Pause, Lower, or Delete? How to Retire an Apple Search Ads Keyword Without Losing the Signal

Every Apple Search Ads account accumulates keywords that stopped earning their place. Maybe a term brought taps but no installs. Maybe it was a good idea in March and a money pit by June. Most indie developers handle these with a vague gut decision: they either leave the keyword alone forever or delete it in a panic. Both habits cost you. The useful move is to choose between three actions on purpose: pause, lower the bid, or delete. Each one answers a different question.

Why the choice matters

A keyword is not just a line item. It is a record of how Apple's auction has treated your ad for a specific query, at a specific bid, in a specific country. When you change or remove it, you change what your future self can learn from the data. If you delete a keyword that was merely underperforming because its bid was too low to win impressions, you lose the evidence that would tell you whether a higher bid would have worked. If you keep a keyword that is clearly bleeding CPT, you keep paying to learn something you already know.

So before touching anything, ask what you actually know about the keyword. Three cases cover almost everything.

Case 1: Pause when the signal is unclear

Pause a keyword when you do not yet trust your own conclusion. Typical situations:

• It has a handful of taps but too few to judge conversion, and the spend is small.
• Your attribution has a known gap, such as a recent change to your paywall or onboarding that makes the install-to-purchase chain noisy.
• You are still waiting on a related change, like a new product page, and want to isolate the effect.
• The keyword is part of a cluster you are reorganizing, and you do not want it competing with its replacement mid-test.

Pausing is reversible and keeps the keyword's history attached to the same entry. It stops spend without asking you to make a final judgment. The downside is clutter: paused keywords pile up, and it is easy to forget why they are there. Add a note in your own tracking sheet, with the date and the reason, so a paused keyword is never a mystery six weeks later.

Case 2: Lower the bid when the keyword works but costs too much

This is the most common case that people mishandle. A keyword might convert reasonably well at its current max CPT but still lose money on a revenue basis, because the CPT is simply too high for what an install is worth to you. The problem is the price, not the term.

In this situation, reduce the max CPT in steps rather than dropping it to a level that stops all delivery. Keep an eye on impressions and taps over several days. If taps fall off a cliff, you have found the bid floor where the keyword stops competing, and you now know that the term's value sits in a narrow band. That is useful information. Lowering the bid preserves the keyword's history and keeps it available as a cheap source of installs if your conversion improves later.

A concrete illustration: suppose a keyword costs a steady amount per tap and you have decided the install is only worth paying so much for. You step the bid down, watch taps, and find that taps hold up until a certain point. That point is your working bid. Nothing about the term changed; only your price did.

Case 3: Delete when the keyword is structurally wrong

Delete only when the keyword itself is the problem, not its price or your current data. Good reasons include:

• The term is misspelled, ambiguous, or matches queries that clearly belong to another app category.
• The keyword duplicates another keyword in a different campaign or ad group, so it is competing with yourself. (This is the overlap problem, and deleting one copy is often the right fix.)
• The term reflects a feature you no longer offer or a name you have dropped.
• It was added as a test, the test is over, and you have a clear answer.

Be careful with deletion as a reaction to a bad week. A keyword with thirty days of weak data is often just underfunded or underexposed, and deleting it erases the one thing that would let you judge it. Also remember that re-adding a deleted keyword does not restore its history in a meaningful way, so treat re-adding as a fresh experiment with a fresh bid.

A simple decision sequence

When you review a keyword, walk through these questions in order:

• Is it structurally wrong, meaning irrelevant, misspelled, or a duplicate? If yes, delete.
• Is the signal trustworthy? If you cannot yet tell, pause it and note why.
• Does it convert, but the CPT is too high for its revenue? Lower the bid in steps.
• Does it perform well at a sustainable bid? Leave it alone.

The order matters. Most people start with performance and jump straight to delete, when the honest first question is whether the keyword is even relevant.

Protect your reporting

Whatever you choose, write it down. A short log with the keyword, the action, the date, the bid before and after, and the reason is enough. Over a few months this log becomes the most valuable part of your account, because it lets you distinguish a genuinely bad term from a term you simply stopped funding too early. Pair it with a revenue view that uses a reasonable time window, since subscription purchases often land well after the tap.

Takeaway

Pause when you are unsure, lower the bid when the term works but costs too much, and delete only when the keyword is wrong for your app or duplicates something else. Each action preserves a different amount of information, so pick based on what you actually know. Keeping a simple log of these decisions is what turns keyword cleanup from guesswork into a repeatable habit.

If you want a second pair of eyes on which keywords fall into which bucket each week, AdsBuddy reads your Apple Search Ads and revenue data and suggests a short, prioritized list of changes. You review each one and apply it yourself.