How Bach AI Optimizes Frequency Caps Per Audience Automatically

Most Indian D2C accounts run a single frequency cap across every campaign. 2.5 impressions per user per week. Sometimes 3. Sometimes nothing at all. It's the marketing equivalent of using one shoe size for the whole family.

Frequency tolerance varies wildly by audience. Your retargeting pool can absorb 8-12 impressions before fatigue. Your cold prospecting audience starts decaying at 4. Your lookalike sits somewhere in between. A single cap either suffocates the cold audiences or burns out the warm ones.

Why Frequency Caps Are Invisible Money Leaks

Frequency fatigue doesn't show up as a red flag on your dashboard. It shows up as gradually rising CPM, gradually falling CTR, and stubbornly flat ROAS. Three weeks in, you assume creatives are tired. You refresh them. Performance returns for a week, then dips again. You're chasing the wrong variable.

Real-world data from Indian D2C accounts in our system:

• Cold prospecting audiences show CTR decay above frequency 4.2 (median across apparel and beauty).

• Lookalike audiences tolerate up to frequency 6.5 before CPM spikes.

• Retargeting from cart abandonment can run frequency 10+ profitably if the offer rotates.

• Existing customer audiences show fatigue earlier — frequency 3.5 for engagement campaigns.

One cap can't honor all four curves. See [how to detect ad fatigue and stop it](https://www.wittelsbach.ai/post/how-to-detect-ad-fatigue-and-stop-it-before-it-costs-you) for the diagnostic side of this.

How Bach AI Calculates Per-Audience Optimal Frequency

Three steps, fully automated:

1. Build the per-audience decay curve

Bach AI plots CTR, CPC, and conversion rate against impression frequency for each audience over the last 90 days. The curve always slopes downward eventually. The question is where the slope steepens.

2. Identify the inflection point

The 'optimal cap' is the frequency at which marginal ROAS drops below your account's blended baseline. Below that point, more frequency is profitable. Above it, you're paying to fatigue your audience.

3. Apply with safety buffers

Bach AI recommends a cap 15% below the inflection point — a buffer for creative rotation and seasonal volatility. Frequency caps are set per adset, not per campaign, because audience composition varies.

The Rolling 7-Day Adjustment Logic

Static caps get stale fast. New creatives push the inflection point higher. Saturation drops it. Bach AI re-evaluates every Monday using the previous 7 days of impression and engagement data.

• Creative refresh detected → cap tolerated higher → Bach AI loosens the cap by 10-20%.

• Audience saturation rising → conversion rate dropping at lower frequencies → Bach AI tightens the cap.

• Seasonal demand surge → audiences less price-sensitive → Bach AI allows higher exposure before fatigue.

Every adjustment shows up in your Bach AI feed with the reason and the predicted ROAS delta. You approve in one click or let auto-execute handle it.

What This Looks Like Inside the Product

Open any campaign in Bach AI and you see a Frequency card per adset:

Prospecting — Beauty Lovers 25-44 — Current frequency: 5.8 (last 7d). Recommended cap: 4.5. Reason: CTR drops 31% above frequency 5.0. Predicted impact: +₹18,400/month ROAS.

Click 'Apply' and Bach AI pushes the new cap into Meta via the API. Click 'Show working' and you get the raw decay curve.

Common Frequency Cap Mistakes Indian D2C Brands Make

1. One cap across the entire account — guarantees over- or under-exposure for most audiences.

2. Setting caps at the campaign level — ignores the adset-level variation in audience composition.

3. Caps too tight for retargeting — abandoning cart pools can absorb high frequency profitably.

4. Never adjusting after launch — the right cap on day 1 isn't the right cap on day 30.

5. No cap at all on broad audiences — Meta's delivery system will gladly burn ₹50K showing your ad to the same 800 users.

How Wittelsbach AI Handles Frequency at Scale

Bach AI runs per-adset frequency analysis on every account, every week, automatically. No spreadsheets, no manual decay curves. Adjustments push to Meta with one approval or run on auto-pilot if you enable it. Bach AI is live at [app.wittelsbach.ai](https://app.wittelsbach.ai). Two clicks to connect Meta.

Frequently Asked Questions

What's a good Meta frequency cap for Indian D2C prospecting campaigns?

Across our customer base, 4.0-4.5 weekly frequency works as a starting point for cold prospecting in apparel, beauty, and home. Jewelry tolerates higher (5-6) because consideration cycles are longer. But the right number is account-specific — Bach AI calibrates yours from your actual decay curve.

Should I set frequency caps on retargeting campaigns at all?

Yes, but loose ones. Retargeting audiences expect repeat exposure. Caps of 10-15 weekly are common and profitable, especially with rotating creative and offer changes. Without any cap, you risk burning out warm prospects with stale creative. Read more on [retargeting funnels for D2C](https://www.wittelsbach.ai/post/retargeting-funnels-for-d2c-beyond-abandoned-cart-sequences).

How does frequency cap interact with creative refresh cycles?

Inversely. Fresh creative pushes the fatigue point higher — you can sustain higher frequency before decay kicks in. Bach AI tracks creative age alongside frequency, so when you launch a new creative pack, caps automatically loosen. When creatives age past 21 days, caps tighten.

Does Meta still respect frequency caps in CBO campaigns?

Yes, but enforcement is looser than ABO. Meta optimizes for cost, not frequency. If you need strict caps, ABO gives you tighter control. Bach AI flags accounts where CBO is undermining frequency goals and recommends a [CBO vs ABO](https://www.wittelsbach.ai/post/cbo-vs-abo-in-meta-ads-which-budget-strategy-wins-for-d2c-in-2026) reconfiguration when it helps.

What happens to my CPM when I lower the frequency cap?

Two things, usually in this order. Reach expands (Meta has to find new users), so CPM can rise 5-15% short-term. Then CTR climbs (less audience fatigue), so CPC and CPA improve. Net ROAS typically improves within 7-10 days. Bach AI predicts the trajectory before you apply.