However, according to new research, a third of CMOs (33%) are more focused on the impact of the rising number of channels and platforms than they are on increasingly complicated consumer behaviour (17%).

Findings from Adverity suggest that related growth in data volumes may be a significant cause behind this imbalance of priorities.

Titled ‘Recession, Resilience, & Marketing Data: What’s shaping the CMO’s roadmap?’, the report surveyed 300 CMOs across the US, UK, and DACH region from small to midsize businesses. It uncovers key challenges for CMOs in the current climate of accelerated change, digitalisation, and uncertainty, and results indicate data handling issues are hindering efficiency.

While 85% of CMOs agree the capacity to make data-driven decisions is a critical competitive advantage, many are struggling to ensure effective data management. This is perhaps explained by the fact that almost seven in ten (67%) feel the volume of marketing data available has become overwhelming, with 99% using 10 or more data sources and 52% using 14 or more.

It’s difficult to remain hopeful when these findings suggest many businesses are at risk of missteps as the global economy declines, with unwieldy data preventing smart pivots, and distracting CMOs from the need to align experiences with shifting consumer needs.

What about budgets?

Although 29% of CMOs are planning to enhance internal capabilities by increasing their data operations budget through 2022, this push towards increased spending doesn’t translate across the board. Nearly as many (27%) say their budget has decreased and the majority (44%) are working with the same spending pot as last year, with those identifying as ‘least resilient’ to ‘black swan’ events in recent years that have been the likely cause of tightening belts.

It’s also interesting to learn that the most common blocker to investment into data tech is lack of knowledge within the business (27%). This is closely followed by the perception that adopting new tech will be too complicated (23%), while only 8% indicate the main barrier to investing in tech is securing budget.

According to Harriet Durnford-Smith, CMO of Adverity, CMOs have become overly bogged down by data challenges and lost sight of their core purpose: meeting the needs of the consumer.

She commented: “How they got here isn’t hard to understand. As channels and platforms have expanded, so has the quantity of incoming data about audience interactions and marketing performance. Failing to implement efficient systems for managing this data, however, means they are now wary of smart tools and battling to bring disordered information into order, when they should be using it to inform efforts that create consistently strong consumer connections.

We don’t know how long this financial instability and uncertainty will last, so it is for the best that CMOs are rapid in their movement to fix any problems surrounding data. Only with a unified supply of accurate insights can they pinpoint the best path to driving relevant, meaningful, and empathetic conversations, which fuel long-term bonds and trust.

The post Recession, Resilience, and Marketing Data – What is Shaping the CMOs Roadmap in the Midst of Financial Crisis? appeared first on PerformanceIN.

The post How Regulatory Compliance Will Protect Your Brand in the Wild West of Programmatic and Affiliate Marketing appeared first on PerformanceIN.

SKAdNetwork was created as an alternative to Apple’s Identifier for Advertisers (IDFA); a familiar advertising ID attached to a user’s device. IDFA allows advertisers to track user interactions to attribute them to a campaign ‘deterministically’. The initial rollout of SKAdNetwork in March 2018 had no immediate impact on the mobile marketing industry, with advertisers continuing to use IDFA as normal, blissfully unaware of what was coming. However, in June 2020, in the context of rising consumer awareness around privacy and the value of personal data, Apple shocked the mobile advertising industry by announcing major changes to how advertisers could access its IDFA. 

The most seismic change was in the form of the introduction of app tracking transparency (ATT). This would require users to opt-in to share their data with advertisers every time they opened a new app. In doing so, they would allow advertisers to continue using familiar deterministic attribution with IDFA. However, when ATT was first introduced, estimated opt-in rates were reported to be in the region of 15%. 

WTF actually is a SKAdNetwork?

Starting from the beginning, the ‘SK’ is shorthand for ‘StoreKit’, Apple’s developer framework; and the technology is an aggregate for measuring mobile ad campaigns for iOS apps. Put (relatively) simply, it’s a privacy-friendly framework designed to provide advertisers and ad networks with aggregated campaign performance data about acquired users, without revealing any user level details. Because it is privacy-centric at its core, it has a range of limitations when compared to the previously dominant IDFA and deterministic attribution. There is also an additional layer of complexity for advertisers that want to comprehend the framework’s data.

SKAdNetwork differs greatly from IDFA and deterministic attribution in that it doesn’t reveal user or device-level data to attribute impressions and clicks to instals of iOS apps. Contrary to popular belief, the IDFA hasn’t been ‘killed’ or ‘ended’, but ATT does now require users to permit advertisers to use IDFA, whereas previously it would work in the background to track users automatically. If iOS users opt-in when they open a new app, advertisers can then track them and access their device ID.

If the user refuses to share their data, the advertisers must rely on more subtle signals under the anonymised SKAdNetwork, or use frowned-upon fingerprinting methods. Fingerprinting attribution runs the risk of advertised apps being rejected from the App Store in the future. Apple calls out fingerprinting as not permissible in their guidelines, but has not clamped down on it yet. SKAdNetwork doesn’t carry such risks; it works for both opted-in and opted-out users while protecting the privacy of each. 

What has its impact been? 

With a 24 hour data privacy threshold and no real-time marketing events available, SKAdNetwork made it incredibly difficult to determine where users were coming from, what their value was, and what marketing campaigns were working. This was especially true at the onset, when SKAdNetwork data was being passed from Apple to ad networks, rather than directly to advertisers themselves – a situation that was resolved in September 2021. From iOS 15 onwards, advertisers have received copies of the ‘postback’ (data) that is sent to ad networks, but only a minority of advertisers are currently making use of this.

That’s because, even with unfettered access to SKAdNetwork data, publishers required commitment from development teams to support it, and data scientists to make sense of it. They were no longer able to lean on traditional mobile measurement partners (MMPs) for install data, standard analytics (DAUs, sessions, revenue, retention), LTV metrics, ROI and ROAS, as they had previously. 

This has driven many app developers and publishers – most (68%) of whom don’t have access to in-house data science – to more seriously explore automating their marketing processes such as the collection and storage of data. Manually collecting, importing, and analysing vast, disparate, anonymised SKAdNetwork data sets is enough to grind any marketing team to a halt. 

What are the benefits? 

Can there be benefits for the mobile marketing industry if it can’t show targeted ads to the majority of users? Potentially. Let’s wake up and smell the coffee; the majority of users, particularly in privacy-aware European markets such as Germany, aren’t going to opt-in to ATT. This makes SKAdNetwork, which functions for all users regardless of their opt-in status, the only white-hat method for advertisers to optimise performance for 100% of iOS users.

If SKAdNetwork does reach mass adoption, it could therefore represent a reset in the relationship between users and advertisers in the long-term. However, given the choice, the vast majority of advertisers still use deterministic attribution for opted-in users, delaying the short-term pain of getting to grips with SKAdNetwork. 

SKAdNetwork, with Apple as the ultimate arbiter, is changing the relationship between advertisers and self-attributing networks (SANs) like Facebook and Snapchat. These networks can no longer limit access to their data and are treated in the same way as smaller ad networks.

Additionally, ad networks have hidden their sources of traffic (app names) from advertisers, but under SKAdNetwork, this is no longer possible. One of the SKAdNetwork fields shared with advertisers is ‘source-app-id’ which corresponds with the app ID in the App Store – giving advertisers much greater clarity. 

Another major upside of SKAdNetwork is the potential to cut down on fraudulent attributions by raising the technical bar for fraudsters. Previously, bad actors could work the system with relative ease and utilise hidden adverts, fake ads or fake clicks because there was no ‘middle-man’ to verify information between user and publisher. SKAdNetwork improves security by adding a cryptographic signature to each ad impression verifying the integral data, without compromising any information on the user who viewed or clicked the ad.

Other improvements include measurement of the performance of web-to-app ads which direct to the App Store, which will now be attributed to the SKAdNetwork, a feature previously left out which meant many users fell through the cracks. With these updates, Apple potentially also gets closer to ending device fingerprinting, a practice it has always been clear it doesn’t allow, but may have tacitly permitted while it gets SKAdNetwork’s full suite of features online.

WTF happens next?

In some ways, SKAdNetwork has been something like a live service video game that was released before it was really ready. It is improving, though. As a part of the recent SKAdNetwork 4.0 update Apple announced it would lift the cap on the amount of data advertisers could pass with campaign ID, in addition to other gradual improvements which will see the framework roll out all features and reach its full potential, at scale, at some point in 2023.

One major impact of SKAdNetwork was a redistribution of ad spend away from iOS towards Android – both Google Play and alternative global app stores such as Huawei AppGallery and Samsung Galaxy Store. This is creating a more diverse market, with more choice for developers and advertisers outside of just iOS and Google. Likewise, Google has responded to Apple’s privacy-first changes with some of its own, announcing an expansion of its Privacy Sandbox to mobile and potentially signalling the end of the Google Ad ID within the next two years. 

It remains to be seen whether a more diverse, competitive market is the long-term direction of travel. Slowly but surely, Apple is refining the SKAdNetwork to better capture data for advertisers while protecting users. It is certain, though, that the long-term direction of mobile advertising is privacy-first and, for those advertisers still mourning the IDFA, it’s time to move on to SKAdNetwork. 

The post WTF is a SKAdNetwork? appeared first on PerformanceIN.

Some even define the target at a business level. For example, we often hear phrases like:

“Our Google ROAS should be more than 150% or Our paid search campaign CPA should not exceed $20.”

This is one of the biggest mistakes we see businesses make; it leads to improper marketing mix optimisation and losing money for the business.

I would say with confidence that it is technically impossible to understand ROAS/CPA at a campaign or channel level. Yes, you can apply different attribution models but all of them are wrong and none of them will show you the real ROAS/CPA of the channel or campaign.

The only ROAS/CPA you can calculate for sure is the ROAS/CPA of your whole marketing mix. And this is the only ROAS/CPA you should use as a target for your marketing mix optimisation.

Let’s take a look at a very simple example of the marketing mix where you have only three channels: Google/cpc, Google/organic, and direct/none, and the monthly statistics using the last non-direct click attribution looks the following way:

Let’s say the CFO is happy with the current results of marketing team performance. The overall ROAS of the marketing mix is 400% and it provides a good unit economy for a business.

CFO is happy with the 400% ROAS of the business. And is ready to invest as much as possible within this ROAS. This is the TRUE KPI.

Taking this information into account, an inexperienced marketing manager might say that their target ROAS for Google/cpc is 200%. But this will be completely incorrect and may lead to horrible results.

Imagine, the client connects SegmentStream and replaces pixel conversions in smart-bidding campaigns with predicted conversions. This way Google campaigns start targeting not only users that convert within the same device using cookies but also users that are likely to convert from other devices and later come as Google/organic or direct/none. The next month the client gets the following results.

Imagine a marketing manager being furious as Google/cpc ROAS drops from 200% to 180% and makes the decision to switch off SegmentStream optimisation. But what actually happened? As a reference, the marketing manager used a vanity metric of a channel ROAS, which doesn’t have anything to do with reality.

The reality is that overall marketing mix ROAS increased from 400% to 500% and optimisation helped bring $1000 of additional revenue with the same budget.

Hopefully, the team manages to educate the marketing manager and persuades them to change target ROAS for google/cpc to 180% instead of 200%. But is the marketing mix optimised? Not quite yet.

As we remember, the CFO confirmed that the marketing team can invest as much as possible as long as the ROAS of the whole marketing mix is at least 400%.

This means that we can push our target ROAS for google/cpc even lower! And after some incremental changes in a few months we might come to the following marketing mix:

This way, we increased investment into Google/cpc from $1000 to $1500. This led to a drop of google/cpc last non-direct click ROAS from 180% to 160%, whilst the overall marketing mix ROAS became 400% and overall revenue increased from $5000 to $6000.

This way, by applying proper marketing mix optimisation, we achieved $2000 revenue increase (+50% growth) while maintaining the same ROAS of the marketing mix of 400% even though google/cpc vanity ROAS metric decreased from 200% to 160%.

That’s a lot of stats! Here are the most important conclusions and key ideas for you to take away:

• Per-channel and per-campaign ROAS/CPA you see in advertising platforms or Google Analytics have nothing to do with real ROAS/CPA.
• There is no technological possibility to calculate real per-channel and per-campaign ROAS/CPA due to cookie-tracking limitations and cross-device interactions.
• The only true ROAS/CPA you can calculate is the ROAS/CPA of the whole marketing mix.
• Target ROAS and target CPA bidding strategies inside the advertising platforms should only be used as a means of budget limitation and should be revised on a monthly basis depending on the whole marketing mix performance. There is no ideal static target ROAS/CPA for the marketing campaign. Target ROAS/CPA can be decreased/increased depending on the overall marketing mix performance.

The post How to Properly Use Target ROAS/CPA Smart-Bidding Strategies appeared first on PerformanceIN.

Businesses, however, need to make the move now as the current version of Universal Analytics will cease to collect any data by the end of June 2023. For brands who may be reluctant to migrate to GA4, they will lose their year-on-year data. Transition timelines will vary from company to company based on the size of the business, how many properties they have, and their reliance on data-driven decision making. 

Making the move in three simple phases

It’s clear that the sooner a business starts its transition to GA4, the more confident its team will feel working with the new platform. The first step involves planning and implementation, focussing on creating the Google Analytics property and configuring the base tracking across a company’s digital properties and apps. Once this has been established, marketers will need to set up platform links between their Google Marketing Platform, Ads, Optimise and BigQuery which is Google’s Cloud Data Warehouse, before testing the initial implementation of GA4.

Following this, brands will need to start establishing the new GA4 capabilities and rebuilding any reporting and data pipelines for the new platform. Marketers can also begin to replicate any Universal Analytics audiences and segments, as well as implement any enhanced e-commerce tracking. It’s vital at this stage to begin training the wider team to help them understand how best to leverage GA4 and its advanced capabilities. 

To fully understand the platform and its capabilities, it’s important for organisations to begin the final stage and trial the new GA4 features. In addition, marketers should undertake regular privacy audits to ensure that their brand adheres to any applicable data protection laws. GA4 is constantly evolving, so businesses need to ensure that they can quickly react to new changes and provide relevant training to help teams upgrade their skills.

Taking this all into account, it is recommended that brands implement a dual set-up, whereby GA4 is deployed on the web alongside the existing Universal Analytics property. When this process is completed, an organisation’s website measurement data is sent to both the Universal Analytics and GA4 property. This approach ensures that marketers have sufficient time to transition reporting before the platform sunsets.

Bringing benefits to brands’ digital assets

Whilst the upgrade may seem like a daunting process, GA4 will bring a host of new benefits to brands. Predictive Audiences is set to allow marketers to forecast when users will churn, convert and uncover lifetime value, which goes far beyond retargeting audiences. By predicting the purchase behaviour of users in the next one to four weeks, businesses have better opportunities for cost-effective paid campaigns and improved return on investment (ROI). Furthermore, Data-Driven Attribution Modelling will become available for non-premium properties. This move to a modelled approach to attribution takes into account each of the touchpoints on a website’s conversion events and ensures that a brand’s advertising spend works harder.

Businesses will also benefit from On-site Optimisation Integration with Google Optimise. This means that experiments can be run using user metrics instead of using the session scope of Universal Analytics, ensuring that a campaign is targeting a group of users who exhibit certain behaviours and attributes. Additionally, BigQuery exports will become available for non-premium properties. Previously, GA360 was needed for brands to automate data exports into BigQuery, but this can now be seamlessly achieved in GA4, opening up a new world of reporting, insight, and modelling capabilities.

GA4 is a privacy-centric platform as Google understands that it is vital for consumers to take charge of their data in order to build mutual trust. The new platform has been built with privacy in mind due to a selection of privacy features that help brands give control back to their customers. For instance, GA4 anonymises user IP addresses by default and can be used with consent mode to make tracking opt-ins easier and more reliable.

Aligning GA4 with performance marketing

It’s no secret that GA4 is going to revolutionise the world of performance marketing and open up a plethora of tools that will bolster future business strategies. Google’s integration with machine learning and AI will allow marketers to make better-informed decisions about their plans for campaigns and content. The platform can track trends and calculate advanced metrics that had not been previously tapped into. As a result, brands will be able to utilise these insights to ensure that they are making the most of their marketing budgets. 

Traditionally, analytics has been restricted to devices or platforms that limited the end-to-end understanding of a buyer’s journey. GA4 can help businesses fill in these gaps and truly gain the knowledge related to how individual customers relate to their website. This customer-centric approach means that brands can better adapt to consumer needs and make updates to the site for an improved online experience which aligns with user demands.

The post GA4: Why the New Platform is Set to Revolutionise Marketing appeared first on PerformanceIN.

This is with the intention that, by the time the end of cookies finally rolls around, the industry will have had more time to develop and test cookieless alternatives.

The statement reads: “The most consistent feedback we’ve received is the need for more time to evaluate and test the new Privacy Sandbox technologies before deprecating third-party cookies in Chrome. This feedback aligns with our commitment to the CMA to ensure that the Privacy Sandbox provides effective, privacy-preserving technologies and the industry has sufficient time to adopt these new solutions.

“This deliberate approach to transitioning from third-party cookies ensures that the web can continue to thrive, without relying on cross-site tracking identifiers or covert techniques like fingerprinting.”

Kate Jervis, Growth Director, Incubeta commented: “While Google’s decision to delay the deprecation of third-party cookies undoubtedly gives the industry more time to prepare, the end of these identifiers is still very much on the horizon and it’s more important now than ever before to prepare your infrastructure for a privacy-first world. In a similar vein, marketers should look to prioritise moving to GA4 in order to protect their future data assets.

“Activations such as complete campaign audits and privacy health-checks, building websites to be capable of tracking a similar level of data to what is gained from third-party identifiers, and focusing on contextual data over demographic data will put businesses in good stead ahead of 2024. Taking responsibility for users’ data and creating a strong data-driven strategy that can handle the change when third-party cookies and Google Analytics are gone should be prioritised now.”

We released very similar news in June 2021, when it was announced that cookies wouldn’t be phased out until late 2023. Again, this time, the announcement is both positive and negative. On one hand, it enables further creation and development of more advanced alternatives, rather than marketers having to take a rushed approach.

However, on the other hand, it once again delays meaningful redesigning of the open web to a more privacy-centred model. What do you think? Is this delay welcome, or is this going on too long?

The post Another Delay – Google Pushes Killing Off of Third Party Cookies Until 2024 appeared first on PerformanceIN.

There’s no doubt that this move, which follows in the footsteps of other browsers, including Apple’s Safari, is greatly challenging for the digital media industry. In fact, we should have been waving goodbye to cookies over the next few months, but the complexity of the project last year led Google to delay the phase-out until late 2023.

The battle for attention

At the same time as this technological challenge, brands are facing a new battle for attention. The growth of digital media has been a double-edged sword for brands, with ever greater numbers of small and medium businesses advertising for the first time. While that’s happening, consumers are enjoying more ways than ever to spend their time. From virtual reality to unlimited streaming services, the media industry continues to offer ever greater choices for consumers to spend their time – both commercial and ad-free.

It seems like a perfect storm. But while the end of cookies marks the death of the old world of digital advertising, marketers shouldn’t panic. Advances in alternative ad technologies – often making use of artificial intelligence – will enable advertisers in the post-cookie age to make their campaigns more effective, without the privacy headache of cookies.

An alternative solution

A number of new approaches to allow programmatic advertising to continue without cookies are in the works. But it is contextual signals, based on anonymous interest cohorts, that are quickly becoming the best data point to maximise metrics like attention. 

Contextual signals don’t make use of personal data, or track users between websites. That means they avoid not just the privacy concerns of cookies, but also the issue of being shown the same ad repeatedly. This is a major aspect of “bombardment”, identified as a key threat to public trust in advertising in a 2019 report by the UK’s Advertising Association.

Instead, contextual signals use non-user-specific metadata from bid requests. As a campaign runs, this data sheds light on which ad buys are most likely to lead to conversions. But to manually analyse this data and make budget allocation adjustments as circumstances change is a Herculean task for anyone. The complexities of digital media buying also result in campaigns that are challenging to scale, due to the need to monitor increased numbers of interdependent items. 

AI in action

This is where AI comes in. With AI, advertisers can build a bespoke learning data set for any given campaign, which then evolves as it learns. As a result, it can make choices that get the best results from a series of variables, therefore delivering the strongest performance. The AI can automatically reallocate budgets in real time, making it much more achievable to scale a campaign. 

This approach means that rather than over-targeting a predetermined set of users (remarketing, for example), a campaign can continually move towards buys that are most effective in capturing the attention of users – leading to greater conversion rates and boosting return on ad spend. 

AI can also make use of other sources of data, such as a brand’s opted-in customer relationship management data or measurement and attribution data, and incorporate custom metrics beyond ROAS (Return On Ad Spend). This gives advertisers greater control of their advertising and allows them to plan campaigns to meet specific business objectives.

A stronger digital media ecosystem

For example, luxury beauty brand Charlotte Tilbury recently utilised an AI designed to be customisable to the digital marketing needs of the brand. They faced problems common in the luxury market, where the relatively small number of (high-spending) consumers means ad campaigns often generate fewer higher value conversions than in other sectors. That then leads to a lack of data that can inform and improve campaign performance and scale. 

Because of this, the AI solution took into account data that is generated in abundance but which legacy approaches to optimisation are unable to utilise: upper funnel signals such as clicks, site visits and ad viewability. The results were significant: a 60% growth in the conversion rate of the brand’s acquisition campaigns, and a 29% drop in average cost per acquisition.

Marketers don’t need to shed a tear for the demise of cookies: the capabilities of AI, when used with contextual signals and first-party data, will let them measure and optimise the performance of their campaigns more effectively than ever before. Along with positive developments in areas such as optimising digital creative, this means a digital media ecosystem that offers better results for advertisers and a better experience for consumers.

The post From Contextual to Attention – How AI is Changing Success Measurement appeared first on PerformanceIN.

Thankfully, FLEDGE API, one of Google’s major proposals, is shaping up to be a positive vision of the future, and Łukasz Włodarczyk, VP of Programmatic Ecosystem Growth & Innovation at RTB House, a key expert operating the Origin Trials, corroborates this claim…

The post Google’s Origin Trials: FLEDGE Testing Would Benefit From More Input by SSPs appeared first on PerformanceIN.