The first $100k of dedicated Meta spend at $5M+ revenue is not a beginner's question. It is the moment Meta becomes a strategic line in the P&L rather than a tactical channel under an agency's discretion. Five decisions decide what the channel produces for the next 18 months: budget split by intent, profitability gates, creative test rhythm, the signal you optimise to, and the 30-day decision matrix. This is the playbook for getting all five right.

There is a difference between a brand spending $100k a month on Meta because the platform has automated its way to that number, and a brand making a deliberate decision to deploy the first material chunk of profitable scale through Meta. This playbook is for the second one.

For founders running a $5M+ DTC brand, the first $100k of dedicated Meta spend is not a beginner's question. It is the moment the platform becomes a strategic line in the P&L rather than a tactical channel under an agency's discretion. The decisions you make in that window set the structural ceiling on what the channel can produce for the next 18 months.

Most brands skip the strategic decisions because their agency does not surface them and their finance team does not ask the right questions. The result is $100k of spend that hits the platform without the founder ever genuinely deciding how it should be split, what it should hit, when it should be killed, or what signal it should be optimising to. This is the version of that decision that compounds.

What the first $100k actually means at $5M+ revenue

For context, the first $100k of dedicated Meta spend at $5M+ revenue is a different proposition than at $1M.

At $1M, $100k is the budget for the year. Mistakes are existential. The job is to find product-market fit through ads, and the spending decision is mostly defensive.

At $5M+, $100k is a single month or a single quarter, depending on stage. Mistakes are recoverable. The job is to make Meta a structural growth lever rather than a maintenance channel. The spending decision is offensive, and the goal is to identify a configuration that you can deploy repeatedly with predictable contribution margin output.

The five decisions that matter at this stage, in priority order: budget split by intent, profitability gates, creative test rhythm, the signal you optimise to, and the 30-day decision matrix for what stays, what dies, and what gets rebuilt.

Decision 1: Budget split by intent

Most accounts default to roughly 70 percent prospecting, 20 percent retargeting, and 10 percent branded search. That is a reasonable starting point. It is also a decision worth questioning hard at $5M+, because the brands compounding at this tier are usually pushing the prospecting share higher, not lower.

The reason is the blended CAC distortion. Retargeting and branded search disproportionately convert customers who would have bought anyway. They flatter platform ROAS. They do not produce genuinely incremental revenue. At $5M+, where the brand has enough audience to retarget into, the temptation to over-invest in retargeting is high. The discipline is to resist it.

A defensible starting split for the first $100k looks closer to 75 percent prospecting, 15 percent retargeting, 10 percent branded search and lower-funnel demand capture. Push prospecting higher (80 to 85 percent) if you are confident in incrementality measurement. Push it lower (65 to 70 percent) only if you have a clear retargeting case based on incrementality tests, not platform ROAS.

The split that almost never makes sense at $5M+ is 50/50 prospecting and retargeting, which is what a lot of agency accounts default to. That split is optimising for platform ROAS, not for new customer acquisition. It will make the dashboard look better and the bank account look worse.

Decision 2: Profitability gates

The profitability gates are the floors and ceilings the campaign must respect before scale is justified. Three gates matter.

nCAC ceiling

The maximum new-customer cost the brand can sustain given its contribution margin and payback model. For most brands at $5M to $30M, this sits 70 to 90 percent below first-order AOV after contribution margin. If your contribution margin per first order is $40, your nCAC ceiling is around $40 (1.0 payback) for cautious accounts, or up to $60 (1.5 payback, assuming second-order margin recovers the difference) for confident accounts with strong retention data.

Contribution margin floor per acquired cohort

The minimum margin the brand needs from acquired customers across a defined window (usually 90 or 180 days). This is the gate that captures the LTV side of the equation. Pulled below this floor, even cheap acquisition does not justify the spend.

MER target across the period

The blended marketing efficiency target, calculated as total revenue divided by total marketing spend. For most brands at this tier, a healthy MER sits between 2.5x and 4.0x depending on category and stage. The gate is whether the first $100k of Meta spend leaves overall MER above that floor.

Each of these gates is a number, not a vibe. Founders who run Meta without explicit gates end up scaling whatever the agency thinks is working, which is usually whatever platform ROAS looks healthiest, which is usually the thing that is least incremental.

Decision 3: Creative test rhythm

The creative test rhythm is the cadence at which the brand introduces, evaluates, and iterates new creative. At $5M+, the rhythm needs to be deliberate enough to produce signal but not so frenetic that it confuses the algorithm.

A defensible starting rhythm: 14 new ad variants every two weeks, organised around four creative angles, on a six-week test cycle. Inside each batch of 14, three or four are intentionally divergent (different format, different angle, different hook) and the rest are iterations on the previous winners. Each ad runs long enough to gather statistically meaningful data (usually 5 to 7 days minimum at meaningful spend), and the kill rules are explicit before launch, not after.

The creative testing layer is also where Entity ID diversity matters. We wrote about this in the Andromeda Q2 update. Shipping 14 new ads that share an Entity ID with last batch's winners is not a real test. The algorithm reads them as the same creative. Real testing requires visual variance at the Entity ID level, not just copy variance.

Decision 4: Signals you optimise to (not ROAS first)

Most accounts default to optimising campaigns for purchase events and reading the resulting ROAS as the primary signal. At $5M+, this is the moment to invert the order.

Optimise for the deepest event you can support with volume (purchase, ideally). But read the signal in this order: nCAC, contribution margin per acquired customer at 30 days, creative-level engagement signals (hook rate, hold rate, unique outbound CTR), and platform ROAS as the supporting tactical view.

Reading nCAC first forces the team to ask the right question of every spend decision: is this dollar producing a genuinely new customer who will pay back, or is it converting existing demand at a flattering ROAS? Reading platform ROAS first lets the report tell you what you want to hear.

The signals at the creative level matter because they predict where the algorithm will allocate future spend. Hook rate of 30 percent+, hold rate of 12 percent+, unique outbound CTR of 1.5 percent+ are the rough benchmarks for a creative that earns continued investment. Below those thresholds, the algorithm will quietly de-prioritise the ad even if the headline ROAS looks acceptable, which means the creative is on borrowed time.

Decision 5: The 30-day decision matrix

Most accounts run on a "wait and see" cadence that means active campaigns rarely get killed and new ones rarely get the budget to learn properly. At $5M+, the first $100k deserves a decision matrix that forces clear calls.

At day 14: which campaigns are pacing against nCAC ceilings? Anything 50 percent over the ceiling at day 14 gets killed, not iterated. Anything within 20 percent of the ceiling continues. Anything below the ceiling gets a budget lift inside the existing structure.

At day 21: creative-level performance review. Underperforming ads inside winning campaigns get cut, not iterated. Winning creative gets prioritised inside the production pipeline for variants in the next batch.

At day 30: structural review. Are the campaigns still aligned with the budget split decision from day 1? Has retargeting share quietly drifted up to 30 percent because platform ROAS looks healthy? Has the team added campaigns without removing any? Most accounts drift away from the original strategic intent within a month. The 30-day review is the discipline that resets to plan.

Where founders most commonly leak the spend

Five patterns we see repeatedly when auditing accounts that have deployed material Meta spend without producing the expected business outcome.

Retargeting and branded search creep. The strategic intent at day 1 was 75 percent prospecting. By day 60 it is 55 percent. Nobody decided this; it happened through small budget reallocations to the campaigns showing the strongest platform ROAS. The result is overall spend looking healthy on the report and overall nCAC quietly climbing in the background.

Optimising to events that do not match the business model. Add-to-cart optimisation produces audiences that look like add-to-cart audiences, not purchase audiences. Most brands at $5M+ have the volume to optimise at the purchase event. Failing to do so is a structural mistake.

Killing winners too early. A creative running 5 days at modest spend has not learned. Killing it at day 6 because the headline ROAS looks soft prevents the algorithm from finding the audience it was built to find.

Letting losers run too long. A creative running 21 days at meaningful spend, with hook rates below 25 percent and CTRs below 1 percent, is not going to recover. Most accounts let it limp on because nobody made the kill call. That spend is acquisition cost the brand should never have paid.

Treating the catalogue as set-and-forget. As Ben pointed out in his Fastlane piece, the catalogue does enormous work in Meta delivery and almost nobody optimises it. Brands shipping new creative every fortnight while running on a stale catalogue are leaking exactly where they cannot see it.

What the first $100k should produce

If the playbook is run with discipline, the first $100k of Meta spend at $5M+ should produce three outputs.

A clear nCAC against contribution margin. You should know, by day 60, what it actually costs to acquire a new customer who pays back inside your gates. That number is the input to every subsequent scaling decision.

A winning creative library, not a winning ad. The goal is two or three creative angles you can iterate on for the next six months, supported by 8 to 12 Entity IDs that the algorithm has learned. A single hero ad is not a scalable output. A library is.

A repeatable structural setup. Budget split, profitability gates, creative test rhythm, and signal hierarchy that you can deploy again with the next $100k, and the $100k after that, with predictable outputs.

If the first $100k produces a flashy ROAS report and no real answers to those three questions, the spend has produced activity rather than infrastructure.

Where to go next

Webtopia runs the first $100k of Meta spend for $5M+ DTC brands as a deliberate playbook, not an open-ended scale push. We measure against nCAC, contribution margin per acquired cohort, and MER rather than platform ROAS. If you want a view on what your first $100k should look like, book a call and we will walk through your budget split, gates, and rhythm with you.

For the broader Full Picture context, the leaky bucket audit is the diagnostic that connects acquisition spend to retention quality, which is the lever that decides whether the $100k pays back.

Frequently asked questions

How much should I spend on Meta ads as a DTC brand at $5M+?

The first $100k of dedicated Meta spend at $5M+ revenue typically covers a single month or quarter depending on stage. The amount matters less than the structural setup: budget split (around 75 percent prospecting, 15 percent retargeting, 10 percent branded search), explicit profitability gates (nCAC ceiling, contribution margin floor, MER target), and a deliberate 30-day decision matrix.

What is a good budget split for Meta ads in 2026?

For DTC brands at $5M+, a defensible starting split is 75 percent prospecting, 15 percent retargeting, 10 percent branded search and lower-funnel demand capture. Push prospecting higher if you have strong incrementality measurement. The split that rarely makes sense is 50/50 prospecting and retargeting, which optimises for platform ROAS rather than genuinely new customer acquisition.

What profitability gates should I set for Meta ads?

Three gates. An nCAC ceiling tied to contribution margin and payback model (usually 70 to 90 percent of first-order contribution margin for cautious accounts). A contribution margin floor per acquired cohort over a 90 or 180 day window. And a MER target across the full marketing engine, typically 2.5x to 4.0x for healthy brands at this tier. Each is a number, not a vibe.

How often should I test new creative on Meta?

A defensible rhythm at $5M+ is 14 new ad variants every two weeks, organised around four creative angles, on a six-week test cycle. Each ad runs at least 5 to 7 days at meaningful spend to gather statistical signal. Kill rules should be explicit before launch. Visual variance at the Entity ID level matters more than copy variance, because Meta's Andromeda system reads visually similar ads as the same creative.