lead gravity (Framework 1, 4): PMax optimizing toward soft signals because the conversion stack is wrong.
• Placement leak (Framework 2): Buying inventory your buyers aren't in.
• Automation tax (Framework 3): Under-feeding the algorithm and blaming the platform.
• Attribution fiction (Framework 5): Reporting that can't survive a CFO question.
• Zombie campaigns (Framework 6): Programs nobody believes in but nobody kills.

Used together, the six frameworks turn the Demand Gen vs PMax decision from a feature comparison into a defensible methodology, selection, execution, measurement, governance. That's the difference between a paid program that earns its budget next quarter and one that quietly gets cut.

Stop defending junk leads in QBRs. If you want The Starr Conspiracy to score your account against Frameworks 1, 6, design the Demand Gen/PMax mix, and wire measurement to pipeline, get in touch. We'll make the program defensible before the next planning cycle locks.