Making Full-funnel Programmatic Work for eCommerce Apps in The Privacy Era
This is the first article on our latest content partnership with Singular. Here we cover how programmatic advertising can support Shopping apps growth on both LAT and no-LAT traffic.
Programmatic is all about approaching the right audience with the right ad at the right moment. For eCommerce apps, that is an excellent way of engaging their user base and maintaining healthy sustainable growth.
Here we look at how programmatic advertising can support eCommerce apps to grow and thrive, despite the new user privacy paradigm.
🔹 eCommerce is running high. Revenue in this market is projected to reach US$2,723,991m in 2021 and to grow at an annual rate (CAGR 2021-2025) of 6.29%, resulting in a projected market volume of US$3,477,296m by 2025.
🔹 Fashion is still one of the most significant segments in the eCommerce market, and China still shows up as a powerhouse when it comes to revenue generated. By this year, the average revenue per user (ARPU) in eCommerce is expected to amount to US$714.11.
What does this mean?
It's worth highlighting that there's apparent user interest in this vertical, but all this demand doesn't simply happen organically.
We can see this demand as an increased opportunity for apps in this space (to continue to grow in size and revenue) and as a competitive challenge (as apps grow bigger and new players enter the market). And to stay on top of both opportunity and competition, programmatic advertising plays the role of growth enabler.
How the industry is shaping up post-iOS 14.5 and how can marketers overcome the challenges
iOS 14.5 and 14.6 adoption is still ramping up but its values are historically low. At the end of May, only 21.6% of the devices had updated to the newer version. ATT prompts to opt-in are being slowly adopted by apps now. The seismic announced changes are happening slowly, giving time to the different players in the industry to adjust to a new privacy reality.
It's unmistakable that user privacy rules are very welcomed, but their effects on app advertising can't be ignored. Slow or not, it’s time to embrace the changes and find new frameworks to continue to deliver top performance to app businesses.
With LAT (Limited Ad Tracking) and through programmatic advertising, performance is very much still achievable for eCommerce businesses. Continued attainment requires reframing strategies and long-term planning. Let's dive into how programmatic UA and re-engagement look like in the post-IDFA for apps in eCommerce.
UA and retargeting for Shopping Apps on the post-IDFA Era
Making programmatic work for your eCommerce app involves maintaining a good balance between acquiring new users and reengaging existing or lapsed ones. This is the recommended way of keeping a healthy and profitable user base, one that is engaged, generates conversions and revenue. This full-funnel approach requires a balanced budget spend between UA and re-engagement campaigns to increase users’ LTV and improve the app's overall profitability.
This overarching strategy works successfully in a deterministic scenario where user identifiers are still available. Without user IDs to work with, the priority becomes to adjust your programmatic strategy to continue driving real advertising impact.
Considering that app marketers operate in these two scenarios, with and without user identifiers, approaches to programmatic strategy might differ while the overall goal of driving incremental growth for eCommerce still stands.
Let's look at how to plan and implement a programmatic plan that works on both scenarios.
Structuring campaigns on LAT and non-LAT
Reframing your approach to programmatic will involve running blended programmatic campaigns for your shopping app designed to bring you incremental value independently of the specific target audience - LAT as well as non-LAT.
At RevX, we call these campaigns PAI (Programmatic App Impact) campaigns. For Shopping apps, these campaigns are run holistically, with a more encompassing goal of generating awareness, acquisition, and re-engagement to capture all of your ads' positive effects.
Programmatic campaigns on LAT traffic
In a LAT context, deterministic retargeting campaigns that require using feed events to target users with personalized offers aren't possible to run. Retargeting campaigns can only be run for users who opt-in on iOS. Otherwise, there aren't any other attribution and targeting methods that would work on a LAT scenario to run deterministic retargeting. But you can still very much run re-engagement campaigns in a blended setting.
Campaigns on LAT traffic demand a distinct strategic approach. The focus of these campaigns is shifting from being firmly CPI/CPA and ROAS focused to adopting a more branding-oriented structure. No longer can advertisers rely on IDFA to target, optimize, and attribute the results of their campaigns.
User acquisition needs now to rely on different attribution methods like SKAN and uplift measurement to get accurate results.
With only aggregated data available, the inability to carve out very defined segments and the lack of deterministic attribution are taking us to rethink the business models that have worked so far. Given that the data available to run and analyze campaigns is different now, app marketers need to embrace a different strategic approach to campaign structure, measuring their efforts, and optimization.
If until now your retargeting strategies relied heavily on identifying active buyers with high volume purchases, that will be far harder to do going forward. Data at the user level and personalization won't be nearly as effective to guide your campaign strategy.
For eCommerce, this is particularly relevant because these privacy changes mean that we no longer have access to user-level data telling us which products a specific user has browsed or purchased. Advertisers lose in LAT the capacity of targeting specific users with highly personalized content ads.
Therefore, retargeting campaigns on LAT users will become part of a broader ad campaign approach encompassing branding and user acquisition. Your budget allocation is subsequently applied to fit "holistic mobile advertising" and reminds us of old-style branding campaigns.
Running programmatic advertising on NON-LAT
For non-LAT traffic, where user identifiers are still available, the app marketers' job is simple: running a campaign strategy that considers both user acquisition and retargeting campaigns to enjoy the advantages of driving results for an app's entire funnel.
Why run retargeting alongside UA campaigns? The data proves that it is a successful combination.
eCommerce apps benefit from a growing market and should use their first-party data to segment and identify opportunities to reengage users. Retargeting is a great way to promote and push products and drive users to purchase. And there are different ways you can structure your retargeting campaign with specific conversion goals.
A one-size-fits-all approach to targeting doesn't work. Start by digging into the data and identify where your most profitable users stall during their app journey to reach them effectively.
It is known that retention rates are higher for retargeted users during the first seven days post-install than new users that haven't been retargeted. And retargeted users are prone to drive more conversions than non-retargeted new users. In addition, engagement rates of retargeting campaigns are 152% higher than those of new user acquisition campaigns.
So, although acquiring new users is the first priority for app growth, re-engagement campaigns sustain that growth and are useful for shopping apps at every step of the funnel.
Re-engagement campaigns can be structured based on the segments you wish to target and what goals you intend to achieve with each campaign. For example, a helpful approach would be to retarget new or lapsed users to complete their first purchase within the first-week post app install. The same method can be applied to buyers to get them to repurchase by offering more relevant products or specially catered offers to loyal, high-value app users.
Conversion funnels can look very different depending on the nature of each eCommerce app. Shopping apps such as offering FMCG and daily consumption items like groceries have a faster-converting path; applying a broader targeting strategy to campaigns works in these cases, so advertisers don't have to sacrifice their ROI on media.
For apps with slower converting funnels such as fashion or electronics, the basket sizes can vary broadly, and users also might present very different types of behaviors depending on their purchasing intent. For this scenario, slicing and dicing your audience to create a more specific targeting works best to achieve exponential results.
The roles of attribution, SKAN, and contextual targeting
The tactics described above are worth applying to non-LAT users and can be run on traditional attribution models.
When it comes to LAT or blended traffic, advertisers will need to rely heavily on supply signals and need to make the most out of the available data to run their programmatic campaigns. Since we lack specific information regarding homogeneous and detailed information for our bid requests and impressions, it's important to shift our focus to what we can do with the contextual data that we still have available.
Advertisers will rely on probabilistic targeting based on contextual data and information coming through bid requests. This requires having the support of a programmatic partner to analyze all this data available. The data referred will allow for a certain level of personalization and enable targeting similar audiences without disclosing who a user is.
What happens to deep linking?
Direct deep linking itself will not be impacted by the privacy changes because the context of the ad engagement and its linking happens directly into the app that's already installed by the user. What is recommended is to use links with app store fallbacks as it would not be known whether the user has the app installed or not.
What is not possible anymore is custom deep linking based on personalized ads. As we understood, there is no matching of users’ surfing history possible, which dynamic content ads are dependent on. Hence, most of the time either the home page or a particular promo page in an app would be the desired landing locations. Product development for the app has to collaborate more closely with marketing to make sure there is no gap in the user experience.
Does creative personalization still work?
Creatives merge your campaign messages with selling your product or services. More than ever, and without IDFAs, ad creatives need to be built innovatively and efficiently to capture users' attention. Personalizing your creatives is always an effective and attractive way to gather users' attention.
On LAT, running dynamic content optimization using asset feeds and A/B testing, will require adaptation. Product feeds won't be useful anymore as a way of catering creatives to particular user behavior in a 1:1 setting, but possibly as a group setting based on ad environment context.
Different target segments in different markets will react differently to creatives, so experiments with different styles adapted to local preferences and traditions would be a way to gather insights on campaign impact.
When it comes to communicating a message in your ads, the focus lies in communicating with users to add value and drive them to convert. This can be done successfully both in LAT and non-LAT environments via a well-thought-out targeting strategy with the information available, a programmatic platform that can support you with optimized ML capabilities, and the right creatives.
Combining these elements is the not-so-secret recipe to offering a strong ad experience to users and achieve your performance goals.
Measuring your efforts effectively.
Both on LAT or non-LAT, the ability to tie advertising spend to business impact is crucial.
SKAdNetwork, the only MMP for LAT campaigns on iOS, provides partially deterministic attribution for installs because performance can be attributed to channels but won't reveal the respective user that has converted.
If you wish to measure the ad campaigns' performance on iOS in a more comprehensive fashion, you will have to implement uplift measurement based on time-series analysis and synthetic control groups.
Be prepared to optimize your app marketing strategy further by looking for the incremental value brought by each campaign, and gather tangible insights throughout your campaigns.
Fighting fraud on all fronts & how fraud post-iOS14.5 can look like
Ad fraud is expected to morph and rise with LAT traffic. User identifiers were used in fraud protection and blacklisting. As that traffic now becomes unavailable, we need new ad fraud mechanisms in place that can use supply information to verify if impressions are genuine or not.
Partners such as Singular are prepared to support advertisers on this.
In the words of Ron Konigsberg, CGO at Singular:
“Fraudsters continuously strive to find vulnerabilities they can exploit, and it is clear they’re working hard to reinvent fraud methods relevant for SKAdNetwork.
We already identified different weaknesses with SKAdNetwork, which require protection methods we provide, including SKAdNetwork postback signature verification, replay attacks reusing genuine postbacks multiple times, and postback manipulation altering fields which Apple does not sign.
These attacks are only the beginning. New fraud methods will continue to surface over time, so advertisers, MMPs, and now Apple will need to stay one step ahead of fraudsters to mitigate the threat.”
Actionable steps for eCommerce apps: creating impact with programmatic
Despite all the privacy changes, programmatic campaigns can still be structured to perform at your expected results. Advertisers should carefully assess their audiences and long-term goals to path a new way forward for their mobile advertising investment.
User acquisition is vital at the start of your advertising journey but shouldn't be your only focus. Keep investing in app re-engagement that gives you better chances of incremental uplift on purchases.
Ad campaigns for eCommerce tend to generate more revenue when running along with seasonal deals and promotions. Seize the opportunity and concentrate your budget spending in these periods to match users' intent to make the bulk of their purchases.
With the eCommerce industry being so diverse and competitive, and with so many parties involved, advertisers should lean towards partners that can offer more streamlined services to capture and analyze what data is available and use it to address target audiences.