Paid media
Model how CPA inflates as you double and triple spend, budgets never scale linearly.
- Category
- Paid media
- Signup
- None
- Your data
- Stays in your browser
- Also in
- Claude, via MCP
How much does CPA rise when you scale ad spend?
Model a fixed rise each time spend doubles. CPA at a spend multiple = current CPA × (1 + rise per doubling)^log₂(multiple). At ₹1,200 CPA and an 18% rise per doubling, CPA is ₹1,416 at twice the spend and about ₹1,560 at three times, 30% above today.
Tripling spend doesn’t triple customers, at 18% decay your CAC climbs to $156 (30% higher). Check that number against LTV before you scale.
Decay compounds per doubling of spend, so returns bend as you exhaust your best audiences and placements first.
About this calculator
"Just double the budget" is the most common way a profitable channel gets wrecked, because CPA rarely scales linearly with spend. This calculator applies a decay rate you set to your current CPA to project blended acquisition cost at 2x and 3x spend, so you can see, before you commit budget, whether scaling breaks your unit economics.
How to use it
- Enter your current monthly spend and the CPA (or CAC) that spend is producing today.
- Enter your CPA inflation rate, how much CPA typically rises each time you double budget in this channel. Estimate it from your own history if you have it.
- Read CAC at 2x and 3x spend, plus the resulting customer volume and the CAC premium versus today.
- Compare 3x CAC against your LTV or max-CAC ceiling before committing to a scale-up.
Methodology
The model compounds inflation per doubling of spend: CPA at a given multiple is current CPA × (1 + decay rate) raised to the power of log₂(multiple). Tripling spend is treated as log₂(3) ≈ 1.585 doublings, so the 3x CPA sits between the 2x and a hypothetical 4x figure, not simply 1.5x the 2x number.
Customer volume at each spend level is that level's spend ÷ that level's CPA, showing that doubling spend produces less than double the customers once CPA rises.
The CAC premium at 3x is the percentage increase in CPA versus your current baseline, isolating how much more expensive each additional customer becomes as you climb.
This assumes decay compounds smoothly and uniformly, real auctions often decay in steps (a specific audience segment exhausts, then costs jump) rather than a smooth curve, so treat the projection as directional, and validate the decay rate against your own spend-history data whenever you have it.
CPA and customers at higher spend
CPA at multiple m = Current CPA × (1 + Rise per doubling)^log₂(m) Customers at m = Current spend × m ÷ CPA at m CAC premium at 3× = CPA at 3× ÷ Current CPA − 1 Marginal CPA from 1× to 2× = Extra spend ÷ Extra customers
- Rise per doubling
- How much CPA increases each time spend doubles, from your own history or a test.
- log₂(m)
- Number of doublings: 1 at 2×, about 1.585 at 3×, 2 at 4×.
- Marginal CPA
- What the extra customers cost, which is higher than the average CPA the tool shows.
The CPA at each level is the average across all spend, not the cost of the extra customers. The extra customers cost more: in the example, average CPA at 2× is ₹1,416, but the additional customers cost about ₹1,727 each. Judge scaling on marginal CPA against your maximum, not on the average.
Worked example: doubling and tripling a ₹5 lakh budget
A D2C brand spends ₹5,00,000 a month on Meta at a ₹1,200 CPA. A past budget increase suggests CPA rises about 18% each time spend doubles.
- Customers now = ₹5,00,000 ÷ ₹1,200 = 417.
- CPA at 2× = ₹1,200 × 1.18 = ₹1,416; customers = ₹10,00,000 ÷ ₹1,416 = 706.
- CPA at 3× = ₹1,200 × 1.18^1.585 = ₹1,560; customers = ₹15,00,000 ÷ ₹1,560 = 962.
- CAC premium at 3× = ₹1,560 ÷ ₹1,200 − 1 = 30%.
- Marginal CPA from 1× to 2× = ₹5,00,000 ÷ (706 − 417) = ₹1,727 per extra customer.
Doubling spend buys 69% more customers, not 100%, and each extra one costs ₹1,727. If the maximum CPA is ₹1,600, doubling is not worth it even though the average CPA stays under it. Set the ceiling with the performance marketing margin protector and watch the trend with the channel saturation estimator.
How do I estimate my rise per doubling?
Use your own spend history or a deliberate test:
- Find two periods with similar creative and season but different spend levels.
- Calculate how many doublings separate them (log₂ of the spend ratio) and the CPA change.
- Rise per doubling = (CPA ratio)^(1 ÷ doublings) − 1.
If you have no history, run a two-week test at higher spend on one campaign. The rate varies widely by audience size and channel, so a figure from another business is a poor substitute.
Scaling ad spend in India vs the US
Large Indian audiences often let consumer brands scale with a smaller rise per doubling, until a city, language or interest segment is exhausted. Expanding to new regions or languages can reset the curve. B2B audiences are smaller, so the rise is usually steeper.
US audiences are smaller and more contested in many categories, so CPA often rises faster as spend grows, and seasonal peaks add to it. Test scaling in steps rather than doubling at once, and check frequency as you go with the ad frequency risk checker.
Common ad scaling mistakes
- Judging scale on average CPA instead of marginal CPA.
- Assuming CPA stays flat because it did at small budgets.
- Estimating the rise from a period that also changed creative or season.
- Scaling one ad set far beyond its audience instead of adding audiences.
- Comparing CPA across channels without allowing for each channel's own curve.
FAQ
Auctions have a finite pool of your best-fit, lowest-cost impressions. As you spend more, the algorithm bids on progressively lower-propensity audience members to hit delivery goals, which raises the blended cost per acquisition, not the cost of your original audience.
Look at your own history: find two periods where spend roughly doubled and compare the CPA in each. The percentage CPA increase between those periods is a reasonable starting decay rate for this calculator.
Blended CAC at a spend level is the average cost across all customers acquired at that level, including the cheap ones from your original spend. Marginal CAC, the cost of the next customer specifically, rises even faster than the blended figure shown here.
Scale in smaller increments and re-measure decay at each step rather than jumping straight to 3x, expand into adjacent audiences or channels to add fresh, lower-cost inventory instead of over-saturating the current one, or accept a lower scale target that keeps CAC within your margin.
Marginal CPA is the cost of the extra customers you gain by spending more: extra spend divided by extra customers. It is always at or above average CPA when costs rise with scale.
CPA here is the paid media cost per acquisition. CAC usually includes all sales and marketing costs. The scaling curve applies to the media part.
Sometimes, when more data helps bidding or a larger budget reaches cheaper placements. It rarely lasts across several doublings.
Numbers looking off?
If the model says one thing and the pipeline says another, that gap is usually where the revenue is. A 30-minute call finds it.
It starts with a 30-minute call. Pick a time below.