The Demand-Led Growth Playbook

Demand-led growth is the next wave of paid media campaigns, and a big step forward from traditional strategies.

To help you get started, we have put together our 8-step framework for implementing demand-led growth in your paid media strategy, covering;

How to audit budgeting behaviour, not just campaign settings
Where to find hidden structural and signal-based constraints in Google’s PMax
How to safely uncap budgets and enable responsive media investment
What KPIs matter most for demand-led growth (beyond vanity metrics)
How to align finance and trading teams to support fluid spend
Ways to empower media teams as orchestrators, not operators
How to structure dashboards, testing, and monthly reviews for sustained optimisation

For more on what demand-led growth is and the impact of fixed budgets on your retail performance, read our blog.

Step 1: Audit budgeting behaviour, not just settings

Your PMax campaigns can only work with the conditions they’re given. Capping budgets and manually adjusting spend may feel like control, but in practice, it constrains PMax’s AI performance. The biggest blocker to demand-led growth isn’t the tech, it’s behaviour.

What to do:
◻ Map where budget caps are regularly hit and for how long, and monitor budget constraint before it reaches ‘Limited by Budget’ status (e.g. 70% constraint scenarios).
◻ Track frequency and causes of manual budget changes (e.g. dips in performance, finance or trading needs).
◻ Analyse auction impression share loss and signs of suppressed bidding.
◻ Run a “budget pressure test”: what would happen if caps were removed tomorrow?

When its working:
Budgets are shaped by real-time demand, not calendar cycles.
No panic reallocations or day-trading led cuts.
Pacing and opportunity signals drive strategy.
Teams trust the system enough to let PMax learn and scale.

This audit should aid alignment with Finance, ensuring budget flexibility decisions are built on shared evidence and agreed guardrails.

Step 2: Align Finance and Trading teams as part of your approach

Demand-led growth requires agreement on what success looks like. Without shared definitions of profitability, payback, and thresholds, flexibility will stall under pressure from Finance or Trading.

What to do:
◻ Confirm ROAS, CPA, or contribution margin targets with Finance partners.
◻ Share diagnostics: auction insights, constraint reports, incrementality testing - more in step 5
◻ Build shared frameworks: category elasticity curves, margin-adjusted thresholds.
◻ Align budget guardrails to trading goals (e.g. stock push, clearance windows).

When its working:
Finance sees PMax as a revenue driver, not a marketing cost.
Budget approvals are linked to opportunity, not just quarterly planning.
Trading and Marketing calendars connected to media pacing.

Step 3: Assess structural and signal-based constraints

It’s not just budgets that restrict growth, poor signals and weak system design (how campaigns, assets, feeds, and audiences are structured) can quietly throttle performance.

How to do it:
◻ Use auction insights to diagnose impression share loss due to rank and budget.
◻ Review feed health, disapprovals, and asset group coverage, engagement levels.
◻ Monitor campaign status, especially prolonged “learning” or stalling after launch.
◻ Track audience match rates, asset eligibility, and conversion delays.

When its working:
Campaigns scale naturally with demand and exit learning quickly.
Impression share loss from rank or budget trending down.
Feed health and asset rotation actively supporting performance.
High match rates on 1st party audiences that fuel lookalike modelling.

Step 4: Uncap budgets in PMax campaigns

PMax is designed to respond to demand in real time. Fixed budgets break that ethos, limiting exposure, learning, and long-term growth.

How to do it:
◻ Remove caps in high-performing campaigns or set generous buffers (e.g. 5× daily spend).
◻ Track how often campaigns hit their cap and what’s being missed.
◻ Align flexibility with key trading periods or category-level seasonal context and objectives.

When its working:
Spend scales naturally with demand.
Little to no impression loss due to budget constraints.
AI takes the lead on pacing and opportunity capture.

Step 5: Set demand-led KPIs and benchmarks

ROAS is a guardrail, not a growth strategy. Demand-led growth requires broader commercial metrics that support strategic scaling.

How to do it:
◻ Incrementality: identify what’s new and additional.
◻ Cost of demand or contribution margin.
◻ TAM coverage: auction share, reach, exposure.
◻ Benchmark spend elasticity: how additional investment impacts returns.
◻ Align with trading priorities and category-specific dynamics.

When its working:
KPIs reflect business outcomes, not just media efficiency.
ROAS is one of many signals, not the only one.

Step 6: Deploy incrementality testing frameworks

If you can’t prove what’s incremental, it’s hard to justify spending more. This is a top priority, because without correct measurement, nothing matters

How to do it:
◻ Design controlled tests: geo splits, holdouts.
◻ Use causal impact models
◻ Use Media Mix Modelling to assess full-funnel and halo effects.
◻ Combine all three measurement approaches for credibility.

When its working:
Regular cadence of clean, credible testing.
Confidence in results across marketing, finance, and trading.

Step 7: Empower your team as AI performance leads

The role isn’t campaign management anymore, it’s shaping Google’s AI’s operating environment.

How to do it:
◻ Define roles around planning, forecasting, and scenario design, not bid/budget changes.
◻ Train teams to interpret signals, guide PMax and define intervention logic.
◻ Run planning sessions based on forecasted demand and commercial priorities with key stakeholder groups across the retail org

When its working:
Humans set strategy, AI executes.
Teams act with clarity, not panic.

Increased cross team collaboration

Step 8: Review and refine monthly

Daily over-optimisation disrupts AI (PMax works to a 30.5 day model). Monthly strategic reviews enable compound performance.

How to do it:
◻ Identify which constraints improved or worsened.
◻ Assess where marginal returns increased or declined.
◻ Reassess guardrails, asset priorities, and measurement approach.

◻ Feed data outputs into Finance and Trading

When its working:
Strategy evolves based on system learning and results.
Cross-functional monthly retros feed next cycle planning.

This checklist sets the stage and removes the brakes on growth. To take the next step to scaling your performance, you need AI.

Talk to our team to find out how our AI & ML platform uses the principles of demand-led growth to scale PMax performance.

Looking for a PDF to download? Click here.