What I'd Do With an 8-Figure Marketing Budget

Managing a large marketing budget isn’t just about finding new ways to spend money – it’s about protecting, growing, and multiplying every pound that goes out the door. So if you handed me an 8-figure budget today, here’s exactly how I’d approach it to drive sustainable, scalable performance.

Step One: Find Out What’s Actually Driving Results Today

Before shifting a single pound/dollar/euro, I’d dig deep into the current performance data. And I’m not just talking about last-click attribution – that’s a narrow lens. I’d be looking at every touchpoint across the journey:

• First-click and multi-touch data
• Platform engagement metrics
• Post-purchase surveys
• CRM insights

You need to see the full picture. What are the channels that initiate demand? Which campaigns influence consideration? And what finally converts the customer?

Once I’ve mapped that against the broader consumer journey, I’d get a clear view of what’s genuinely moving the needle. That becomes the baseline. From there, I’d calculate how much of the current budget needs to stay put just to sustain existing growth. That’s my foundation – the cost of maintaining momentum.

Step Two: Run Incrementality Tests (In the Quiet Periods)

The next move? Figure out what’s truly incremental. I’d design a series of controlled experiments to isolate the impact of key campaigns and channels. But I’d do it during non-peak periods – not Q4 when noise is at its highest.

You’d be surprised how many brands over-invest in channels that are really just harvesting intent rather than generating it. Incrementality testing helps uncover where the net-new customers are really coming from – and where budget could be better deployed.

Once those insights are in, I’d adjust the media mix accordingly. Some channels may shrink. Others – the ones actually fuelling growth – would get scaled aggressively.

Step Three: Expand Into 3 to 5 New Channels (With Purpose)

With the core engine running smoothly and the incremental channels dialled in, I’d earmark budget for strategic expansion. But I wouldn’t just throw money at whatever’s trending.

I’d choose three to five channels with real potential to capture attention and generate demand. Think YouTube, Connected TV, podcasts, high-quality newsletter sponsorships. Not coupon sites or affiliate arbitrage plays.

Here’s the key: I’d treat each test like a full campaign launch. That means a clear hypothesis, a tailored media strategy, bespoke creative assets, and a commitment to optimising over time. No one-off banner ads slapped onto a new platform. If you’re going to test, do it properly.

Step Four: Build a Layered Measurement System to Track It All

When you’re working with this level of budget, you can’t afford to rely on surface-level reporting. I’d build a multi-layered measurement framework aligned to both short-term performance and long-term profitability. Here’s how I’d structure it:

• Daily – One-day click ROAS targets for digital, and ratios of paid vs organic/owned traffic. This gives visibility into pacing and channel efficiency.
• Monthly – Channel-level budget shifts based on actual contribution to revenue. What’s over-performing? What’s dragging?
• Quarterly – Incrementality studies and media mix modelling to gauge long-term impact. This is where strategy sharpens.

In short, I’d be measuring everything that matters – not just what’s easiest to track.

A Worked Example: Splitting a £10m Budget

Talking about allocation in the abstract is easy. Actually putting numbers on it is where most people freeze. So here’s an illustrative split on a round £10m to show the shape of it – not a prescription. It borrows the spirit of the classic 70/20/10 rule (roughly 70% proven, 20% emerging, 10% experimental), which is best treated as a flexible planning heuristic rather than a fixed formula.

Here’s how I’d think about it, step by step:

1. Protect the core first. Start with the maintenance line from Step One – the spend proven to hold current revenue. Fund that before anything else.
2. Fund the proven scalers. Give the channels that incrementality testing confirmed are generating net-new demand room to grow.
3. Ring-fence the test budget. Carve out a fixed pot for the three to five new-channel bets so they’re never raided when a proven channel has a good week.
4. Pay for measurement and creative. Treat data infrastructure and creative production as a line item, not an afterthought – they’re what make the other three buckets trustworthy.

Budget Bucket	% of Spend	Example GBP on a 10m Budget	Primary KPI
Protect the core (maintain proven performance)	60%	£6m	Blended CAC / ROAS held at baseline
Scale incremental winners	20%	£2m	Incremental ROAS, marginal CAC
New-channel tests (3–5 bets)	10%	£1m	Incremental lift vs holdout
Measurement, data & creative	10%	£1m	Test velocity, model confidence

The percentages should flex with your stage – an earlier brand still hunting for product-channel fit might push the test-and-learn slice closer to 20–30%, while a mature account tilts further toward protecting the core.

Designing an Incrementality Test That Proves Something

Naming incrementality testing (as I did in Step Two) is the easy part. Designing one that stands up is where it earns its keep. The workhorse method is a geo holdout: split your markets into two matched groups, keep spending as normal in one, and go dark – or hold to a baseline – in the other. The gap in outcomes is your incremental lift.

A few design rules of thumb:

• Duration: run it for at least two weeks, and longer for considered purchases where the sales cycle stretches results beyond the test window.
• Threshold: decide your minimum detectable effect up front. A commonly used bar is around a 15% lift – smaller effects need far more conversions to detect reliably, so low-volume channels are harder to read.
• PSA variant: for upper-funnel video and brand work, a PSA test (showing a public-service ad to the control group instead of nothing) isolates the impact of your message rather than mere ad presence.

If a channel clears your lift threshold with confidence, it graduates to the “scale incremental winners” bucket. If it doesn’t, it stays a test – or gets cut.

How Fast to Scale a Winner (Without Blowing Up Efficiency)

Finding a winning channel is only half the job. Scaling it too fast is one of the quickest ways to wreck efficiency, especially on Meta, where large budget changes reset the learning phase and send performance sideways. As a rule of thumb, step budgets up by around 15–20% every few days rather than doubling overnight, and let the algorithm restabilise between increases. Meta’s 2026 systems have tightened what counts as a “safe” change, so smaller, patient increments matter more than they used to.

Alongside the step-ups, hold a CAC or ROAS guardrail. If marginal CAC drifts past your payback threshold as you add spend, that’s the market telling you you’ve found the ceiling for now – pull back, consolidate, and redeploy the difference into the next test. Scaling is a series of controlled nudges, not a single lever. We go deeper on the mechanics in here’s how we scale Meta ads the right way and the smarter way to test and scale Meta ads.

The Measurement Stack Behind the Layers

The layered framework in Step Four only works if something feeds each layer. In practice that means three complementary methods – the “triangulation” approach that’s become best practice: platform and web analytics (GA4) for the daily click-based view, incrementality tests for causal read-outs, and marketing mix modelling (MMM) for the quarterly, top-down picture. For MMM there are now credible open-source options – Google’s Meridian (a Bayesian model that handles reach, frequency and dozens of geographies) and Meta’s Robyn (built on ridge regression) – so even the long-term layer is buildable without a six-figure vendor contract. No single method is the source of truth; you triangulate. For more on choosing what to measure, see why ROAS is misleading and what you should be tracking instead.

It’s Not About Spending More – It’s About Spending Smarter

At this level, it’s easy to get distracted by scale. But an 8-figure budget doesn’t give you permission to be lazy. It demands rigour.

The goal is not to spend the budget – it’s to invest it. Intelligently. Strategically. With a clear path to measurable returns. Because the brands that win in today’s landscape aren’t just outspending the competition – they’re outthinking them.

FAQ

How would you actually split an 8-figure marketing budget?

Start by protecting the spend that’s proven to hold current revenue, then fund the channels incrementality testing shows are genuinely driving net-new demand, ring-fence a fixed pot for three to five new-channel bets, and treat measurement and creative as their own line. On an illustrative £10m that might look like roughly 60% to protect the core, 20% to scale winners, 10% to tests and 10% to data and creative. Treat those numbers as a shape to adapt, not a rule to obey.

Does the 70/20/10 rule still apply at this budget?

The spirit of it does – roughly 70% on proven activity, 20% on emerging channels, 10% on genuine experiments. But it’s a planning heuristic, not a formula. Earlier-stage brands still finding product-channel fit usually need a bigger test-and-learn slice, while mature accounts tilt further toward protecting what already works.

How fast can you scale spend on a winning channel?

Slowly enough not to break it. On Meta especially, big budget jumps reset the learning phase, so step increases of around 15–20% every few days work better than doubling overnight. Hold a CAC or ROAS guardrail as you go: when marginal CAC pushes past your payback threshold, you’ve hit the current ceiling and should consolidate before pushing again.

What’s the best way to prove a channel is incremental?

A geo holdout is the workhorse: split your markets into two matched groups, keep spending in one and pause or hold to baseline in the other, and measure the gap. Run it for at least two weeks (longer for considered purchases), set a minimum detectable lift up front – around 15% is a common bar – and use a PSA variant for upper-funnel brand work to isolate your message from mere ad presence.

Which tools measure whether a large budget is working?

Triangulate rather than trust one number: GA4 and platform analytics for the daily click-based view, incrementality tests for causal read-outs, and marketing mix modelling for the quarterly top-down picture. Open-source MMM options like Google’s Meridian and Meta’s Robyn now make that long-term layer buildable without an enterprise vendor contract.