The Anatomy of Apple Search

How to Build A Successful Apple Search Strategy

To understand Apple search, you must understand its three core elements:

1. Campaign
2. Ad Group
3. Keyword

Campaigns

Campaigns are shells which can contain unlimited ad groups.

You don’t have many levers to pull at the campaign-level: only your promoted app, country storefront and budget.

You should generally treat your campaign as your audience. I always recommend that advertisers choose one campaign per audience.

The campaign creation screen

Ad Groups

You set your audience at the ad group level. You can choose to target according to City/State, Age, Gender and New User vs. Existing User status.

The audience selection screen within an ad group

All ad groups within a campaign should contain identical audience targeting. You should create multiple ad groups when you want to cluster together different sets of keywords against the same audience.

Apple defaults all ad groups into Search Match, which automatically generates a list of keywords based on your app’s metadata.

The search match selection module within an ad group

Search Match can be useful for discovering new keywords, but never include an ad group with both keywords and Search Match enabled, as you’ll lose your ability to control the levers of performance. Instead, flight a dedicated ‘Search Match’ ad group with no keywords.

Keywords

These are the keywords you want to show your ads against. You set bids at the keyword-level and bid on a cost per tap basis. This is equivalent to bidding on a cost per click on Google.

The keyword selection screen for Bloom, a secure identity app.

You can choose to target either ‘Exact Match’ or ‘Broad Match’ for each keyword. This follows the same logic as Google Search.

With Broad Match, you’ll target any search term that includes the keyword, including misspellings, synonyms and phrases that include it. With Exact Match, you’ll only target search terms that include your keyword in order or close variants of your keyword in order.

You can also exclude keywords from every ad group using negative keywords.

Matrix of keyword results via Incipia

Creative Sets

By default, your ad creative on Apple Search will be the first three screens from your App Store listing.

While Apple doesn’t offer advertisers much in the way of creative freedom, they recently introduced a degree of freedom with Creative Sets.

Creative Sets allow advertisers to specify which screens are shown and in what order at the ad group level. The main function is to personalize the ad creative to the keywords.

An example of the creative set screen selection, in which you can select any 3 screens from your app store to use in your Apple Search ad creative. By default, your ad is comprised of the first 3 screens.

Apple doesn’t allow a/b testing on the app store, so you can setup some clever albeit limited experiments through Creative Sets to determine which order of app screens drives the highest click rate. If you find that a specific order of screens is driving a higher tap through rate than your default ad, you might want to consider adjusting your main App Store page to drive higher conversion rate on all traffic.

Optimal Campaign Structure

While the optimal structure will vary from app to app, I’ve created a chart that most apps can adapt for their Apple Search Strategy.

Each campaign holds at least 7 ad groups, differentiated by category and exact vs. broad match. Each ad group excludes keywords from the prior ad group, such that each ad group is targeting a mutually exclusive set of keywords in a waterfall fashion. You should never be bidding on the same keywords across multiple ad groups within the same campaign/audience.

For example, an app like Headspace would set their campaign up as follows. This only includes one keyword per ad group for illustrative purposes.

You’ll notice that I included a search match ad group at the bottom of the waterfall. The function of this ad group is to pick up long tail keywords that you may have missed. If the performance on any of these Search Match keywords is competitive, promote the keywords to a different ad group and add it to a negative keyword to the Search Match ad group.

When to Create Additional Ad Groups

Functional keywords can consist of more than two ad groups. Here’s the rule of thumb: you should create unique ad groups wherever a given set of keywords could benefit from differentiated creative.

For example, while ‘Social Travel’, ‘Travel Journal’ and ‘Travel App’ all would be considered functional keywords, it makes sense for TripTrek to create unique ad groups for each keyword so that they can assign a unique creative set to each, which allows TripTrek to increase their performance efficiency.

When to Create Additional Campaigns

Apple’s UI is extremely inefficient and optimization becomes laborious with too many separate campaigns. However, there are two situations when you’ll want to break out audiences into multiple campaigns:

1. Value difference. You should create multiple campaigns across targeting dimensions that differentiate user value.For example, if you find that installs from the 18–21 age group are worth 20% less than users age 22+, you should create separate campaigns for age 18–21 and reduce your install bid by 20%.
2. Creative difference. You should create multiple campaigns across targeting dimensions that differentiate ad creative response. For example, if you find that female users respond better to female screenshots, it makes sense to break out separate campaigns for male and female users so you can assign appropriate creative sets to each.

So let’s say you have a US-only app where (a) users 18–21 are worth 20% less than users who are aged 22+and (b) you notice that pairing the user gender with the ad creative gender drives a significantly reduced cost per tap. You’ll want to setup your campaigns as follows:

While clustering your campaigns together would make your day-to-day Apple Search workflow easier, breaking things out in this fashion allows you to (1) calibrate your bid according to predicted user value differences for younger users while (2) driving better ad performance via serving gender-appropriate ads.

You’ll notice I included a final campaign with no targeting and a significant bid reduction. This is our catch-all audience, designed to collect all users with limit ad tracking enabled who would otherwise be excluded from our targeting. You’ll typically want to bid down aggressively on this audience because they’ll show up as organic users due to the lack of IDFA tracking.

Measuring Beyond the Install

You’re not going to find success on Apple Search if you optimize on the basis of which keywords drive the lowest cost per install.

Imagine you’re promoting a shopping app and analyzing the relative performance of the exact match keywords [Free Stuff] and [Luxury].

You’re getting installs from [Free Stuff] for $.25 and installs from [Luxury] for $2.50. You decide to shift your budget to [Free Stuff] because that’s driving much cheaper installs! Unfortunately it turned out that the average install from [Free Stuff] yielded $.10 in purchase value, while the average install from [Luxury] yielded $10 in purchase value, so you made an objectively poor media allocation decision.

You can avoid this situation by always leveraging a mobile attribution partner like AppsFlyer, which can help you measure what post-install behavior you’re getting from each keyword. This is in fact the only way you can truly measure the ROI from Apple Search campaigns.

I always recommend that advertisers allocate their marketing budget on the basis of driving the highest return on investment and not the lowest cost per install.