Negative keywords in Apple Search Ads: the cheapest ROAS win

Negative keywords feel “minor,” but in Apple Search Ads they’re often the fastest way to improve ROAS: you’re not trying to win more taps—you’re blocking the taps you already know won’t turn into installs (or purchases). When spend drops while revenue from qualified users stays the same, ROAS climbs.

Why negatives can be the cheapest ROAS win

Apple Search Ads runs a cost-per-tap (CPT) auction. Your ad shows for search queries that match your keywords (including your Search Match / Discovery-style matching). If Apple sends you traffic that’s clearly off-intent, you pay for taps anyway.

A negative keyword flips the equation from “optimize what we attract” to “remove what we shouldn’t.” The key benefit is mechanical:

• Spend is proportional to taps (at whatever CPT you’re paying).
• Revenue is proportional to the users who convert (install → purchase/other event).
• If a query produces taps with near-zero conversion, negative keywords reduce spend without materially reducing the installs you actually care about.

Think of it like tightening your funnel at the top. You don’t need a better product page yet—you need fewer irrelevant searchers hitting it.

Where negative keywords matter in Apple Search Ads

In Apple Search Ads, negative keywords are a Search Results–level guardrail. Practically, they help you prevent your app ads from showing for specific search terms within the scope where you’ve set the targeting.

A few practical constraints to keep in mind:

• They’re most valuable for Search Results campaigns (where most indie spend starts).
• One campaign targets one country/region, so be mindful: negatives are not “global across the world.” If you run multiple countries, you’ll manage negatives per campaign/country.
• Keywords live inside ad groups. Your negatives are also managed at that level—so if you split keyword sets into separate ad groups (common for control), you may need to add negatives in more than one place.

The playbook: add negatives using data, not vibes

Here’s the workflow I’d use if I were trying to improve ROAS quickly without touching bids.

Step 1: Pull your query/search term report

In Apple Search Ads, look at the Search Results search term data (the report that shows what users searched for when your ads were eligible to show).

Make sure you’re looking at:

• the correct country/region
• the correct campaign
• the correct time window (use enough days to avoid one-off noise)

If you only have partial data, start with what you trust: usually, the worst queries by spend are the most reliable.

Step 2: Score queries by “waste,” not just by taps

You’re trying to find searches that generate cost with poor conversion.

Use these metrics in combination:

• Taps and CPT: tells you whether you’re paying real money for that query.
• TTR (taps/impressions): if TTR is low, the query likely isn’t aligned with your ad’s promise (low relevance).
• Conversion rate (installs/taps): this is the fastest signal of intent mismatch.
• CPI/CPA (and ultimately ROAS if you track purchases): tells you whether the install quality is poor.

A practical way to prioritize without inventing benchmarks:

• Sort queries by total spend.
• From the top spenders, focus on queries where conversion rate is extremely low and/or where downstream purchase signals show poor ROAS.

Step 3: Turn “bad queries” into precise negative keywords

When you pick negatives, be conservative and literal.

Good negative candidates:

• Clear intent mismatch (e.g., your app is a paid habit tracker, but the search term is “free” + a totally different use case).
• Competitor/service confusion (depending on your strategy—see the warning below).
• Content type mismatch (e.g., users searching for “PDF,” “wallpaper,” “lyrics,” etc., when your app isn’t that).
• Device/format confusion (“iPad app only” when you’re marketing a universal app, or “iPhone 15 pro” style queries for something that doesn’t relate).

Avoid negative candidates when:

• The term is closely tied to your brand name or your core product terms.
• The term is a common synonym for your value proposition. Overblocking can quietly erase your best traffic.

Tip: add the negative using the exact phrasing you see in the search term report. Don’t try to get clever with stemming or “close enough” guesses.

Step 4: Add negatives in small batches (and keep an eye on volume)

Treat negatives like guardrails, not a rewrite of your strategy.

A safe batch approach:

• Start with high-confidence negatives (queries with meaningful spend + very poor conversion).
• Add a small set (think “top offenders,” not “everything that looks related”).
• Wait for data to accumulate, then reassess.

What “good signs” look like after adding negatives:

• Spend drops for the ad group/campaign.
• Taps from irrelevant queries disappear.
• ROAS improves (or at least CPI/CPA improves) without a dramatic crash in qualified install volume.

If spend drops but installs also collapse, you likely blocked queries that were converting—time to remove the least certain negatives.

Common mistakes that erase the benefit

Mistake 1: Negatives too broad

If you negative a term that’s an important part of your market (even if it’s not your best converting segment), you can choke off the right users.

Fix: begin with the exact query strings you’re blocking, and expand only after confirming the effect.

Mistake 2: One country uses negatives that were learned in another

Because each campaign is tied to a country/region, your “bad” searches in one market may be normal in another.

Fix: learn and apply negatives per campaign.

Mistake 3: Forgetting that ad group structure matters

If you have the same keyword theme in multiple ad groups, the negative you add to one ad group may not protect the others.

Fix: track where the search terms are coming from in the report and apply negatives to the right ad group(s).

A concrete example (illustrative)

Say your search term report shows three queries for a given ad group:

• Query A: high spend, TTR moderate, conversion rate near zero
• Query B: medium spend, conversion rate decent
• Query C: low spend, conversion uncertain

Your fastest ROAS win usually comes from Query A: add it as a negative first.

After the change, you’d expect:

• fewer taps from that irrelevant search pattern
• lower spend
• roughly stable installs from the still-eligible good queries

Then you can consider Query C once you have more data (because low spend terms can be misleading early).

How often to refresh negatives

Negatives are not a one-time chore. A simple cadence that works for indie teams:

• Review search terms on a weekly or twice-weekly basis for the first month after you start or re-structure campaigns.
• After things stabilize, move to bi-weekly.
• After any big change to targeting (new keywords/ad groups) or product page updates, do a fresh look.

The goal isn’t perfection—it’s continuous removal of obvious waste.

Quick checklist before you add negatives

• You’re looking at the Search Results search term data.
• You’ve filtered to the correct country/campaign.
• You prioritized queries by spend + poor conversion / poor purchase outcome.
• Your negatives are exactly the terms you saw (no guesswork).
• You’re adding in small batches so you can detect overblocking.

Closing takeaway

Negative keywords are “cheapest” because they require minimal operational change: no new creative, no bid rewrites, no redesign of your product page. You’re simply preventing irrelevant searches from spending your budget.

If you want to make this even faster, a tool like AdsBuddy can read your Apple Search Ads performance alongside your revenue signals and surface the negative keywords that are most likely to improve ROAS first—so you spend your time approving changes that are actually worth it.