Search Match to Manual: A Safe Way to Gain Control in Apple Search Ads (Without Tanking Delivery)

Apple Search Ads can feel like two different games: one where you let Apple do the matching (Search Match), and another where you micro-manage keyword targeting (exact/broad). The mistake most indie teams make is switching too aggressively—then they wonder why delivery drops or CPI spikes.

This post gives you a “safe control” workflow: how to move from Search Match to manual keywords using search-term evidence, while keeping auction participation steady.

The real problem with Search Match isn’t control—it’s attribution clarity

Search Match is useful because it can find queries you didn’t explicitly buy. But because it’s automatic matching, you can’t directly reason about performance at the keyword level in the way you can with exact/broad.

So your goal isn’t to “replace” Search Match instantly. Your goal is to:

Identify which search terms are actually producing installs (and ideally profitable revenue)
Recreate that intent with manual keywords
Keep enough inventory in the auction while you ramp manual spend

When you should start the Search Match → Manual transition

You don’t need to wait forever, but you also shouldn’t act on thin data. Start once your Search Match ad group (or campaign) has enough signal to separate “noise” from “intent.” Concretely, look for:

A stable spend pattern over at least ~7 days (or enough impressions for your top terms to show up)
Multiple recurring search terms in the Search Terms report, not just one-off queries
Search terms that produce taps, then installs (not just clicks)

If you don’t have enough data, the transition will turn into guesswork—exactly what we’re avoiding.

Step 1: Export your Search Terms and score them by intent

In Apple Ads, open the Search Terms report for your Search Match-based campaign/ad group. For each search term that appears, record these columns (or the closest equivalents you see):

Impressions
Taps
TTR (taps / impressions)
Installs
Install conversion rate (installs / taps)
CPT
CPI (or CPA)

Then split terms into three buckets:

Bucket A — “Keep / Expand”

Install conversion rate is meaningfully above your account baseline and
CPI is within the range you can sustain for payback

Even if you don’t have ROAS by term (you typically won’t), CPI + conversion rate is still enough to decide what intent you want to own.

Bucket B — “Monitor”

Decent taps/TTR but weak installs (or installs but inconsistent CPI)

These are often near-miss queries. You can keep them on Search Match for a while while you tighten other pieces.

Bucket C — “Exclude / Deprioritize”

Lots of taps but near-zero installs
Or impressions/taps without any install signal after a reasonable sample

This bucket is where negative keywords start paying off.

Step 2: Convert top search terms into manual keywords (with a safety rule)

Now take Bucket A terms and convert them into manual keywords. Here’s the safe part:

Use exact first for the most profitable intent

For your top 5–15 Bucket A terms (depending on volume), create an ad group for exact match keywords.

Add broad only when you have at least some install history

Broad can generate volume fast, but it can also reintroduce the “automatic matching chaos” you’re trying to reduce.

So for broad, either:

Start with a narrower broad list (fewer terms), or
Use broad only after you’ve confirmed your exact match terms continue to convert

Safety rule: don’t translate everything at once

If you convert 100% of your Search Match intent into manual keywords immediately, you can starve the auction while the new ad group “learns.” Keep Search Match running during the transition.

Step 3: Stage the campaign structure so auctions don’t collide

A common failure mode: you add manual keywords, but they end up competing with Search Match in a way that’s hard to interpret.

Your structure should make intent separation obvious.

Recommended staging:

Campaign 1 (existing): Search Match ad group stays on
Campaign 2 (new): Exact manual ad group contains the Bucket A exact keywords
Campaign 3 (optional): Broad manual ad group contains a smaller broad set after exact stabilizes

Why separate campaigns? You keep country/region rules clean and avoid confusing bid interactions. Also, it’s easier to compare performance week-over-week.

Step 4: Use bid caps that protect CPI while delivery ramps

You’re not trying to “win” the auction on day one—you’re trying to participate without exploding CPI.

For each new manual ad group:

Start the max CPT bid at or slightly below what your Search Match is effectively paying for similar outcomes (use your Search Terms metrics as your guide)
If you see high taps but poor installs, lower bids rather than immediately expanding targeting

Important: CPT bidding is auction-based. Two keywords can have the same intent but different auction pressure. So treat bid levels as hypotheses, validated by taps → installs.

Step 5: Add negatives to stop manual broad from re-learning your mistakes

Once your manual campaigns start delivering, build negatives using your Search Terms report.

In particular:

Add negatives for Bucket C search terms (or near-identical phrasing)
If broad is generating a lot of taps without installs, add negatives sooner

You don’t need a massive negative list. You need the top 20–50 worst offenders—whatever shows up repeatedly.

Step 6: Let the transition run on a schedule (not emotion)

Here’s the operational cadence that prevents churn:

Day 0–2: implement the manual keyword campaign(s), keep Search Match live
Day 3–7: monitor taps, install conversion rate, and CPI
End of week 1: decide which manual keywords deserve more budget via higher bids (or simply add more terms)

If you change bids every day, you won’t learn anything. You’ll just confuse auction pressure with your actions.

How to decide when it’s time to turn down Search Match

You shouldn’t pause Search Match until your manual keywords can “own” the intent it was finding.

A practical stop condition:

Your manual exact keywords show stable install conversion rate (over several days)
Your manual CPI is not drifting upward beyond what you consider acceptable
You have enough spend from manual campaigns that Search Match isn’t the only source of installs

Then you can reduce Search Match coverage by lowering its max CPT bid (or gradually pausing it if volume is replaced).

If you cut Search Match early, you’ll likely experience one of these:

Delivery drop (less inventory for your best intents)
Higher CPI (manual bids are still ramping)
Unclear learning (you can’t tell what stopped working)

Common pitfalls (and how to avoid them)

Pitfall 1: Converting search terms into keywords without controlling scope

If you turn a broad search term list into broad match keywords all at once, you often get the same “automatic matching” behavior again.

Fix: exact first, narrow broad second.

Pitfall 2: Translating one winning term into a generic keyword

Example: “game for kids ages 5” is not “kids game.” Different intent, different auction.

Fix: start with the phrase-level terms that actually show up in your Search Terms report.

Pitfall 3: Ignoring taps → installs

Clicks alone can be misleading. Apple Ads can deliver taps even when install conversion is weak.

Fix: prioritize install conversion rate and CPI when evaluating candidates.

Where AdsBuddy can help (lightly)

If you want a fast way to identify which Search Terms to promote into exact/broad (and which to negative), tools like AdsBuddy can read your Apple Search Ads performance and revenue signals and propose the next small changes to approve—without you manually stitching reports together.

Takeaway

A safe Search Match → Manual transition is about staging, not replacement. Keep Search Match running while you recreate its best-performing intent with exact keywords first, add broad carefully, apply negatives from your search terms, and only dial down Search Match once manual conversions are stable.

If you do it this way, you gain control without sacrificing delivery—and your CPI stops being a mystery.