Frequently Asked Questions
This FAQ is organized into two main categories:
- Setting up and optimizing user acquisition campaigns
- Considerations about ad creatives.
Before you dive in, here is a glossary of terms you will find throughout this documentation:
|
|
For additional terminology, visit the Unity Ads Glossary
Disclaimer: The information contained in this FAQ is provided for educational and informational purposes only. Any recommendations or guidance described herein are based on Unity's generalized experience in providing its products and services to its clientele and are provided with the intention to be helpful and to get you thinking in a more sophisticated manner about how to launch your user acquisition campaigns. The performance of your campaigns will be based on your particular circumstances and a multitude of different factors, some of which may be outside of your and/or Unity's control. If you use this FAQ, you agree that Unity has not made any guarantees about the results of taking any action recommended herein, and you acknowledge that prior results obtained by yourself or others do not guarantee a similar outcome.
Setting up and optimizing user acquisition campaigns
Think about your campaign’s structure as the easel you’ll use to start your painting. Having the correct setup, tools, and idea in mind will determine not only the result but how smooth the whole process will go.
This section will cover common questions clients ask the Unity Ads team when setting up their campaigns to help you have everything in place when it comes to setting up and optimizing your own campaign structure.
- Dashboard: 1–2 hours
- Report Builder: 24 hours for the daily data cube, up until the last hour for the hourly report
Certain elements in the Unity dashboard run under a standardized time zone – mostly universal time coordinated (UTC). Other elements use a timestamp on your local time.
Here is the breakdown:
- Campaign history and history and performance tools both use UTC.
- When you hover over the creation/uploaded date for creative-packs and creative pages, you can see a timestamp marked as local time.
- Once a campaign is live, The campaign's creation date is displayed on the campaign details page in relative time (i.e. five years ago). However, if a user modifies the start date of the campaign through the scheduling feature, then this new time will be displayed.
- Scheduling uses UTC.
The campaign budget dictates how much marketing spend the campaign could consume. Consider fixing a daily budget of $500 per country ($50 on iOS for CPM billed campaigns) and a minimum of $2,000 campaign budget. The best practice is to set the campaign budget to be, on average, at least five to seven times the amount of the daily cap to make sure that the campaign budget does not limit daily spend and that there is always enough budget left.
Campaigns with a lower budget and daily cap per country might not get any exposure due to budget restrictions. Having a low daily cap can also result in overspending during the first few days, as the system might not be able to predict the performance of your campaign accurately.
The traffic is highly dependent on competition, and we cannot estimate the scale. However, if the daily cap is low, the campaign will experience issues with scale or potential problems with overspending.
Consider analyzing device reporting at the user level for your game and using the device targeting feature to exclude devices delivering lower-end quality for the campaign. Make sure you use the different reporting dimensions available on Unity’s report builder or Statistics API to inform your decisions.
You can accomplish this by filtering by the “Device Model” or “Device Manufacturer” dimensions on Unity's report builder, then configuring a campaign to include or exclude specific devices for best performance.
- Retention campaigns:
- For passing data via a software development kit (SDK) only, you need to have a minimum of 10 retained users in that geo over the 30-day cohorts. This is counted over all installs, not just Unity-attributed installs.
- For passing data via mobile measurement partner (MMP) only, you need to have a minimum of 100 Unity-attributed installs in that geo over the 30-day cohort. The D7 retention rate over the 30-day cohort must be at least 1%.
- For passing data via both SDK and MMP, you need to have a minimum of 10 retained users in that geo over the 30-day cohorts. This is counted over all installs, not just Unity-attributed installs.
- In-app purchases (IAPs) – ROAS campaigns:
- For passing data via SDK only, you need to have a minimum of 10 paying users in a specific geo over the 30-day cohorts. This is counted over all installs, not just Unity-attributed installs. Only users making IAP through the platform provider’s store are considered paying users.
- For passing data via MMP only, you must have a minimum of 10 paying users in that geo over the 30-day cohort. This is counted over Unity-attributed installs.
- For passing data via both SDK and MMP, you must have a minimum of 10 paying users in specific geo over the 30-day cohorts.
- In-app advertising (IAA) revenue campaigns:
- Passing data via MMP only, you need to have a minimum of 200 installs (attributed to Unity) in a specific geo over the 30-day cohort. The cohort must also have generated some ad revenue passed to Unity via the MMP, but the cohort can also include users who have not generated ad revenue.
- You cannot run ad revenue campaigns if you are passing data via SDK only.
Learn more about the ROAS and Retention eligibility threshold in our knowledge base.
Unity’s most recent report, Mobile Insights 2022: Growth and Monetization, gives guidance on what locations are most popular to conduct soft launches.
Our findings suggest that most studios look for countries that have a high volume of DAU and devices for larger sample sizes or countries whose primary language is English. This makes Canada the top choice for publishers to soft launch in, followed by Australia and the Philippines.
If your game is targeting the US mainly, consider the following approach:
- For a technical soft launch where you are looking to run validations for things like MMP integrations, gameplay, or bugs, consider starting in the Philippines, India, and other countries where the English language is widely used, and the cost is low.
- For a KPI soft launch, consider smaller English-speaking markets such as Canada, the Netherlands, New Zealand, and the United Kingdom, as they are not as expensive as the US.
The allowlist will let the campaign bid specifically only on the Source App IDs targeted. Allowlists can be used to single out and target higher quality sources for your game with higher bids, ensuring you are receiving more traffic from the sources bringing in high traffic.
- For install campaigns (measuring CPI)
Keep in mind that source optimizations should be made based on matured data (recommended after at least 50 installs per source), so consider waiting to make changes after the ad creatives have passed the 2–3 day exploration phase and reach the 50 installs threshold.
For scale issues, bidding changes should be made in 20% increments every two hours until you reach the desired results.
- For ROAS and retention campaigns
Consider waiting until the data is matured, at least seven to ten days before making a bid or goal changes. Since the ROAS campaigns are optimizing towards day seven retention (D7), we suggest not making more than two changes per week, so you can evaluate the impact accordingly in the next D7 matured cohort.
There are additional considerations when setting bids on iOS for CPM billed campaigns:
- For install campaigns
The eCPI you see might be larger than the target bid. After three days (or more), eCPI will get closer to your target bid when delayed install postbacks are received from SKAdNetwork, and our algorithms have had time to optimize your campaign. Consider waiting at least three days from the last change for accurate comparisons between eCPIs and target bids before gradually adjusting your target bids.
- For ROAS campaigns
After nine days (or more), the eCPI will get closer to your max bid when delayed install postbacks are received from SKAdNetwork. That way, our algorithms have had enough time to optimize your campaign for the full install cohort for day seven (D7) post-install data.
Consider waiting at least nine days from the last change for accurate comparisons between eCPIs and max bids before gradually adjusting your max bids.
Given that the observed conversion rates are within an acceptable rate, consider increasing your CPI once each creative pack has reached at least 5000 impressions, then evaluate for 12 -14 hours. This is when our delivery system takes these impressions into account (when we need to retrain our models to ensure that they are actually in production).
Avoid taking the bid down too much from this optimized level, but take it down to a CPI range acceptable to you.
We do not offer source category targeting as a generally available product. You can, however, use the report builder to view the source categories of source app IDs.
For example, if you are looking to target the arcade category, you can copy and paste the respective source app IDs for targeting. Source targeting can also pinpoint specific sources performing well against your KPIs, such as ROAS or retention, and add them as part of targeting for best performance.
If your account is managed, please contact your Unity point of contact if you are interested in a list of the top sources per category and other metrics. If your account is unmanaged, please reach out using this email address.
You will see installs reported in a paused campaign due to the lookback window (the period after an ad is clicked or viewed within which an install can be attributed or matched to that ad). Once the lookback window has passed, there will not be any traffic coming into the campaign.
Keep in mind that the length of the lookback window varies, and you can set this in your MMP dashboard. Refer to your MMP for more information.
Here’s an easy one. The answer is: immediately.
The conversion rate can be improved by doing an ad creatives clean-up. A way to do this is by unassigning a lower-performing creative pack to have only the top-performing creatives in rotation. If the campaign is experiencing a slower exploration phase, consider adjusting the CPI to a higher level to widen the targeted audience. Once the creatives have established a stronger CVR, you can gradually adjust the CPI to lower levels.
Another option is to analyze the device, source, and category performances and adjust the campaign’s targeting accordingly to improve the conversion rate.
It is important to note that the Unity dashboard receives all of the data from the MMP (tracking provider) postbacks. Due to data delay, it can take longer for the metrics to be available in the Unity dashboard depending on which data source and metric you are viewing.
At Unity, your account can be managed or unmanaged.
- If you have an unmanaged account: Focus on 3–5 countries where you already have data on your top markets. Try to run for seven full days to catch up with weekly trends and then make source optimizations using the allowlist, blocklist, or source bid list.
- If your account is managed: Source lists can be pulled based on category or competitor’s performance via your designated contact.
You can find the Source App IDs targeted by a campaign using the report builder feature in the Unity dashboard or through the MMPs (tracking provider’s) dashboard.
Please reach out to your Unity point of contact if you are interested in targeting specific sources based on the network's performance ranking or specific source game names you would like to target using the source level bidding or the allowlist features. If your account is unmanaged, contact our support team for help.
Yes, you can add the source app ID you wish to block to the blocklist located in the “App Targeting” section of the dashboard.
On iOS for CPM billed campaigns, we recommend an initial daily budget of $50 to mitigate the risk of overspending. CPM billing with Unity enables lower daily budget caps than had previously been available. Lower daily caps at the beginning of new campaigns will help you avoid overspending on high eCPIs before real performance can be observed.
The short answer is no. However, if you set a higher bid for the second campaign with another set of creatives and targeting, it could help you increase the overall volume.
The eCPI (effective cost per install) you see might be larger than the target bid. After three days (or more), the eCPI will get closer to your tCPI (targeted cost per install) when delayed install postbacks are received from SKAdNetwork, and our algorithms have had time to optimize your campaign.
We recommend you wait at least three days from the last change for accurate comparisons between eCPIs and target bids before gradually adjusting your target bids.
- Verify your MMP settings and, most importantly, that attribution works.
- Double-check your campaign configuration
- Before launching an iOS campaign, ensure that the advertiser app is configured correctly with SKAdNetwork according to the documentation from Apple.
- Update your Unity Ads SDK to the latest version if you have not done so already. (Android and iOS)
- For launching an iOS ROAS or retention campaign, follow the Unity knowledge base article on Audience Pinpointer.
In this scenario, if the campaign has CPI-based billing, a new campaign might overspend if the CVR and CPI bids are high. High bids cause the campaign to have a strong buying power and it acquires users fast. When this happens, the information about the installs is not transmitted immediately to us so we can’t react to the spend.
In addition, when there is a new campaign, we don’t have sufficient data of its performance, making the CVR difficult to predict. After the first days, our algorithm will throttle down the campaign’s performance and try to adjust the spend closer to the daily cap.
In the case of budget-based campaigns, overspend should not happen. However, a sudden and drastic change in bids or budget may have an impact on the performance.
Consider grouping similar countries. If the target country's language is the same, put them in one campaign. That way, it will be easier for you to manage them and make adjustments.
Things get a bit more complex for iOS. Due to new limitations in the SKAdNetwork framework around removing country information, advertisers must limit their combinations of live campaigns and creative packs to 50 per app. This will help you obtain more accurate SKAdNetwork reporting by supporting the correct association of installs to campaigns and creative packs.
This change increases the need to group countries within campaigns. However, consider targeting your most important countries and regions in their campaigns for better performance and reporting accuracy.
For many advertisers, the US is a significant market. If the US is grouped with other countries and a campaign daily cap is in place, please pay close attention to the daily cap distribution. This will help you ensure that not only one geo consumes most of the daily cap and leaves the other geos unexplored during the day.
Considerations about ad creatives
Ad creatives are the nuts and bolts that can make or break your campaigns. They are also the most exciting elements of user acquisition as they allow you to tell your game’s story and attract the right people to experience it. Below, our experts have compiled a list of answers to the questions we see coming our way more often.
Different operating systems behave differently. For Android, we recommend using three to five creatives with a minimum $500 daily cap.
On iOS, consider starting with a daily budget cap of $50 to mitigate the risk of overspending. Cost per mille (CPM) billing with Unity on iOS enables lower daily budget caps than previously available.
Using lower daily caps at the beginning of new campaigns will help advertisers avoid overspending on high effective cost per install (eCPIs) before real performance can be observed. After several days, advertisers can increase daily budget caps to achieve more scale after the initial exploration.
A common explanation is that the high-scaling ad creatives have been removed. Consider keeping one or two top performing/scaling creatives alongside new creatives.
Another possibility is that the exploration phase for the game has passed, and the predicted conversion rate (CVR) was higher than the observed CVR, so the scale was adjusted accordingly.
Here’s a situation where one size doesn’t fit all. The best practice is to test different ad format options and video lengths to see what works best for your game’s genre.
Consider the following steps:
- Create a new campaign (do not start it yet). Assign a budget and set the targeting relatively wide and the CPIs higher than usual CPIs. We suggest starting with 20-30% higher.
- Upload your new creatives.
- Add all the new creatives only to the new campaign.
- When all creatives are approved, start the campaign. The new creatives are being seen by the system exactly at the same time and get the same starting conditions.
- Once the creatives have been explored, pause those from the test campaign and assign them to run in your main campaign only.
The test creatives campaign will help hasten the exploration phase of the creatives. Once the learning of the creatives has been established, you can add these to the main campaign. Please note that the exploration phase of the creatives has been completed once a creative pack has received at least 5K impressions.
Another approach if you are targeting different countries is to run your ad creative test in locations where the cost is low.
Consider targeting tier three countries with a lower cost. Test creatives with different variations/lengths, and check if you have too many creatives with a low daily cap. Make sure that the campaign targeting tier three countries does not include an existing creative that has already performed well in the past.
Ensure your campaign targeting is left wide enough to capture a bigger audience. Check if the campaign has used all of its budget, or the daily cap is being spent repeatedly. Consider increasing the campaign budget or the daily cap for more volume.
Video performance varies by app and game category. The best practice is to include both orientations of a creative within the same creative pack. Consider testing all ad types and orientations to find out what works best for you.
We measure creative learning in terms of impressions. Once a creative pack has reached at least 5K impressions, the creative pack is considered to have passed the initial exploration period. Typically this takes a few days, but this process can take a few hours or several weeks.
This period is defined by the bids you set for your campaign and the exploration phase might take longer if the bid is not competitive enough.
It can be difficult for new creatives to be explored as older creatives have an established learning and stronger CVR. As the creatives compete with each other for traffic, we suggest following the workflow:
- Create a new campaign (do not start it yet). Assign a budget and set the targeting relatively wide and the CPIs higher than usual CPIs. We suggest starting with 20-30% higher.
- Upload your new creatives.
- Add all the new creatives only to the new campaign.
- When all creatives are approved, start the campaign. The new creatives are being seen by the system exactly at the same time and get the same starting conditions.
- Once the creatives have been explored, pause those from the test campaign and assign them to run in your main campaign only.
The test creatives campaign will help hasten the exploration phase of the creatives. Once the learning of the creatives has been established, you can add these to the main campaign. Please note that the exploration phase of the creatives has been completed once a creative pack has received at least 5K impressions.
All new creative packs undergo a human moderation process when their associated campaigns go live for the first time. Creatives can be rejected due to content or technical checks.
If the creative was rejected due to its content, please read our policies and adjust the creative accordingly. You can then upload the new version through the self-serve dashboard.
If the creative (usually playable assets) was rejected due to a technical problem, please test the creative with the test app.
To understand more about our moderation processes, please visit our documentation page.
Yes, you can test different combinations and get the reporting on the creative pack level.
We measure creative learning in terms of impressions. Once a creative pack has reached at least 5K impressions, it is considered to have passed the initial exploration period.
Exploration and performance are always contextual. For example, if a creative pack has been explored in India, our algorithm does not know how well it would perform in Sweden. This is why “exploration” is always shown when a creative pack gets assigned to a new campaign. This is also why a creative pack can have a different status in different campaigns.
However, remember that our algorithm uses any information available, so knowing how well the creative pack performs somewhere (or with some audience) is better than not knowing anything at all. This means a creative pack that has already been explored in one campaign usually goes through exploration much faster when assigned to a new campaign since our algorithm does not have to start from scratch.
This is why we recommend setting up separate campaigns with higher CPIs for exploring creative packs; the more similar the audience is to the “actual” campaign you intend to use the creative pack in, the better the exploration will be.
At the campaign level, there are no limits on the amount of creative packs you can have. However, there is a limit of 700 per advertised app. Keep in mind your iOS app is considered separately from your Android app, so each one can have up to 700 creative packs.
However, for iOS, due to new limitations in the SKAdNetwork framework around the removal of country information, advertisers are required to limit their combinations of live campaigns and creative packs to 50 per app. This will help you obtain more accurate SKAdNetwork reporting by supporting the correct association of installs to campaigns and creative packs.
- Video and end-card assets: 12–24 hours
- Playable assets: 24–48 hours.
Remember that seasonality affects the volume of creatives in the moderation pipeline and the service level agreements (SLAs).