- 1. Cost Per Action (CPA)
- 2. How to Calculate ROAS and ROI
- 3. Cost per Tap (CPT)
- 4. Conversion Volume (CV)
- 5. Conversion Rate (CR)
- 6. Optimizing mobile advertising metrics with SearchAdsHQ
Cost-per-tap is an Apple Search Ads advertising model where you pay for every tap on your banner. The trick about CPT is that you don’t earn from the tap itself. You earn from what happens after.
After users tap on your ad, they find themselves on an app store product page, then (if things shape up well) they download, install and open your app. Only upon these actions users may make an in-app purchase, buy a subscription or view native ads.
Users can get stuck or drop out at any stage of the funnel. How to make sure you buy those who will bring in revenue? There is a whole bunch of Apple Search Ads metrics that we can use to measure the performance of the funnel. Let’s bring into focus 5 of them:
- Cost Per Action (CPA);
- Return on Ad Spend (ROAS) & Return on Investment (ROI);
- Cost Per Tap (CPT);
- Conversion Volume (CV);
- Conversion Rate (CR).
For the latest TTR, CVR, CPT, CPA benchmarks, check out our Apple Search Ads Performance report.
1. Cost Per Action (CPA)
Post-install target actions are specific for each app and depend on mobile app monetization. These can be:
- In-app purchases;
- Starting of a trial, etc.
Marketers who optimize their campaigns for a particular action keep the Cost per Action (CPA) metric top-of-mind. Basically, CPA is the average cost of conversion actions performed by users. If you want your campaigns to be profitable, this cost should be lower than the revenue generated from the action.
How to calculate your current CPA? Divide the overall spend by the number of target actions performed by users.
CPA = Spend / Action
Let’s say you’ve spent $100 on an ad, which resulted in 10 in-app purchases. $100/10 = $10 – this is the average amount you pay for a target action.
CPA optimization in Apple Search Ads
CPA optimization is about how to get users who will perform a target action at a lower target cost.
For example, take a subscription-based fitness app. Subscription is a target action in this case. Let’s assume that we know the average LTV of our paying users and this LTV stays the same. Let’s say an average user pays $40 per year (ARPU) and stays with you for 2 years (LT). Knowing ARPU and LT, we can calculate LTV:
LTV = LT x ARPU = 2 x $40 = $80
So, in this case the cost of a target action should not exceed LTV = $80. But that’s not the amount we want to pay, right? We want to somehow get profit.
The thing is, LTV is gross profit, but there are fixed acquisition and support costs as well. You will probably have to pay a 30% fee per user’s subscription to Apple’s App Store and around 20% as additional taxes, customer support or content production costs.
Let’s get back to your fitness app. Taking into account 50% (30% + 20%) off the user LTV, your profit per subscriber will be:
Profit = LTV x 50% = $80 x 50% = $40
$40 is CPA which will keep you at a break-even point. But if you want your profit to be greater than costs, try to bring CPA further down.
Target CPA = 75% x Profit = 75% x $40 = $30
There, you have it, $30 is the optimum CPA which will produce a $10 profit per your fitness app subscriber with 2 years of LT.
To sum it up, here’s how it works:
|LTV||$80||ARPU x LT|
|Subtract 30% of App Store fee|
Subtract 20% of support costs
|$40||Fixed costs for subscription services|
|Break-even CPA||$40||If we use this CPA, we will get the same amount that we pay|
|75% x CPA|
|$30||If we use this CPA, we will get a profit of $10 from every paying user during their LT|
Now that you know your Target CPA, you can adjust the CPT bid of a keyword according to the following funnel:
Max CPT Bid = Conv. Rate * Target CPA * 1.2
Conv.Rate = Conversions / Taps = 20 / 400 = 5%
Max CPT Bid = 5% * $30 * 1.2 = $1.8
You may wonder where 1.2 came from. In reality, CPT is usually around 10-20% lower than the set Max CPT Bid because of a second price auction. For instance, if you’ve set Max CPT Bid to $1.00, you are likely to spend around 20% less – $0.80 per tap on average. So to compensate the difference, multiply the bid by 1.2.
Conversion Rate = Conversions / Taps = 20 / 400 = 5%
Max CPT Bid = 5% x $30 x 1.2 = $1.8
This method of CPA bid optimization has to be applied to each keyword according to its specific funnel.
2. How to Calculate ROAS and ROI
Post-install events are not limited to subscriptions only. There are in-app purchases where users pay variable prices (e.g. one user buying virtual currency in a game can contribute from $1 to $1,000). This monetization model makes it hardly possible to calculate Target CPA. In such situations, marketers assess advertising profitability by using ROAS.
Simply put, Return on Ad Spend is the number of dollars you make from an ad campaign per each dollar spent on advertising. The Return on Ad Spend formula is:
ROAS = Revenue / Ad Spend x 100%
Suppose you allocate $100 for advertising, which brings you a $120 revenue. Here’s how to calculate ROAS:
ROAS = Revenue / Ad Spend x 100% = $120 / $100 x 100% = 120% (1.2x)
How can the result be interpreted? You’re on the better side of a cost-return balance, as per $1 invested in advertising you earn $1.2. In a perfect world, revenue from ad campaigns should exceed advertising costs, that is why healthy ROAS is over 100%. Conversely, ROAS under 100% is a warning signal that optimization is needed.
ROAS usually comes in tandem with Return on Investment (ROI). While ROAS measures the performance of a dollar invested in advertising, ROI takes into consideration marketing efforts as a whole – overheads, salary, employee training and other factors including dollars spent.
ROAS is all about the performance of a particular marketing channel, while ROI is a broader scale measure of the overall marketing performance. Another difference between the two metrics is that ROAS always shows a positive number, while ROI can be negative.
Here’s a formula to calculate ROI:
ROI = Profit / Spend x 100% = (Revenue – Spend) / Spend x 100%
ROI = ($120 – $100) / $100 x 100% = 20%
And here’s how ROI measures up against ROAS:
ROI = ROAS – 100% = 120% – 100% = 20%
ROAS Optimization in Apple Search Ads
The idea behind ROAS optimization is to keep return at a particular level. It’s not our advertising costs that matter here, but the amount of return you earn from your investments.
Let’s imagine you market a game app where users can buy gold for real money. One in-app purchase can bring you from $1 to $1,000, which means you can’t calculate the average CPA. So, ROAS comes into play.
Similar to CPA optimization, you will have to contribute 50% of the generated revenue – 30% as a fee to the App Store and 20% as support costs.
Let’s calculate your margin per revenue of $1,000:
|30% – the App Store fee|
20% – support costs
|$500||Acquisition and support costs needed to raise the revenue|
|Profit per Sales||$500|
|Break-even Ad Spend||$500|
|Max ACoS||50%||ACoS stands for Ad Cost of Sales. As long as you don’t spend more than 50% of the revenue ($500) on ads, you’ll break even.|
|Min ROAS||2.0x (200%)||ROAS = 1 / ACoS = (1 / 0.5 = 2.0x)|
Return on Ad Spend is 200%
|Target ROAS||2.5x (250%)||If you want to make a profit, maintain return greater than Min ROAS, in other words, reinvest a smaller portion of revenue into ads.|
Target ACoS = 1 / 2.5 = 40% (not 50%)
Some businesses want to hit Target ROAS of 2.5x, others – 4x. It’s up to you to decide on the portion of Profit per Sale you want to spend on ads.
Let’s get down to bid optimization. Below is the formula to count Max CPT Bid for this funnel:
Max CPT Bid = (Revenue / Taps) x (1 / Target ROAS) x 1.2
Max CPT Bid = ($1,000 / 400) x (1 / 2.5) x 1.2 = $1.2
This is the way you optimize bid for a ROAS-based model. You have to apply this calculation to each keyword based on its specific funnel.
True ROI/ROAS in Apple Search Ads
PPC is a highly transparent way to measure performance, that’s why marketers like it. But when it comes to Apple Search Ads, the process gets tricky, and the reason is data discrepancies.
Data discrepancies between an advertising network and MMPs are natural. In most cases, they don’t exceed 5%, so marketers see them as a margin of error.
Apple Search Ads is different. Data discrepancies range from 30 to 70%, which means up to 70% of revenue from this channel can be hidden, i.e. attributed to the organic traffic or other marketing channels. Why does it happen? Here are a few reasons:
- Re-downloads. Re-downloaders are users who download the app more than once. Depending on their settings, MMPs can recognize that this is a re-download by an existing user and attribute it to the media source this user initially came from. Some app categories, for example, dating, have over 60% of re-downloads.
- Limit Ad Tracking (LAT). LAT helps Apple users prevent their devices from being tracked by media sources and MMPs. When LAT is turned “On”, the Identifier for Advertisers (IDFA) associated with the device is substituted with zeroes. As a result, none of the user data, including app installs, can be collected by MMPs, thus Apple Search Ads is not reflected in the statistics of your mobile tracker and you don’t get a full picture of attribution.
On average, LAT-based discrepancies make 15-20%.
- API glitch, open latencies, attribution settings. These causes for discrepancies usually amount to 5-10%. For other reasons of mismatch read Why Search Ads and trackers show different data.
You may wonder what discrepancies have to do with ROAS and ROI? As the mismatch of data can reach 70%, the ROAS calculated ignoring that mismatch is distorted. Let’s introduce an example for clarity. Suppose you promote an app, and your marketing activities produced the following results:
|Apple Ad Spend||$200|
|LAT On Conversions||8 (8%)|
|API glitch, click latencies, attribution settings||5 (5%)|
First of all, let’s calculate ROI ignoring data discrepancies:
Strict ROI = (MMP Installs x ARPU – Spend) / Spend x 100%
Strict ROI = (50 x $5 – $200) / $200 x 100% = 25%
ROI equalling 25% means that you spend more than you get, and either optimization is needed or the advertising channel doesn’t line up with your unit economics.
Now, let’s take into account hidden installs. For the sake of simplicity, let’s add re-downloads and LAT On Conversions together (assuming that they don’t overlap).
Adjusted MMP Installs = 50 + 27 + 8 + 5 = 90
Adjusted ROI = (90 x $5 – $200) / $200 x 100% = 125%
Look at the dramatic difference: the Adjusted ROI is 5 times bigger than Strict ROI. By doing the calculations this way, you can see the true picture of your Apple Search Ads performance and thus bid more competitively and efficiently scale your campaigns.
In fact, the same method can be applied to CPA calculation. Find more about true ROAS and ROI in the AppsFlyer blog post by Thomas Petit.
3. Cost per Tap (CPT)
Cost per Tap (CPT) is an underpinning pricing model of Search Ads. In the Apple’s advertising network you pay only when users tap on your banner, and CPT shows how much this tap costs you. For instance, you’ve spent $1,000 on advertising and received 500 ad taps, your CPT will be:
СPT = Ad Spend / Taps = $1,000 / 500 = $2
The higher your CPT, the sooner your ad budget fades, and at some point you’ll find it hard to stick to the desired ROAS. On the other hand, low CPT decreases your chances of hitting the auction. As a result, your ad falls short of impressions.
Similar to CPA and ROAS, CPT traces the overall health of your marketing campaign. But is there such a thing as good CPT in Apple Search Ads?
There is. But first of all let’s explore the benchmarks. Our SearchAdsHQ team has researched the key App Store categories and their average CPT, and here are our findings:
With the average Apple Search Ads CPT of around $0.84, the top 4 categories are as follows:
- Shopping: $3.48;
- Finance: $2.81;
- Travel: $1.27;
- Games: $0.96.
Now let’s shift your focus from heavy bidders to the most popular App Store categories – Games, Business, Lifestyle and Education. Look at the statistics above: app publishers who fall under these 4 categories hover somewhere around $1. Such CPT makes Apple Search Ads stay competitive with other advertising networks.
CPT Bid Optimization in Apple Search Ads
We’ve already touched upon it: either high or low CPT bid won’t work well in Search Ads. Look at the ways to find the most optimal Cost per Tap:
- Adjust bids up to target metrics. This is the most straightforward optimization technique. Keep your Target metrics – CPA and ROAS in focus and increase or decrease your CPT bid accordingly (see previous chapters).
- Use negative keywords. Search terms with a low relevance score eat up your ad budget. By using negative keywords, you can prevent your ad from being matched to irrelevant search terms. This, in turn, will decrease your CPT bid and improve other metrics across the funnel, and you will be able to scale your campaign with no harm to efficiency.
- Pause keywords. Similar to the previous method, irrelevant or non-performing keywords can be paused. By redirecting your advertising budget towards more efficient keywords, you can bring the CPT bid down.
- Work with segments. It’s common knowledge that marketers split the audience into groups to deliver tailored messages. Segmentation can equally work to optimize the CPT bid. For example, statistically the female audience is more likely to convert compared to men. So, if you decrease the CPT bid for the male segment, the average CPT will go down as well.
Here’s another common example. Audiences may behave differently throughout a week, in other words, they are active Monday to Friday and inactive at weekends. If it sounds familiar to you, there’s a way to optimize your CPT bid – just turn your ads off at weekends and watch the average CPT decrease. But remember to maintain the reasonable balance between cost per tap and conversion volume.
4. Conversion Volume (CV)
Reaching the target values for CPA, ROAS and CPT is vital, but only half of the job. Ideally, good metrics should be backed by a reasonable amount of conversions. This is where the conversion volume comes into play.
Simply put, conversion volume is the number of conversions generated for a certain period, e.g. week or month. The metric has nothing to say about how efficiently you advertise, but you should maintain it at a certain level for that efficiency. Moreover, high conversion volume contributes to the statistical significance of your decisions, because you accumulate enough data to rely on.
In a perfect world, conversion volume should grow as time goes by, while the cost of advertising should stay on the same level or even drop. The good news is that, in comparison with CRO, conversion volume is much easier to optimize. So, what are the ways to boost it?
- Add more keywords. By searching and adding relevant keywords, you increase your chances of getting extra impressions, and thus more conversions. Practice the keyword expansion technique from time to time and include efficient keywords to your top performers list.
- Run discovery campaigns. Apple Search Ads discovery campaigns are one of the methods to expand your list of keywords by means of Search Match and Broad Match. These wide reach match types do a good job of finding relevant terms, synonyms, long-tail keywords, misspellings, etc., associated with a keyword. The most high-performing of the generated keywords are then transferred to exact campaigns.
- Increase/reallocate budget. A very tight daily budget can significantly limit your impressions. It’s reasonable to revise your allocations sometimes and identify the areas worth investing. Focus on localizations, ad groups and keywords with a strong performance.
- Expand targeting. If you run high-performing campaigns in the US storefront, chances are that you may be that successful in other English-speaking countries. Give it a try.
- Increase your bid. Low bids may be a reason for not entering into an auction, and consequently for the traffic loss. Thus, raising bids is a smart way to improve competitiveness. Apple’s bid insights hint at the areas needing optimization.
5. Conversion Rate (CR)
Сonversion rate (CR) is the perfect metric to estimate whether the funnel you’ve built is poor or healthy. It shows the overall number of conversions generated from taps to a particular in-app action within a period. Here’s the formula to calculate CR:
Conversion Rate = Conversions / Taps x 100%
Let’s see how to calculate a conversion rate for the funnel above:
Conversion Rate = Conversions / Taps x 100% = 20 / 400 x 100% = 5%
You’re doing good if the conversion rate is high. In such a situation, you pay less for a particular target action, thus the allocated budget allows to acquire more users. On the contrary, the poor conversion rate is a red flag that optimization (CRO) is needed.
You can optimize conversion at any funnel stage, including post-install events. For now, let’s focus on pre-install conversions – from impressions to taps (TTR) and from taps to downloads (Download Rate). What can you do to improve the rate?
- A/B test creative sets. We’ve run a lot of experiments and can say for sure that smartly updated metadata can bring spectacular results – conversion boosts by 18% for screenshots and by 13% for icons. Learn how Lab Cave have reached a 45.8% conversion rate growth for their screenshots.
- Use creative sets more relevant to user queries. Apple’s creative sets are a blessing as they gave way to more ad variations. Instead of a default ad, you can play around with up to 10 screenshots and 3 video previews. More than that, align those creatives with specific keyword themes or audiences, and here it comes – a highly-performing ad ready to convert!
- Localize and adjust your app metadata. If you position your app in a different region, keep in mind cultural differences. What is attractive and visually appealing to one culture, won’t resonate with another. So, try to adjust your app’s metadata to the corresponding market, as ZiMAD did for the Japanese storefront. By the way, they saw an impressive 36% conversion rate uplift.
- Improve your app’s rating. Multiple App Store users sharing their positive experience with an app… Is there a thing more encouraging to download? As many as 70% of users read at least one app review before giving it a download. High ratings and reviews push you higher in ranking as well, so go for 4+ ratings and positive featured reviews.
- Work with segments. When you use this technique to decrease the average CPT, you influence the conversion rate in the first place. For example, ads switched off from Saturday to Sunday push the conversion rate up, which results in the CPT adjustment.
As for benchmarks, we at SplitMetric have dug into the TTR values across various App Store categories. Here’s what we’ve found:
The findings show that on average TTR is slightly over 7%, while Business is the winning category with a Tap-through Rate of 11.70%. Other leaders are Stickers and Lifestyle with TTR of 11% and 9.30% respectively.
As a rule, conversion from impressions to taps is higher for Brand campaigns than for Competitors campaigns. 5% is a threshold level, and if TTR is below 5%, the ad group performs worse than it could and you need to find the reasons why.
6. Optimizing mobile advertising metrics with SearchAdsHQ
All the above metrics provide insights into the performance of your marketing funnel. Having them at hand is a must for a marketer, but obtaining them can be quite a task. With this in mind, we’ve built SearchAdsHQ – a platform which facilitates data collection and management, so that you focus on making higher-level optimization decisions.
How can you benefit from SearchAdsHQ?
1. View and optimize the entire funnel.
Spend, impressions and taps, among others, are conversions tracked by Apple. MMPs, on the contrary, record events that happen after an app is downloaded. To receive some important metrics, you have to endlessly switch between Search Ads and your mobile tracker, and then make tedious calculations.
SearchAdsHQ joins together metrics from Apple and your MMP in one interface and takes the load of time-consuming calculations off your mind. Instead of performing manual Excel tasks, you can invest your efforts into strategic planning.
2. Automate your efforts.
If you’ve got a big ad account, you definitely know how hard it is to handle it. Each day you waste an hour or two on managing bids or researching keywords. When you’ve got dozens of campaigns in multiple storefronts and thousands of keywords out there, such manual operations consume even more time and efforts.
Things are different with SearchAdsHQ automation features. All that you usually have to perform manually, can be put on autopilot using customizable rules. You can configure rules for campaigns, ad groups or keywords, and set actions, frequency and conditions for any of the metrics available in SearchAdsHQ:
There’s no need to constantly monitor your Search Ads campaigns. You can save your valuable time by creating as many rules as you want. Flexible filtering helps identify areas where rules can be applied. Here’re 10 SearchAdsHQ filters for Apple Search Ads optimization.
3. Get your metrics visualized.
It goes without saying, the graphic representation of data helps view, analyze and present data in a more comprehensible way. SearchAdsHQ charts demonstrate changes in metrics over time, correlations between metrics, as well as detect patterns to streamline your decision-making process.
In comparison with Apple Search Ads, SearchAdsHQ provides charts at any level, and they are located right in the dashboard above the table. There’s no need to switch between tabs to see the visualized metrics, while Apple has placed its charts in a separate Reports tab.
Additionally, as we’ve already mentioned, Apple tracks only pre-download events, thus you won’t find post-install metrics visualized there. SearchAdsHQ, in its turn, offers charts for every stage of the funnel, including in-app conversions.
Your Apple Search Ads campaigns can get more relevant users! Learn the keyword expansion technique and several approaches to adding new keywords to your list.
SearchAdsHQ is a platform to manage, optimize and automate your Apple Search Ads campaigns. Request a free demo at SearchAdsHQ.com.