ConsultMyApp: how the Apple Ads auction really seems to work (relevance gate, then performance auction)

ConsultMyApp dug into a real UK Apple Search Ads dataset (132 keywords, 627 keyword to app observations across the top 5 ad slots) and asked a simple question: how much does “semantic relevance” actually explain the auction order once an ad is eligible?

• Source: ConsultMyApp, “How Does the Apple Ads Auction Actually Work?” (published Mar 13, 2026)
• Read: https://www.consultmyapp.com/blog/how-do-apple-ads-work

The core model (useful, even if you disagree with the exact numbers)

Their best-fit mental model is:

1. Candidate generation via exact match, broad match, and Search Match
2. Eligibility gating (relevance threshold depends on the pathway)
3. Auction ordering that is mostly bid plus predicted performance once you’re through the gate

In plain terms: relevance decides who gets in the room. Money and expected outcomes decide who gets the best seat.

What this changes operationally

• When you see an “irrelevant” competitor at #1, don’t jump straight to metadata blame. It might be Search Match (or broad) plus a strong bid/performance profile.
• When you’re highly relevant but stuck at #4 or #5, treat it as a commercial problem first. You might be underbidding, underconverting, or letting discovery traffic distort your account structure.
• Exact vs Search Match is not a preference, it’s a control system. You need clear separation so you can keep discovery noise from leaking into your core intent terms.

Tiny win (30 minutes)

Pick one expensive keyword where you “should” win.

• Put it in its own exact-match ad group.
• Add a control group or at least a clean baseline window.
• Write one line in your reporting doc: “If we lose rank here, is it relevance (eligibility) or commercial (ordering)?”

Read the source: https://www.consultmyapp.com/blog/how-do-apple-ads-work