The great migration by marketers to first-party data is encountering some bumps in the road, even as many have been careful in managing the transition from a privacy standpoint.

Chief among the issues is that consumers are growing more reticent to fork over sensitive personal information unless there’s a clear benefit to doing so on their end. That’s one of the biggest takeaways from Gartner’s latest CMO research, which draws on surveys of nearly 400 marketing professionals that were conducted in late 2022.

Just 36% of those surveyed by Gartner strongly agree that they are prioritizing first-party data to create more immediate customer value, while a little less than half (47%) somewhat agree with this statement. Still, there are signs of progress. For instance, 42% of respondents stated they are able to execute one-to-one personalization, a goal that was previously seen as a “holy grail.” More than three-fourths (78%) are also encouraging customers to manage their own data.

Of the surveyed marketers, 60% believe executing first-party data strategies that balance a value exchange with privacy will be more difficult this year. Privacy headaches persist even as 85% of leaders in the industry have now introduced a formal policy around customer data management. Nearly one-third of respondents to Gartner said they have cut ties with an agency or channel partner over the past year over trust- or privacy-related concerns. The trend follows not just growing consumer wariness, but also an uptick in privacy laws and regulations that can mete out hefty fines for violations. 

The report also argues against the belief that a focus on fewer channels makes pivoting to first-party data easier. Gartner asserts that the opposite is true and that many organizations with a broader channel mix have been more successful on this front. Close to half of marketers that manage 11 or more channels increased their first-party data collection, while just over a quarter of organizations that work with 10 channels or fewer did that same. Those overseeing a more diverse channel mix were able to drive growth through new experiences, while 24% were applying artificial intelligence to engage consumers.

Persistent challenges in balancing privacy with personalization come as organizations face the looming deprecation of third-party cookies and other forms of what’s termed “signal loss.” Google has set 2024 as the expiration date on cookies, though the tech giant has extended the deadline several times already. Regardless, it’s an expected policy change that has roiled the industry and set off a mad dash for alternatives, few of which have achieved scale or a consensus as a viable replacement. 

“Third-party cookie depreciation is causing sources of data to disappear, and many marketers are still scrambling to shore up their first-party data strategy,” said Ant Duffin, senior director analyst at Gartner Marketing, in a statement. “As this data becomes more challenging to collect, marketers must adapt by leveraging new sources of data to fuel personalization and build deep customer relationships.”

Even as the benefits of first-party data are becoming clearer, many marketers have trouble weaning themselves off old mainstays. A Datonics survey revealed that more than half (53%) of digital marketing campaigns still use third-party data derived from a variety of sources. Behavioral data, interest intent and demographic information were among the most valued data categories.

Artificial intelligence (AI) and algorithms are increasingly key pieces of technology for Procter & Gamble (P&G) as it aims to improve advertising efficiency and effectiveness, brand chief Marc Pritchard said during a livestreamed keynote at the Association of National Advertisers (ANA) Media Conference in early 2023. 

The packaged goods executive, who is board chair of the trade group and delivered the confab’s opening address for the eighth time in its 11-year history, started his talk by acknowledging that P&G has lost its status as the world’s top marketer by media spend. It’s a change he attributed in part to a focus on precision versus volume, suggesting that others in the industry could benefit from adopting similar thinking.

“Celebrating a company as the largest spender implies that it's better to spend more and more each year,” said Pritchard. “We see it a little differently, and that it's time for a reset.”

P&G has made efforts to in-house marketing functions over the years, notably in media buying. This strategy has extended to more tech-forward areas like data and analytics and AI development, Pritchard added at the conference.

Multiple units within the company, which owns Tide, Downy and Pampers, now rely on proprietary algorithms that draw on “a vast repository of consumer research data” from across decades, Pritchard said. With P&G’s in-housing bets reaching fresh levels of maturity and sophistication, its relationship with marketing services providers — which has been contentious in the past — continues to shift.

“More than ever, we need agency partners to see around corners,” said Pritchard, again calling out AI as an example. “We're finding that we can do more work in-house productively and we can strengthen agency partnerships at the same time to create more value.”

Algorithms in action

Pritchard joins a chorus of executives who have put renewed emphasis on AI and adjacent technology in recent months. The rapid rise of solutions like OpenAI’s ChatGPT has captured the interest of both consumers and investors and set off a firestorm of debate around the ethical implications of ceding more work to automation. Top agencies also appear eager to jump into the space.

For P&G, AI intersects with a concentrated push around first-party data. Pampers, for instance, needs to engage the relatively small subset of parents who have children in the diapering stage. In turn, the marketer has switched its model to focus less on broad-based channels like TV and more on owned channels, such as a due date calculator, baby name generator and its Pampers Club rewards program. Data acquired from these assets is then leveraged by an in-house media unit for targeting purposes. 

“With first-party data and other sources, the Pampers media team has created precise smart audiences using a proprietary algorithm that defines consumer groups across the various stages of baby and child development,” said Pritchard.

Advertising creative is also being tested more frequently against AI. Pampers and other P&G brands like Tide rely on an AI studio offering that uses a “neural data network” built on decades of P&G consumer testing, according to Pritchard. Consumer feedback to new efforts is uploaded to this system and measured against existing ads to predict in-market performance.  

“This allows us to make changes to images, music, copy or text to iterate and improve the ad for the very, very best performance, very quickly,” said Pritchard. 

The executive claims this approach is cheaper and more efficient than traditional ad-testing methods. Pampers is now able to optimize more than 140 different ads per year for various channels, including TikTok, YouTube and e-commerce. The brand grew its U.S. sales by 10% in 2022 while increasing return on investment by 17% and media savings by 15%. 

Tide applies AI differently because it’s aimed at a wider swath of consumers, giving TV and streaming greater prominence. The detergent brand now has planning and scheduling algorithms that help it place ads across dozens of networks and thousands of program combinations while cutting down on redundancy, a sore spot in connected TV. P&G claims it saved $65 million in 2022 through its fabric care in-housing.

“Now, they're using an algorithm to test a new way to buy ads that's spread evenly across all TV programs, channels and day parts, which should increase reach another three points and save another $40 million,” said Pritchard of the fabric care team.

New frontiers

With automation bolstering aspects of P&G’s legacy media business, an algorithmic bend is bleeding into more nascent channels as well. Retail media networks represent one of the CPG marketer’s largest and fastest-growing buckets of spend, according to Pritchard. 

Together with retailers, P&G is creating what he called an “auto-bidder.” The digital solution opens direct access to retailer search programs and lets P&G adjust its ad buying and content “every 15 minutes automatically,” the executive said. The end result is four times higher return on brand sales and bigger basket sizes for merchants. 

During a Q&A portion of the event, Pritchard was asked about how he thinks about AI practically for P&G amid a period where marketers are trying to discern what’s hype versus substantive.

“The best way, we found, to do it: Don't talk about the algorithms, don't talk about the technology, don't talk about AI,” said Pritchard. “Talk about the outcome you want. What are we trying to achieve?”

Sponsored content

By ARM Treasure Data

Leveraging the power of machine learning and the increasing availability of data, companies increasingly use predictive analytics to identify customers’ intent, design experiences tailored to individual preferences, inform product development decisions, and improve business operations.

Predictive Analytics Use Cases

Marketers must constantly find new ways to engage customers so it’s no wonder that predictive analytics are most commonly used to predict customer behavior. However, other functions such as supply chain, operations, product, and service can also leverage predictive techniques to improve business results. Here are six use cases where data insights can quickly make a difference.

1. Identify Customers Likely to Churn
   Predictive retention models can identify which customers are most likely to churn — companies can reach out to them with education on products or tailored promotions. Predictive models can also identify common behaviors of loyal customers. With that knowledge, companies can attempt to steer likely churners to adopt those desirable behaviors. 

2. Recommendations for Cross-sell and Upsell
   When selling a variety of products, predictive modeling can help improve the “recommended for you” engine by analyzing historical data and applying customer profiles to target optimal conversion. For example, someone who has purchased hiking boots might be shown other outdoor gears. 

3. Prediction of Features for Product Development
   Predictive analytics can help determine which features to include in upcoming releases. A recommendation engine can be used to score features with various attributes. Information such as what competitors are doing can also be incorporated. 

4. Inventory Forecasting and Pricing Strategy
   Businesses can use predictive analytics to get insights from combinations of data, such as inventory across a large number of SKUs, categories, and locations overlaid with geographic sales data. This can identify buying patterns to inform pricing, promotional strategy, and optimal inventory levels. 

5. Calculation of Risk
   Predictive analytics is also useful for the banking and insurance industries: to calculate risk. Predictive scoring models can evaluate prospect lists and provide an indication of the relative risk of a potential customer. The predictive model could examine hundreds of risk indicators, narrowing characteristics down to small details, and predict expected losses.

6. Predictive Maintenance and Quality Fulfillment with IoT Monitoring
   If you are responsible for customer support or hardware maintenance and repair — perhaps monitoring through IoT devices — knowing which components are most likely to fail would allow you to be proactive rather than reactive, thus avoiding unscheduled downtime. Replacement or maintenance of parts can be planned based on real-time data rather than schedules. 

Benefits of Predictive Analytics

Predictive scoring gives marketing teams the ability to allocate budgets wisely — focusing on getting the right message to the right people at the right time — and improve campaigns ROI. An analytics system makes it easier to identify customers with shared traits and finding look-alike customers from your install base. Once the customer segments have been identified, marketers can create custom messages and campaigns targeted to those segments, creating special offers designed just for them and activating on the channels they prefer.

Making Predictive Analytics Accessible

Arm Treasure Data’s Customer Data Platform (CDP) is purpose-built for marketers — developed with both data availability and accessibility in mind, designed to bring marketing and data teams together. The technology behind the Treasure Data CDP is a scalable, enterprise-grade big data platform. It combines pipelines of clean, unified customer data with a robust and reliable open source machine learning library.

This puts powerful machine learning tools into more hands, allowing more people to gain greater understanding of their customers, and building business practices around the customer.

Data clean rooms (DCRs) are a priority area for many marketers, but raise tough questions for the industry as adoption spreads, a report from the Interactive Advertising Bureau (IAB) found. 

Steep investments are required to maintain DCRs, both in terms of building the underlying tech and the talent needed to manage it. That puts smaller brands, agencies and publishers at a “tremendous market disadvantage,” per the IAB — one that may curtail in-housing trends. The IAB also views DCRs as needing data interoperability to reach their full potential, an area where marketers have historically struggled. Among surveyed DCR users, 39% pointed to challenges with data interoperability and customization, while 38% called out a lack of internal resources and 32% did so for privacy hurdles. 

DCRs have become a fixation point for marketers as they try to address the pending deprecation of third-party cookies. The technology allows advertisers and platforms to dump their aggregated first-party data into one place in a privacy-safe fashion. Different data pools can then be matched up to refine targeting and address problems like audience duplication, among other functions.

Platforms and publishers have quickly wised up to the business potential. Amazon has expanded clean room capabilities powered by its AWS cloud-based computing juggernaut. Disney Advertising also enhanced its existing DCR through a VideoAmp integration. 

Broadly speaking, companies are expected to invest 29% more in DCRs in 2023 compared to 2022, according to the IAB. The research conducted by Ipsos draws on 200 surveys and 20 in-depth interviews with stakeholders from across different industry categories. 

“There is growing demand due to loss of signals, and wanting to share data with partners without sharing sensitive data,” said Jeffrey Bustos, vice president of measurement, addressability and data at the IAB, in a statement. “This may show that companies are understanding what’s required and ramping up their investments accordingly.”

But the trade group’s report also emphasizes that the resource intensiveness of DCRs could leave smaller and mid-size marketing players behind at a critical juncture. 

DCRs are often leveraged in tandem with other ad tech like customer data platforms, data management platforms and consent management platforms. The compounded annual costs of these stacks can climb above $2 million, per the IAB.

Nearly one-quarter (23%) of marketers spent over $500,000 on their DCRs alone in 2022, while 62% put a minimum of $200,000 toward the technology. The heavy price tag could be more of a drag in a downturn, when marketing departments are often among the first to feel cutbacks. 

Time is another consideration, with respondents stating that standing up a DCR offering can take up to two years. Meanwhile, Google has slotted cookie deprecation to take effect in 2024. On the talent front, roughly half of DCR users have teams of six or more people dedicated to the tech and 30% have a minimum of 11 specialists.

Looking beyond the barriers to entry, the IAB believes DCRs could be held back without full data interoperability. That’s a level of standardization and unification that’s been called for in digital marketing and ad tech for years, while the landscape in reality has often remained fragmented and controlled by walled gardens.

“It’s in everyone’s interests for closed ecosystems to provide access so that advertisers can effectively analyze their campaign activity,” said Bustos. “Advertisers need a window into the data.”

Heineken applied attention-based measurement for the first time in a globally tested campaign and achieved better-than-expected results, per details shared with Marketing Dive.  

The beer brand’s Rock in Rio 2022 sponsorship was deployed with agency Dentsu and programmatic ad-tech platform Teads, which used Lumen´s Research LAMP program and a metric called APM to track success. The technology is based on a predictive attention methodology that’s based on knowledge of consumer behavior in digital environments. Heineken’s APM was five times higher than Lumen’s digital benchmarks while its view rate was two times higher. The marketer also received the highest level of brand recall among Rock in Rio sponsors, according to a news release. 

Discussing the Rock in Rio results, Heineken representatives said the brand prioritized an attention-based approach to see a “more qualified vision of impact.” 

“In an environment where people dedicate less than 10 minutes of their attention each day to advertising, we need to find other ways to evaluate engagement,” said Fernanda Saboya, director of consumer connections at Heineken Brazil, in a statement. “By surpassing expectations, results show we are on the right path to creating more assertive campaigns that capture attention in a much more effective way.”

The effort utilized “disruptive” display formats and social placements highlighting giveaway promotions before and during the event and used data to target three sets of festival goers: Gen Zers, alternative music fans and vintage enthusiasts.

The APM metric, or average attention per 1,000 impressions measured in seconds, is part of the emerging attention-based methodologies that solutions providers like Teads are championing. The attention economy has become a marketing buzzword as holding people’s interest for even a few seconds proves challenging in a busy digital world. Attention-based forms of measurement have subsequently gained industry traction as brands try to assess not just whether a person saw an ad, but also if they engaged with the media on a deeper level. For example, someone served a spot on YouTube or connected TV might check their phone until the video ends, meaning the marketer reached a degree of viewability without actually producing a high-value interaction. 

Agency partner Dentsu has also attempted to push past a square focus on viewability, believing the metric alone is insufficient for gauging effectiveness based on complex factors like an ad’s format size, length and page position. 

The search for richer forms of digital measurement comes as mainstay tactics experience disruption. Third-party cookies, one of the most popular ways of keeping tabs on consumers online, are set to be phased out in 2024. Mobile platforms like iOS are also implementing privacy changes that make targeting and measuring campaigns more difficult.   

NEW YORK – Emerging technologies such as AI and virtual try-ons have enhanced the consumer experience in the post-COVID landscape, said Ophelia Ceradini, vice president of digital technology and innovation at The Estée Lauder Companies.

Ceradini, whose comments were made during a panel at National Retail Foundation’s (NRF) Big Show, pointed to the quickly changing relationship between brands and consumers. Despite returning to stores, consumers do not want a separate online and offline identity, per the exec. Instead, the beauty consumer’s online identity should follow her when she reaches the store and vice versa. 

“It’s all about using the data you have...making her feel like one person,” said Ceradini.

Ceradini was joined by Nicolette Bosco, vice president, divisional business manager for beauty at Macy’s and Kelly Kovack, the founder and CEO of BeautyMatter. Kovack served as the moderator of the panel, titled Beauty’s blush with technology. The trio discussed how technology is changing the industry, including the impact of AI and the difficulty of keeping up with trends. 

AI is for everyone

Despite increases in both online and brick-and-mortar sales, the rate of merchandise return has remained steady or dropped, indicating a more satisfied consumer. Technology is one potential reason behind the shift. Software enabling virtual try-ons leaves consumers guessing less about what they want. This means that the consumer is usually more informed and has a better idea about what they are purchasing. 

However, technology carries its own assortment of issues that marketers must deal with. For example, the human population is incredibly diverse, and catering to all body types and ethnicities has proven elusive. Making sure all customers are seen is incredibly important, especially for beauty brands, per the panelists. For example, Estée Lauder has made an effort to cater to every type of eye shape with its virtual try-on technology.

There’s a virtuous circle for brands when they can use technology to provide value for consumers, who are more comfortable sharing data with the company if they are able to get something out of it, according to Ceradini. Such data can help marketers further optimize their messaging and activations. 

A more individualized experience has also been beneficial on the retail side, according to Bosco. For example, Macy’s has introduced an auto-replenishment system which allows consumers to set how often they want a product sent to them.  

“Once you've found something that you absolutely love, you can go on and set it and almost forget it,” said Bosco.

Power to the beauty experts

Both on the retail and brand side, there has been an effort to empower beauty advisors, according to panelists. By granting consumer-facing employees the ability to keep up with trends, they are able to help consumers find what they are looking for even if the brand itself has not had time to pivot toward what the latest TikTok trend is. 

That doesn’t mean the brand itself shouldn’t be on top of trends, according to Ceradini. It’s important that brands try to predict what they can and pivot their advertising when necessary. On the retail side, it’s important to make the most of the inventory in stock, according to Bosco.

However, allowing floor beauty advisors to be on devices and to look up the latest trends gives them the power to meet the consumer where they are. From there, the imperative becomes letting consumers know where the latest trendy product can be found. 

“But then really pivoting the market to tell consumers, ‘Hey, we have it here,’” said Bosco.

Nearly 61% of hotel brands have implemented a first-party data strategy, and of them, 81% reported an increase in revenue after implementing the strategy, according to findings by travel marketing company Sojern.

First-party data strategies have become increasingly valuable as marketers prepare for the looming end to third-party cookies. With their widely used loyalty programs, credit-card partnerships and other data-gathering initiatives, hotel brands are sitting on a wealth of first-party data that can be used to better cater to target consumers. 

Sojern’s study, which surveyed 74 executives representing a cross-section of hotel companies in North America, Europe and Asia Pacific, indicates how hotels are already seeing results from leveraging their first-party data. Among other benefits, the majority of respondents saw an increase in campaign performance, increased brand value and stronger customer relationships. Two-thirds of respondents expect to see more benefits in 2023.

Additionally, 57% of respondents indicated they saw an increase in guest satisfaction at their properties, 76% reported that they saw an increased campaign performance, 68% reported stronger customer relationships and 64% said brand value increased. 

The survey also indicated that 59% of hoteliers have already implemented a first-party data strategy, 62% of whom said the strategy was “very important” for digital marketing. Of those who had not yet built a first-party data strategy, 75% said they want to build one. Eight-six percent view a first-party data strategy as being useful in maximizing overall revenue. 

While first-party data is clearly important for hotels’ marketing efforts, it could also prove valuable should any more hotel chains opt to create their own media networks. Already a growing channel for retailers, proprietary media networks offer companies an opportunity to use their owned media channels — like in-room television and digital screens in hotels — to host paid advertising from other brands. 

Marriott International launched the Marriott Media Network earlier in 2022, the first such network for the hospitality industry. The media network will initially target travelers in the U.S. and Canada, but Marriott said it would expand the platform globally giving advertising access to the 164 million members of its Bonvoy loyalty program. 

“It’s no surprise that three-quarters of hotel executives want to build a first-party data strategy,” said Kurt Weinsheimer, chief solutions officer at Sojern, in a release. “With a cookieless world coming, it will disrupt the entire industry, and marketers must use this time wisely to prepare.”

Meta is rolling out a new machine-learning solution for reducing algorithmic bias in its advertising in 2023, according to a blog post. The Variance Reduction System (VRS) was developed over the past year with the Department of Justice (DOJ) as part of a settlement agreement over alleged discriminatory practices in the way Meta’s housing ads were served. 

VRS, which is rolling out now in the U.S., carries implications beyond the housing category as well. Meta plans to implement the system for other areas of its sprawling ads business that are affected by federal discrimination laws, including employment and credit ads, later this year. The DOJ is positioning the new resolution as not just a power check on Meta, specifically, but also a wake-up call for other platforms that rely on algorithms to deliver digital advertising at scale. 

“This development marks a pivotal step in the Justice Department’s efforts to hold Meta accountable for unlawful algorithmic bias and discriminatory ad delivery on its platforms,” said Kristen Clarke, assistant attorney general of the DOJ’s Civil Rights Division, in a press statement. “Federal monitoring of Meta should send a strong signal to other tech companies that they too will be held accountable for failing to address algorithmic discrimination that runs afoul of our civil rights laws.”

VRS is the outcome of a lawsuit that saw Meta, formerly known as Facebook, come under fire for its housing ads. According to the case, the company applied algorithms that could serve ads based on characteristics federally protected under the Fair Housing Act (FHA), including sex and race and ethnicity, resulting in discrimination. 

The social networking giant in June settled with the DOJ, with VRS one of the more substantial ways it’s trying to remediate the situation and ensure its advertising is equitable. As part of the agreement, Meta is also sunsetting a Special Ad Audience tool that targeted ads at lookalike audiences and is removing targeting options for housing advertisers that describe or relate to FHA-protected characteristics. 

A third party selected by Meta and the DOJ, Guidehouse, will review the VRS to ensure compliance, with the courts having the ultimate authority to resolve any disputes that arise. The DOJ called the resolution “groundbreaking” and said it hoped other platforms would follow the social media firm’s lead in tackling algorithmic bias.  

The news shows Meta’s advertising coming under direct court oversight for the first time. It’s another sign that years of controversy surrounding data and ad-targeting practices are starting to catch up with the Facebook and Instagram owner. Meta was also recently fined roughly $275 million by Irish authorities over its approach to ads personalization.