How to Optimize Ads in the Shopify App Store

Ads on the Shopify App Store launched in February 2020 to provide a new way for app developers to accelerate their business growth by directly advertising to merchants looking for apps. For those taking advantage of this new channel, learning how to optimize ads in the Shopify App Store is the key to finding success. 

In this article, we discuss the different considerations to take into account when planning an ad campaign for the Shopify App Store.

How ads work

Before we dive in, let’s take a quick refresher on how ads in the Shopify App Store work. 

Advertisers compete in an auction on specific keywords to serve ads to merchants searching those keywords. 

When you create an ad, you specify which keywords you want to bid on. When a merchant searches for that keyword in the app store, an auction occurs with all participating apps who bid on that same keyword. The winner(s) of that auction gets to show their ad to that merchant. 

A cost is only incurred if a user clicks on the ad, and the cost per click is the bid amount at the time of the auction.

Getting started

Before you get started with ads, there are some essentials that you should have ready so that you are clear on what you want to achieve, how you measure results, and how you go about achieving these results. 

To help make your first foray into ads a bit more fruitful, think about the following three considerations.

1. Keywords

Do keyword research. This is really important because this determines your starting point for optimization—for example, do you already know what keywords your customers are searching for or is there some discovery still remaining?

Choosing strong keywords to bid on means understanding what kind of terms and phrases your potential users search for in the app store. To figure that out, you’ll need to understand your users’ needs and problems at every stage of the conversion funnel, so you can figure out how to best target users at each level.

The conversion funnel

The conversion funnel is a way of visualizing the steps a user goes through before deciding to buy (or in this case, install) your product. A simple example of a conversion funnel looks like this:

At the first stage, users have a specific problem they want to solve and are looking at different ways they can solve that problem. At this stage, you’re really just looking to become one of the potential solutions your user considers. 

At the second stage, users have a set of features they’d like to learn more about, and are searching for more information about those specific features. 

At the third and final stage, users have a clear idea of what they’re looking for, and are looking for more information specifically about those offerings.

For an in-depth look at how the funnel affects how you market your app, read our article on how to get more downloads in the Shopify App Store.

Users at different stages of the funnel search for different information, which means that depending on the stage you choose to target, your keywords will change. As an example, let’s look at Shopify. Business owners search different things, depending on how seriously they’re considering using Shopify. Some examples of keywords for each of the three stages of the funnel could be:

• Awareness: online store, ecommerce, selling online, online sales
• Consideration: shipping online orders, online payments, online store competitors
• Conversion: Shopify pricing, Shopify free trial, Shopify monthly plan

Conducting market research for your app will help you get closer to your users and better understand the terms they may be searching for at every level of the funnel.

Leverage data

In addition to considering the conversion funnel, you may be able to find promising keywords from the data generated by your website, other analytics platforms, and your app listing. While these sources likely won’t translate perfectly to success with app store ads, they will provide guardrails and inspiration for further keywords you can leverage. 

The following tools can help you with this research:

• Google Search Console: Visibility into what keywords are bringing people to your website 
• Google Ads Keyword Planner: Tool to research potential keywords
• Google Analytics: Visibility into what terms merchants are searching in the Shopify App Store before they arrive at your app listing (as long as you have an existing app listing and have added your Google Analytics ID) 

You might also like: 7 Insights From the Shopify App Review Team to Set Your App Up for Success.

2. Budget

An important facet of planning your ad campaign is figuring out how much you’re comfortable spending. When you’re starting out without any historical performance data, the best way to get familiar with ads is to set aside some money to experiment with, and increase or decrease the amount as you exceed or underperform expectations. 

As you start to get results on how many installs you acquire at what cost, you can start to make more long term adjustments to how much you should spend. 

If this is your first time advertising on the Shopify App Store, you might be eligible for a $100 ad credit to get you started. Learn more and claim your credit.

3. Key performance indicators (KPI)

In the section on objectives above, we discussed choosing your goals for your ad campaign, and how that ties into how you measure success. Charting your KPIs in advance means that you’ll be ready to gauge whether your ad campaign has been successful.

If the objective is user growth, you probably want to look at Cost per Install (CPI) and start developing a performance benchmark for the keywords that you are bidding on. That is, keep track of your lowest and highest CPI ads/keywords, as well as your overall average CPI. 

If the objective is customer growth, you’ll want to look at your Customer acquisition cost (CAC). Similar to the CPI objective above, start developing a historical performance benchmark. Look at where your average sits, and what factors are contributing to high or low CAC.

If the objective is to improve Return on Ad Spend (ROAS), similar to the above two, you’ll want to start building your own performance benchmarks. The key difference with ROAS is that you are aiming for higher ROAS rather than lower costs (CPI and CAC). 

How to interpret performance metrics

On the Ads page in the Partner Dashboard, you have access to visibility, average position, CPI, CAC, ROAS, spend amount, volume of impressions, clicks, and installs metrics. Each of them tells you something different about the performance of your ads. Below, we look at each of these metrics.

1. Visibility 

The visibility metric tells you how often you’re winning a position on the first page in auction. Generally speaking, you’ll want higher visibility where possible. Higher visibility means getting your app in front of more merchants on the first page, which means more opportunities for clicks and installs.

Typically, visibility is correlated with bid amount and relevance, but other factors can make 100 percent visibility not possible for certain keywords. These factors include relevance, overlapping bids, and others. If you’re getting an average position of one, but your visibility isn’t increasing no matter how high you bid, chances are the bid is not the limiting factor.

2. Average position 

The average position metric tells you which position in the search results you appear on average. Like visibility, this is correlated with bid amount and relevance, but unlike visibility, there are no additional factors that prevent advertisers from achieving an average position of one. 

It’s important to keep your CPI in mind when you look at and make adjustments to bids and keywords for the sake of improving average position. A good average position does not guarantee a good CPI, so you’ll want to strike the right balance between good average position and CPI. 

For example, an average position of two may yield the same CAC or CPI as an average position of one. In such a case, it may not be worth it to bid the amount needed for first position. You’ll need to experiment with your bids and monitor results to see if this is true for your particular app and keywords.

Bid suggestions 

Bid suggestions are the approximate bid that is required for your app to be competitive in auction. For exact match keywords, the bid suggestion represents the bid range required for a first page auction win. For broad match keywords, the bid suggestion represents a measure of competitiveness against other advertisers. Because broad match keywords match so many different search terms, actual performance in terms of average position and visibility may vary greatly.

Use these bid suggestions to help guide your bid setting. They do not guarantee results, but are meant to give you a more accurate starting point to better performance. 

3. Cost per install (CPI)

If the objective of running ads is to grow your user base, CPI is likely one of the most important metrics. Compare your actual CPI to your historical CPI range, and you have the direct measure of how your ads are performing relative to your own benchmark. 

Look at where your average CPI falls in your range. What you want to see is your average CPI sitting at the lower end of your range. If not, you’ll want to identify which ads or keywords are high cost and make the requisite adjustments or optimizations (detailed later in this article). 

For accounts with a lot of history that have been heavily optimized, it’s okay if the CPI sits in line with your average vs. at the lower end of the range. 

4. Customer acquisition cost (CAC)

If the objective of running ads is to grow your paying customer base, CAC is likely one of the most important metrics for you. Like CPI above, use your historical benchmark to identify if some of your ads or keywords are high cost, and adjust accordingly. 

Because customer data is reported against the date the ad impression was served, when looking at CAC performance, it’s best to look at data older than 60 days to account for the lag in customer conversion times.

5. Return on ad spend (ROAS)

If the objective of running ads is to maximize the efficient revenue acquisition, ROAS is likely one of the most important metrics. Compare your actual ROAS to your historical ROAS range, and you have the direct measure of how your ads are performing relative to your own historical performance. 

Look at where your average ROAS falls in your range. What you want to see is your overall ROAS sitting at the higher end of your range. If not, you’ll want to identify which ads or keywords are low ROAS and make the requisite adjustments or optimizations (detailed later in this article).

For accounts with a lot of history that have been heavily optimized, it’s okay if ROAS sits in line with your average vs. at the higher end of the range. Because revenue data is reported against the date the ad impression was served, when looking at ROAS performance, it’s best to look at data older than 60 days to account for the lag in customer conversion times.

6. Spend amount or install volume 

CPI the metric by which you measure your success, but by itself, it can lack context. That’s where the spend amount or install volume comes in. 

Some keywords can have very low or very high CPI, but represent only a small fraction of your total spend or total installs. Ideally you want to see both a high install volume or spend amount, and a low CPI on the keyword or search term. Bid adjustments made to your largest spend and highest install volume keywords tend to have the largest impact. See a comparison of two example results below.

In the above example, the keyword with the greatest amount of spend ([pixel tracking]) is the one that has the highest number of installs at the lowest CPI.

In the example above, the keyword with the lowest CPI has only $50.00 ([google ads]) of spend against it.

You might also like: Tips for Creating Effective Technical Documentation for Your Shopify App.

Identifying poor results and what optimizations to make

In this section, we provide some examples of what poor results look like, and what you can do to improve them.

Low visibility 

By itself, the low visibility metric simply shows that you’re not winning a large portion of your first page auctions, and in many cases is directly influenced by two factors: your relevance score and your bid. Raising your bid is the most direct way to improve visibility, but as mentioned above, relevance and other factors also impact visibility.

It’s best not to make changes based on only low visibility. Instead, you’ll want to look at this metric in conjunction with other indicators, like average position (see below). 

In the example below, we see that visibility is low for [pixel tracking] but that average position is 1.1. In such a case, increasing the bid is not likely going to improve visibility as the average position is already one.

In the example below for [pixel tracking], we see that both visibility and average position are low (40 percent and 5.6 respectively), and relevance is not low or very low. In such cases, increasing your bid is likely to have an impact on visibility.

Poor average position 

Poor average position is usually defined as the average position of four or greater (average impression showing up on page two or three). 

Low average position combined with low visibility and medium or better relevance likely means that your bid is not high enough. You can increase your bid to try and improve both. If you’re able to get a first page average position and visibility, this can lead to improved performance as you have much greater visibility with merchants.

In the example below for [pixel tracking], we see that there is an average position of 5.6, a visibility of 40 percent, and high relevance. Increasing the bid is likely to have a significant impact on average position.

In the example below, a high average position (1.1) combined low visibility (40 percent) and low relevance (see image below) likely means that your bid is sufficiently high, but other factors prevent visibility from improving. Overlapping bids, or bidding on broad match keywords are two common causes. 

High CPI or high CAC

When you start to see high overall CPI or CAC (relative to your historical range), it’s a sign that you need to make some optimizations. Some causes and the optimizations you can make include:

• Continuing to bid on keywords with no installs or customers that have a high spend amount or high impression volume. Each advertiser needs to set their own threshold for what they determine to be sufficient spend or impression volume. The solution here is to pause this keyword or substantially lower the bid as it’s not worth the cost to keep this keyword active.
• Overpaying for clicks. If your average position is already one or under 1.3, it’s unlikely that increasing your bid further will yield much improvement on CPI or CAC.
• Relying only on broad match keywords. This can lead to high CPI or CAC because of all the search terms that often match to a broad match keyword. Identify which search terms for each broad match keyword are driving installs at or below your target cost, and add them to a new campaign as exact match keywords.

If you’ve tried experimenting with bids and match types, and still can’t lower your CPI, it might be worth looking at a different set of keywords to see if you are able to acquire installs at a low cost that way.

Low ROAS

When you start to see low overall ROAS (relative to your historical range), it’s a sign that you need to make some optimizations. Common causes and the actions to take are very similar to what you would do for high CPI or high CAC:

• Identify keywords with high spend and low (or no) ROAS. If you’ve spent quite a bit but are still seeing very little ROAS, it’s probably not worth keeping this keyword active.

• Overpaying for clicks. If your average position is already one or under two, it’s unlikely that increasing your bid further will yield much improvement to ROAS.
  
  
• Relying only on broad match keywords. This can lead to low ROAS because of all the search terms that often match to a broad match keyword. Identify which search terms for each broad match keyword are driving installs at or below your target cost, and add them to a new campaign as exact match keywords.

No impressions served 

When you’ve added a keyword to an active ad and see zero impressions for a given ad or keyword, there are usually four reasons. Here are some actions you can try to show ads for certain keywords:

• Very low relevance. Having low relevance does two things: increases the bid required to win auctions, and decreases your ability to participate in auction.
• Incorrect date ranges selected: The ad was only created on June 2nd, but the date range you’re looking at is May 1 — 31.
• Low bid: Try gradually raising your bid in controlled increments to see if it’s a low bid that’s causing this.
• Keyword added as a negative keyword: If you bid on the keyword [marketing] and have [marketing] added as a negative keyword, this will prevent any impressions from being served.

You might also like: Research 101: How to Conduct Market Research for Your App.

Other ways to optimize your ads

Outside of the poor results identified above, listed below are some optimizations you can make to improve overall performance.

Adding keywords

If you can’t achieve your target CPI with the keywords you’re currently bidding on, try adding other keywords from your research or from our keyword recommendations (see image below) in the Partner Dashboard and see if you can acquire installs at a lower cost.

Reorganize ads

If your ads are constantly out of budget, and you feel your high performing keywords aren’t getting as much of your daily budget, it might be worth reorganizing the way you group your ads so that your best performing keywords get allocated as much budget as possible.

The best way to think about this is to organize your ads by performance groupings, and allocate budget according to what performs best, rather than have all of your keywords lumped together in one ad. The thinking is that some keywords perform better than others—you want to isolate and maximize how much budget is allocated to those. 

For example, dividing your keywords by exact match marketing keywords, exact match Facebook ads, and broad match Facebook ads, like in the image below, helps you funnel more money to the highest performing batch (in this case, exact match marketing keywords).

Using ads for app store success

We hope this optimization guide was useful, and that you’re able to identify some actionable optimizations for your own campaigns so you can better achieve your growth goals, and improve your overall advertising efficiency.