Yes, we know: programmatic trends & predictions are among the TOP-5 hackneyed topics in the industry. This is quite mind-numbing to see all these statements roaming from one blog to another. We will not tell anything globally new here (we can't change the trends, not yet) but offer a refreshing perspective on applying them.
After combing through tons of upcoming programmatic predictions for 2022 and interviewing a dozen of the greatest ad tech innovators, we know what the major trends are today for sure. Yet, the thing we're not sure about is whether your company needs to chase them altogether.
Thus, we created a quiz that will help you decide what to implement first based on your annual goals. Also, the article features direct commentaries from ad tech leaders on the media buying trends in 2023 and actionable insights for brands, publishers, and ad networks. Does it sound alluring enough to check out? Let's start then.
(Or, you can always watch a video first and then come back to the article if you need extra info):
Programmatic Trend #1: The Final Trial for Post-Cookie Data Solutions
As Google postponed its termination of 3rd-party cookies to 2024, many of us breathed a sigh of relief in 2022. But the new year means new challenges. In 2023, all industry players will wrap up their cookie-related vacation and get down to business with renewed vigor. It will be our time to prepare and dry run new solutions.
Which exact solutions are expected to arrive? Data management platforms will likely re-architect their algorithms to collect consented, personally identifiable data and transfer it to programmatic platforms without defining anonymous segments.
“Solutions developing themselves in a privacy-first environment and companies adapting their strategies to be still able to approach and acquire customers effectively will come out stronger than before - building a strong relationship with their consumers while also retaining the main benefits of performance marketing.”
Google's New Initiative: Topics
One of the ways to collect personally identifiable information data initially proposed by Google was Federated Learning of Cohorts. Cohort solutions were all about local data processing: they use machine learning algorithms that run on the device and form audience segments based on behavior such as browser history.
Thus, advertisers were able to target users based on shared interests rather than using a pseudonymous identifier.
However, on January 26, 2022, Google announced that it killed FLoC and replaced it with Topics. What the heck is this, why cohorts were doomed from the start, and how Topics will work?
FLoC was repeatedly criticized for gathering too much user information without even asking for consent. Cohorts were formed based on hundreds of identifiers and, therefore, many loopholes for fingerprinting. Although being quite beneficial for advertisers, this raises concerns about user privacy since users could not opt-out, unlike with cookies.
Furthermore, cohorts included sensitive demographic data, which could lead to potential misuse of personal information and discriminatory ad targeting. These concerns made FLoC incompatible with GDPR, so Google decided to pull it back.
Topics follow a slightly different approach. They gather browsing data to define interests, assigning a user up to 300 different topics. These topics are equivalent to IAB site categories like "books & literature," "fashion," or "rock music."
And now, let's figure out how this will work. Let's assume that a user visits a website that utilizes Topics for advertising. This is what happens next:
- The browser randomly draws three topics out of 5 top user's interests for the past three weeks;
- The website sends this data to the advertisers through bidstream;
- The advertiser can target relevant users based on their recent interests rather than granular identifiers.
Industry players seem to be less enthusiastic about innovation, though. Advertisers disregard Topics because the categories proposed are "too broad" and "only slightly better than nothing."
Shared User ID Solutions
Another workaround here is shared user ID solutions. Publishers collect 1st-party cookies on their website and forward them to the ID provider through a Prebid API. An ID provider creates an anonymous ID based on the user's consent. This ID will later be shared with ID consortium partners like brands and ad tech companies and sent back to the publisher to store the ID in their 1st-party cookie.
The identifier assigned to the user can be declared or inferred. Declared ones are consistent identifiers like emails or phone numbers. Users are not prone to leaving such info, so inferred identifiers seem more viable. Those are passive identification signals like IP addresses or the device's user agent.
Tip: What shared ID solutions are already on the market? Check out DigiTrust from IAB, ID5, Britepool, LiveRamp, and Unified ID solution.
The third solution that has already been there but little known on the mass market is bidstream data. It represents a code sent from a publisher to an advertiser within a bid request during an RTB auction. Bidstream contains a lot of data about the user, but it is not personally identifiable.
Instead, advertisers receive GPS and location, device type, Wi-Fi or IP address, and many more technical signatures. At first glance, this is not the data that helps identify the user. Yet, since Safari has already blocked 3rd-party cookies, advertisers use a method called fingerprinting to get around identity solutions.
How does fingerprinting work? A publisher collects dozens of technical data points even if the user doesn't accept cookies on the website. Then, it becomes possible to match this dataset to the existing database and define if the user with the identical signatures has already visited the website and what their behavior was.
Though fingerprinting is a probabilistic approach, it can largely help advertisers target rough segments of users based on bidstream data, which technically consists of anonymous information. And actually, iAB's oRTB 3.0 allows passing deterministic ID with bidstream.
Now, it likely seems that we are fully set for 2023. But the abundance of ways to leave the cookie deathbed without mourning creates frustration itself. 60% of marketers believe they will need multiple identity solutions to do business as long as they can combine them.
From the legal side, consent management platforms will be a new must for publishers. With a CMP, it will be easier to automate obtaining consent from consumers and comply with GDPR and CCPA.
“It's overall good to see that the ad tech industry stopped arguing about GDPR (or CCPA) but built with TCF 2.0 framework that integrates these legal requirements into the programmatic process chain. The technical centerpiece to handle all these privacy requirements is the Consent Management Platform (CMP).”
Programmatic Trend #2: Further Growth of CTV, OTT, and In-App Ad Formats
Connected TV, OTT, and mobile-friendly ad formats will continue to rise. 1 in 3 Americans cut the cord and banish the cable to a dump. 55% of them pay for at least one OTT service. And lastly, 3 out of 4 people use their CTV devices daily. This all means we're breaking into the new era of TV commercials - now programmatic.
“CTV component will be a fundamental element of the media planning of many advertisers with an approach that we call “total video” advertising. We conducted a study with CoLab, a survey of marketing professionals from the European countries, where 68% of German marketers forecast to increase their investments in Connected TV devices in the next year.”
Other facts revolve around such ads' effectiveness. OTT ads have an average completion rate of 98%. It's much harder to skip the ad on a Smart TV device, and in some cases, viewers have no choice but to watch if they want to proceed with the video.
During the Covid-19 pandemic, CTV and OTT programmatic ad spend increased by 40%. After many countries loosened quarantine restrictions and we sort of returned to our shaky new normal, users haven't rushed to step away from their TV sets. It can be confirmed by 2021's CTV ad spend - the figures rose by 32% compared to almost 60% for 2020.
The big screen is an absolute new favorite arena for advertising, but let's not forget about small yet mobile ones. Mobile has a 68% share of digital ad spending, as there is no more accessible device to browse the web, play games, and watch videos anytime. With advanced HTML5 formats at their disposal, advertisers will keep benefiting from in-game and in-app campaigns in 2023.
Ad tech mastodons also picked it up and now surfing the app tsunami: we observed huge acquisitions in the segment. Apploving acquired MoPub and Adjust, while Digital Turbine purchased Fyber, AdColony, and mobile DSP Appreciate.
Mobile and CTV get along together: cross-device retargeting is a pure blast decision here. CTV is a perfect arena for branding campaigns but is the underdog speaking about good CTRs for obvious reasons. Thus, it makes sense to retarget the users who have seen an ad on CTV and reach them on their mobile.
Cross-device advertising is still about handling issues with measurement standardization. There is no unified way to track performance among streaming services, and the industry's final standardization is yet to come.
“Currently, there is no single benchmark for measurements in CTV performance. As a result, different ad players, like streaming services and agencies, measure performance differently. Even though the industry (from ANA and IAB to VAB) is seeking ways to standardise data-sharing and enable cross-management measurement this will continue well into 2023.”
Programmatic Trend #3: AI and Machine Learning Development
Both COVID-19 and the data revolution have dramatically accelerated the use of artificial intelligence (AI) and machine learning in programmatic advertising. And this trend will likely persist in 2023. First, it will be fully applied in targeting to process loads of data faster and generate better audience segments. Machine learning is a must in such targeting methods as contextual and behavioral targeting.
Machine learning will also be used to improve bidding optimization to make programmatic advertising more cost-effective. For example, Epom DSP offers a system of bidding rules and multipliers to automatically adjust the bid price based on the given conditions.
“Utilizing machine learning in advertising has been a trend over the past years and its adoption is only about to accelerate. Many ad tech providers and ad networks are introducing more products relying on machine learning to best optimize their client's campaigns and take off the load for granular optimizations.”
What's more, AI can deliver predictive insights by taking into account a customer's browsing history, installed apps, past purchases, past interactions with ads, and a customer's resemblance to previously identified high-value customers. These insights also help advertisers to improve their targeting and bidding and to increase their ROI.
Programmatic Trend #4: Audio and Podcast Ads Scale-up
Mysterious programmatic audio will likely scale as the popularity of online music streaming, and online podcasts grows. Digital audio ad spending in the United States is estimated to reach $6.31 billion in 2022, up 10.8% year-on-year from 2021. Programmatic is set to represent $1.31 billion of that ad spend, increasing to $1.8 billion in 2024 (eMarketer).
While it’s growing, taking care of audio recording and creating engaging audio advertising is key in the podcast industry since it introduces alternatives to Spotify merchandise and other monetization methods. Almost 40% of people find podcast ads less intrusive, according to Semrush research. When podcast ads have high-quality sound, consumers are less likely to skip them. There’s nothing worse than listening to audio with crackles, echoes, or other distracting noises.
Only the minority of demand-side platforms currently support podcasts and other audio ads, yet this will likely change in 2023. Time to look at how to execute programmatic advertising within this promising channel.
Linear audio ad formats work like video ads: you listen to pre-roll, mid-roll, or post-roll third-party insertions placed inside the main program, respectively.
Another type of programmatic audio is host-read ads. It's kind of a native ad in an audio format. With programmatic, host-read ads will be also placed dynamically, being pre-recorded to use during any episode of a podcast.
“The sale of smart devices has exploded in recent years, which is a lucrative avenue for advertisers to grab. For instance, the concept of smart audio ads that would perform action-based responses is also in trials.”
Programmatic Trend #5: Shift from Hyper-Personalization to Hyper-Relevance
The focus in targeting will shift from hyper-personalization to hyper-relevance. What does it mean? Personalized ads create experiences based on the user's overall interests. Relevant ads create timely experiences based on the user's current interests, which means more room for making decisions here and now.
The example of hyper-relevant advertising combined with smart device targeting is geofencing on mobile. It focuses on highly localized areas to provide relevant ads - making it useful even when you lack additional user persona information. Geofencing places higher value and precision on location - and it works well, often becoming a crucial factor in purchase decision making.
This is only one of the factors that can be taken into account - the right location combined with the person's potential needs. When it comes to advertising relevance, we ask ourselves questions whether the user NEEDS the product or not, rather than identifying their INTEREST in that product.
That's a much more complex task to implement than just personalizing the message. Thus, hyper-relevant ads will be powered by the abovementioned AI and machine learning algorithms due to the ever-changing, dynamic nature of relevance.
Programmatic Trend #6: DOOH Advertising Upliftment
By 2023, the DOOH industry is projected to hit US $17.77 billion in ad spending. The cost per person for DOOH advertising is estimated to average at $2.31 this year.
“I'm quietly optimistic that many areas that were hit hard by the COVID19 pandemic will see a resurgence, such as DOOH.”
Out-of-home advertising is one of the oldest forms of promotion, yet its popularity seemed to be fading away once online ads were introduced. And, especially the mass obsession with mobile phones did the trick. But as you know, fads come and go, and now when everybody becomes nauseous about online, OOH will see its great renaissance, now in the new wrapper.
Not that many supply-side platforms now integrate with DOOH. The segment remains more tied to direct forms of media buying rather than traditional online advertising, which experiences a great shift toward programmatic. Yet, DOOH is gaining momentum among SSPs, so the company that owns a DSP and has access to custom supply partners can greatly benefit from the early adoption of programmatic DOOH.
Programmatic Trend #7: White-Label Software Will Loom Larger
"Owning a DSP? Millions and years to spend. Nah, better use the existing one." - advertisers, especially smaller ones, may show their protest. That is the #1 reason why we predict the growth of the white-label software segment, which allows brands to reap all the benefits of platform ownership but spend zero time and energy on in-house software development.
Transparency, modularity, and data richness - these words are seen across the loudest headlines in advertising-related media as trends of 2023. White-labeling your advertising platform, you kill three birds with one stone. Let's see how.
- White-label tech makes your media buying process 100% transparent. You become a root user of a system, see all the insights of the platform, and even can get access to the back-end. As a platform owner, only you decide which supply or demand partners to integrate with and pay no bid markups to middlemen.
- White-label tech is modular and highly customizable. The root user can give out accounts, set up roles and permissions for their own clients. It's possible to adjust interface, optimizations, and even request custom feature development (obviously cheaper than development from scratch).
- White-label tech allows for data collection from the supply side. The platform owner controls all data streams and can use them for their specific needs. For example, you can request audience data from the SSP and enrich your targeting options with custom parameters based on data values of any kind.
Being so well aligned with the overall programmatic trends, white-label demand-side and supply-side platforms might get a new round of recognition this year.
White-label software is a godsend for the companies with $20,000+ ad spend who have money but would prefer to invest it in anything else rather than development. Paying a flat fee to the white-label provider usually turns out to be less spendy than building the platform themselves.”
Summing up, the ad tech industry will still be driven by how the world pandemics will unfold. Other significant factors to affect programmatic are new data regulations, phase-out of traditional targeting algorithms, and some castling in device use. So, seven trends to expect in 2023 are:
[Quiz] Your Personal Programmatic Prediction for 2023
And now, we're going to predict your personal #1 trend for 2023. This is not like tarot card reading or magic ball based on the pseudo-random number generator, but a data-based projection on what's may be the most relevant for your company.
Press "start", and let's find out: should you focus on 4K visuals for TV, create a literal buzz with programmatic audio or reach your zen with a white-label programmatic platform.
Your Personal Programmatic Prediction for 2023
Based on your advertising goals and values, and not on a random number generatorGet now
What about trying a white-label DSP that strives to keep up with all these trends? Sign up for a free trial and enjoy machine learning-based optimization and full mobile and CTV support.Get free trial