Promoting Winning Search Terms to Exact-Match Keywords in Apple Search Ads

If you’re running Apple Search Ads (ASA) yourself, you already know the fastest path to better results is to stop treating every tap the same. The highest leverage move is usually: take search terms that are clearly working and “lock them in” as exact-match keywords. That reduces wasted discovery traffic while keeping the queries that actually convert.

Below is a practical way to promote winning search terms to exact-match keywords—without accidentally promoting one-offs, seasonal spikes, or irrelevant lookalikes.

Why promote to exact match?

On the Search Results placement, ASA can match your keywords using different match types. Exact match focuses on the term you chose (with Apple’s matching rules), while Discovery/Search Match (automatic matching) can reach related queries that you didn’t explicitly write.

Exact match matters because:

You reduce ambiguity. Discovery match may bring “close enough” queries. Exact match keeps the intent tighter.
You get cleaner optimization signals. When the query is closer to what you bid on, it’s easier to attribute performance to the keyword itself.
You can scale intentionally. You choose which terms deserve higher bids and which should be capped.

The goal isn’t to turn off discovery entirely. It’s to graduate winners into exact so you can scale them with control.

Step 1: Define what “winning” means (use install chain metrics)

ASA doesn’t give you revenue per keyword the same way you might expect; revenue comes from the install→purchase chain using Apple’s attribution token (often mapped to revenue by tools like RevenueCat). So the most useful “winning” definition is based on metrics you can observe and that correlate with downstream purchases.

Start with a shortlist of search terms that meet conditions like:

Conversion quality is solid: installs per tap (conversion rate) is meaningfully above your baseline.
Efficiency is good: your CPI/CPA is at least stable or improving.
Profitability is not broken: if you have ROAS/revenue mapping, confirm the term isn’t only buying cheap installs that don’t monetize.

A simple rule indie devs can use:

Treat a term as “candidate winner” if it has enough volume to trust the signal (avoid tiny sample noise).
Only promote if CPI/CPA is acceptable and conversion rate isn’t an accident.

If you only look at taps or impressions, you’ll promote traffic that’s interesting but doesn’t actually produce sustainable installs.

Step 2: Pull the Search Terms report and sort by performance

In ASA, use the search terms / performance view for your Search Results-targeting keywords/ad groups.

What to extract for each search term:

The search term text (the actual query users entered)
Impressions and taps (to gauge volume)
TTR (taps ÷ impressions)
Installs and conversion rate (installs ÷ taps)
CPT (your max bid constraints matter)
CPI/CPA (as available)
Any mapping you use to estimate revenue/ROAS

Then sort and filter so you’re reviewing the top few “highest quality” terms, not just highest taps.

Step 3: Check the context—are winners coming from the right place?

Before you promote anything, verify these practical details:

Placement: Make sure these terms are actually driving results on Search Results (since that’s where most discovery happens).
Country/region: Remember: one campaign targets one country/region. Don’t mix winner terms across regions in your head—performance can change sharply.
Ad group + match type source: If the winning term came only from a single discovery match setup, it may be broader than you think.

This is also where you protect yourself from “promoting an accident.” If the term only looks great because it had unusually high conversion on one day or in one campaign period, pause and wait for stability.

Step 4: Create an exact-match keyword ad group for promoted terms

Don’t drop promoted exact keywords into a messy existing ad group with mixed intent.

Instead, create a dedicated structure like:

Campaign: same country/region
Ad group: “Exact Winners (Promoted)”
Targeting: exact-match keywords only (for terms you’ve verified)

Why separate? Because it keeps performance signals clean and makes future bid adjustments safer.

Include one keyword per term (initially)

For your first iteration, add the exact-match keyword for the promoted search term as a separate keyword entry.

This gives you:

clearer reporting (you can see which exact terms keep winning)
fewer unintended interactions (especially if you later adjust bids)

Step 5: Set the bid for exact match (start from your current reality)

Exact match keywords usually require a different bid strategy than discovery match. There’s no universal bid number you should copy, so use your existing data.

A safe starting approach:

Note the CPT you’re effectively paying when the search term is generating installs.
Start the exact-match keyword at a similar level, then allow a short test window.
If your exact-match keyword gets fewer impressions than discovery matched traffic, you may need a small bid adjustment upward to regain share.

If you’re constrained by budget, prioritize the winners that have the best downstream quality (conversion rate and CPI/CPA, and ROAS if you map it).

Don’t overbid to “win the auction” on day one

Exact match can ramp quickly if you bid too aggressively—sometimes you’ll get a lot of impressions with mediocre conversion. Let the keyword prove itself again under the tighter intent.

Step 6: Keep discovery running, but stop it from re-paying the same bill

Here’s the common mistake: promoting exact keywords but leaving discovery match to keep buying the same queries at the same cost.

To avoid paying twice for the same intent, do one of these (depending on your setup):

Lower bids on the discovery ad group(s) that were bringing those winning terms, or
Add negative keywords to prevent the discovery ad group from matching the exact winning terms (where supported and appropriate)

The guiding principle: discovery should be responsible for finding new winners, not repeatedly buying known converters.

Step 7: Validate within a controlled time window

After you add exact-match keywords, watch:

TTR: Does the exact keyword get reasonable engagement, or is it too narrow?
Conversion rate: Are taps still turning into installs at the same quality level?
CPI/CPA stability: Is it competitive versus your baseline?
ROAS (if you track): Are purchases actually coming through for this tighter intent?

If the exact-match keyword performs worse than the original search term:

the original search term performance might have been influenced by broader intent/context from discovery
or the exact-match keyword isn’t matching as tightly as you assumed (language/grammar differences matter)

In that case, keep the exact keyword or adjust based on the direction of the metrics—but don’t immediately nuke it after a single day.

Step 8: Promote in batches, not as a one-time cleanup

A reliable workflow for small teams:

Every 1–2 weeks, pick 5–15 search terms that meet your criteria.
Promote them to exact-match keywords in a dedicated ad group.
Keep the number small enough that you can attribute performance changes to what you just did.

This batch approach prevents you from flooding the account with new exact keywords and losing the ability to tell what moved.

Common gotchas (worth checking)

Low volume mirages: A term that only ran a handful of times can look amazing or terrible—don’t promote just because one day was good.
Seasonality and app freshness: If your app launches updates or seasonal events shift behavior, winners can rotate. Promote based on stable patterns.
Attribution mapping delays: Apple attribution resolves within ~24h via AdServices, but your revenue mapping pipeline may add extra lag. Don’t judge profitability too early.
Budget fragmentation: Creating too many campaigns/ad groups can slow learning. Keep it focused.

Closing takeaway

Promoting winning search terms to exact-match keywords is about tightening intent with evidence. Use search term performance to identify candidates, move them into a clean exact-match ad group, start bids based on what you’re already paying effectively, and protect discovery from re-buying known converters.

If you want an easier version of this workflow, AdsBuddy can review your ASA performance alongside your mapped revenue signal and suggest the exact terms to promote (and the bid/structure changes to apply)—but you’ll stay in control and approve every step.