The impact of Artificial Intelligence (AI) on marketing teams continues to grow and 2023 could in fact be a watershed year for shaping how AI, marketers, and consumers interact.
Personalization, Privacy, and Profit – The Impact of Artificial Intelligence on Modern Marketing
SPECIAL REPORT by jenalea howell, vp,
applied intelligence group, informa tech
AI’s Impact on Creative
(and the ‘Creators’)
AI and Privacy
Enhancing Your Marketing Mix with AI
01
02
03
In this special report, we will explore how personalization and privacy are likely to manifest themselves across three key themes in 2023
AI’s Impact on Creative
(and the ‘Creators’)
01
ChatGPT and Content Creation:
New Possibilities for Marketing?
What is it?
ChatGPT uses machine learning (ML) and natural language processing (NLP) to mimic human-generated interaction. A user can enter a query using everyday language and ChatGPT provides an answer that sounds like a person rather than a computer has responded. Third-party companies are now using the ChatGPT API to create applications intended to enable new use cases for marketing departments.
The Promise and Impact
For marketing departments, ChatGPT’s greatest asset could be its ability to quickly create personalized content. For instance, a marketer could enter information about a particular audience and ask for ChatGPT to return separate emails for each segment. ChatGPT can also help with SEO by tailoring communications to include relevant keywords. With responses returned almost instantly, creating product descriptions can also be streamlined. Marketers could potentially enter customer feedback and ask ChatGPT to analyze the response for trends that can inform future work or aid customer care teams. Microsoft thinks so highly of ChatGPT functionality that it plans to integrate the platform into Bing, perhaps finally creating a true rival to Google’s search engine.
As with any technology, however, potential exists for ChatGPT to be put to not just beneficial, but damaging, use. One area of growing concern that marketers should keep in mind is ChatGPT’s impact on cybersecurity. Given its ability to rapidly craft unique messaging, ChatGPT functionality could be used to develop email phishing campaigns or business email compromise (BEC) attacks. Deployed in this way, AI could be used to develop unique or even personalized phishing content that will be much harder to detect and, from a marketing perspective, could put brand reputation at greater risk.
The Current Reality
Content powered by ChatGPT functionality is not fact-checked and links or references are not provided. Because the learning data was cutoff in 2021, it’s also impossible for ChatGPT to answer questions about current top trends or anything that requires real-time data. Moreover, ChatGPT will at times repeat answers in slightly different contexts, leading to potential redundancy. Finally, as with any AI, ChatGPT is only as good as its training material: any biases within the training data can be reflected in the answers it provides. For all these reasons, marketing teams should be ready to provide a close editing pass across any messaging produced by ChatGPT.
The Takeaway
ChatGPT is a huge advance. But there are limitations that marketers should keep in mind. Responses from ChatGPT sound like a human creates them, but they often need heavy editing and, as noted above, fact checking. At this point, ChatGPT can likely best help with creating initial drafts that can then either be edited or used as a brainstorming tool. It can also help create efficiencies by taking existing content and returning variations for different channels, like transforming a blog post into complementary social media posts. Creating email subject lines or headlines that are targeted to specific audience segments is also possible.
While the hype is real and the future looks (largely) bright, at this point ChatGPT is better as a tool to help a marketing team work efficiently rather than a means to fully replace marketing work itself. Finally, although not a direct threat yet, marketers should keep an eye on the potential for ChatGPT functionality to be used in support of phishing campaigns that could put hard-earned brand equity at risk.
OpenAI graphic design:
painting a new future?
What is it?
Taking advantage of advances with natural language processing (NLP) and machine learning (ML), OpenAI has created text-to-image capabilities. It takes the prompts and parameters a user provides to return a visual representation in near real-time. In short, it understands the text you enter, connects those inputs with existing visual concepts from its vast library of machine learning, and then creates an original design. Want to see an astronaut peddling a blue ten-speed bike on the water? Done. It can also manipulate or edit an existing image that you upload to add new elements. With either new or existing images, it can then create variations on the original theme.
The Promise and Impact
For marketers, OpenAI graphics could mean greater personalization of content and incredibly fast design work that generates greater efficiencies. Need an image for a blog post or an email campaign? OpenAI graphic design can quickly create variations for each of your audience segments. You can even ask it to offer images appropriate for different formats (i.e., a banner ad, a billboard, or even a layout for website design). There’s fantastic stylistic flexibility as well: you can ask for images to be photo realistic, resemble pixel art, mirror the style of your favorite painter, etc.
The Current Reality
While OpenAI graphics design offers incredible speed, most industry experts agree that it isn’t ready to fully replace the human touch. Creating the perfect image, even with OpenAI graphics tools, requires strong creative direction. OpenAI graphics are therefore likely to best function as a creative assistant or as a tool to enhance workflows rather than as a full replacement for all design work.
The Takeaway
As a creative tool, OpenAI graphics engines can help with artistic brainstorming, add quick examples to a creative brief, or help illustrate early concepts so that a design team can show (rather than tell) what they have in mind when interacting with clients or internal stakeholders. Teams can also upload existing and fully polished designs to have variations on that original work quickly created to support campaigns. Using OpenAI graphics tools to modify and edit images could also be a time saver for many teams. Small businesses, which might lack a full design team, could be especially well positioned to take advantage of OpenAI graphics.
Driving both sides of this dichotomy is the ability of AI and machine learning to mine vast amounts of marketing data to develop more comprehensive and nuanced understandings of audience needs, wants, and preferences. In an inverse relationship, as the pool of data available to marketers grows larger, smaller and smaller audience segments can be targeted with enhanced precision.
To cite two current examples, the emergence of ChatGPT and new AI-powered graphic design tools could mean greater ability for marketers to create unique, custom content in support of personalized experiences and messaging to drive increased conversions and profit. In short, AI is helping not just enable hyper-personalization but has created the possibility of doing so at scale.
Indeed, thanks to AI’s ability to mine vast amounts of data, the traditional “4 Ps” of Marketing (Product, Price, Place, and Promotion) have three new critical additions: marketers using AI must balance between the desire for greater Personalization with the need to protect user Privacy, whilst still showing Profit from marketing dollars spent.
However, AI’s impact on marketing isn’t necessarily that straightforward. To borrow from a caricature of advertising, “but wait, there’s more!” The flip side of AI collecting and analyzing incredibly vast amounts of data is a growing concern that the desire for personalization can easily bleed into an intrusion of user privacy. From changes to data collection introduced in iOS 14 to the White House weighing in with a new AI Bill of Rights to a shift to a cookieless future of advertising, gathering the data needed for hyper-personalization is becoming more difficult for marketers.
To navigate these two sides of the same coin successfully, marketers will need to prioritize agility. The good news is that here too AI can help by creating responsive marketing mixes that can shift quickly as brand goals, user interests, or regulatory requirements change.
02
AI and Privacy
A cookieless future?
Theme
What is it?
Cookies, those small pieces of data stored on a user’s browser to help track activity and online behavior, have long been a foundation of digital advertising and programmatic campaigns. However, due to increasing privacy concerns, advertisers and content platforms are preparing for a cookieless future.
The Promise and Impact
Part of a broader shift toward increased transparency, eliminating cookies is meant to restore privacy and agency to consumers. For those in marketing, this shift has enormous implications for activities like retargeting, tracking user behavior, and attributing conversions correctly. Moving forward, advertisers will find it more difficult to serve ads targeted to audiences based on behavior and demographics, while content publishers likely won’t be able to sell their platforms as effectively to those same advertisers. The loss of cookies, in short, promises to disrupt long-standing digital advertising practices.
The Current Reality
The industry is at a crossroads, and the immediate future for digital marketing looks to be fragmented as providers search for a replacement to cookies. The most promising options are using AI to develop enhanced targeting that does not intrude on user data or privacy. Google, for instance, recently announced its plan for what it calls Topics API, which works in conjunction with a user’s browser history to identify weekly topics of interest for a user. No sensitive information like gender or race is ever kept, and topics are retained for only three weeks before being deleted. Information, moreover, is never shared with any external servers. When a user visits a participating Topics API site, one topic from each of the last three weeks is shared with an advertising partner to help deliver more personalized messaging to a user.
AI also promises to expand opportunities for contextual advertising as a replacement for targeting based on cookies. Contextual advertising places ads on webpages that have content that will appeal to those with like interests or collective behavioral attributes. It’s about the pages and content rather than individual users or attributes. AI promises to play a critical role in contextual advertising by helping create algorithms that consider multiple and rapidly shifting inputs like the content of a web page, the location of a user, or even the season or weather to calculate messaging that is likely to resonate in that particular scenario.
The Takeaway
It’s not a matter of ‘if’ but ‘when’ cookies disappear. Now is the time to prepare alternatives that can best position your marketing work for success moving forward. Collecting more first-party data (i.e., information that your users willingly provide to you) is a key step. Businesses, for instance, can consider offering discounts or other bonuses in exchange for users completing surveys or taking quizzes that can help a company shape more effective communications. It’s likely that second-party data companies (those who collect first-party data and sell it to others, with the original user’s consent) will grow in importance. From mining first-party data in order to develop user-specific content, to creating new systems like Topics API, to enhancing the capabilities of contextual advertising; AI-powered algorithms are helping shape a future where digital advertising can not only abandon cookies but actually enhance both personalization and the trust between brands and users.
02
AI and
Privacy
A cookieless
future?
Theme
The Takeaway
It’s not a matter of ‘if’ but ‘when’ cookies disappear. Now is the time to prepare alternatives that can best position your marketing work for success moving forward. Collecting more first-party data (i.e., information that your users willingly provide to you) is a key step. Businesses, for instance, can consider offering discounts or other bonuses in exchange for users completing surveys or taking quizzes that can help a company shape more effective communications. It’s likely that second-party data companies (those who collect first-party data and sell it to others, with the original user’s consent) will grow in importance. From mining first-party data in order to develop user-specific content, to creating new systems like Topics API, to enhancing the capabilities of contextual advertising; AI-powered algorithms are helping shape a future where digital advertising can not only abandon cookies but actually enhance both personalization and the trust between brands and users.
A cookieless future?
What is it?
03
Enhancing Your Marketing Mix with AI
AI for
optimizing campaigns
Theme
The Promise and Impact
The Takeaway
Companies that embrace AI could have an advantage for developing impactful marketing mixes, even as information sources shift and decision-making variables increase. That said, implementing an in-house AI solution isn’t necessarily an overnight endeavor. Long development cycles and the need to update AI modeling can complicate the act of operationalizing AI into a marketing mix. The next wave could be a rise in companies that offer AI-as-a-Service to bring the best benefits of AI, with the least operational difficulties.
Capable of handling multiple inputs with both ease and speed, AI modeling enables marketers to assess how marketing mix performance is affected by variables like: seasonality, promotions, geography, economic trends, pricing, weather, and any other key elements for a business. Reducing time to insight from months to weeks, AI promises to help improve agility and decision-making velocity, all with greater granularity for forecasting, and enhanced adaptability when market conditions shift. Better and faster predictive modeling will help marketers fine-tune their marketing mixes on the fly and ultimately improve ROI.
iPhone 14 as a
catalyst for privacy
What is it?
Each Apple device has a unique number known as an Identifier for Advertisers (IDFA). Starting with iOS 14, Apple has given users greater control of what personal data is collected through this IDFA and how it is used. In the past, a user’s IDFA has enabled a level of cross-app and website tracking that provided advertisers with aggregated data about a user’s interests and behaviors. With the shift to iOS 14 and beyond, however, users can opt out of this type of tracking.
The Promise and Impact
As with the move toward a cookieless future, the iOS 14 changes fit within the broader cultural shift toward giving consumers greater data privacy controls. Apple’s changes have proven extremely popular with consumers. Reports indicate up to 90% of users choose to opt-out of IDFA tracking when given the chance. For marketers, the iOS 14 changes complicate attribution, measurement, frequency capping, and the ability to create personalized, targeted campaigns.
The Current Reality
The changes have taken a particular toll on marketing strategies built around Facebook. Advertisers have long relied on the Facebook pixel to understand what users are clicking on, downloading, and purchasing to inform campaign work. Apple’s privacy controls have cut off this ability and have reportedly cost Meta up to $10 Billion in advertising revenue.
There are, however, other levers that marketeers can (and must) pull in order to strike a new balance between respecting user privacy, and the benefits generated by delivering personalized and relevant content. Greater use of first-party data is high on the list of alternatives that have grown in importance. Understanding user behavior can also be enhanced by greater use of Google Analytics and setting Urchin Tracking Module (UTM) parameters. For instance, by associating a unique UTM with each ad, source, broader campaign name, and more, advertisers can gain back some of the user insights lost by blocking IDFA tracking.
The Takeaway
The privacy changes introduced in iOS 14 sent shockwaves through the advertising industry. Tracking has become more difficult and there is no longer a one-size-fits-all approach to gathering user data. But savvy marketers should also focus on how wildly popular these data controls are with consumers. Creating trust and transparency with consumers and ensuring that users receive value in exchange for any personal information provided is the new opportunity frontier. This is precisely where AI can play a pivotal role. In the past, striking a balance between personalization and privacy was much more difficult. AI, however, is helping generate insights and messaging recommendations by passing through, and modeling against, vast amounts of data that can be walled off from any individual user. Results from these AI-enabled algorithms can then be continuously tested and fine-tuned to ensure relevance and, ultimately, help advertisers deliver personalization at scale. In short, AI could be the key for brands to succeed at striking a powerful balance between privacy and personalization
moving forward.
Data governance
and ethical AI
What is it?
As AI grows and enters the mainstream, so too must the rules of governance evolve. Recognizing this need, in October of 2022, the White House Office of Science and Technology Policy (OSTP) released “Blueprint for for an AI Bill of Rights: A Vision for Protecting Our Civil Rights in the Algorithmic Age.” In short, there is growing recognition that AI’s ability to gather and mine vast amounts of data must not erode the rights of individuals, overstep privacy boundaries, or foster discriminatory practices. For marketers, this critical focus on ensuring that AI is used in an ethical way brings both far-reaching repercussions and new opportunities to differentiate your brand and earn the trust and loyalty of consumers.
The Promise and Impact
Enhanced data governance and oversight of AI in marketing should be seen in conjunction with the inverse desire to create greater messaging customization and improved audience targeting. Moving forward, personalization and privacy must walk hand-in-hand (as counterintuitive as that might at first sound). Marketers have access to more data and customer journey information than ever before, but this creates a greater responsibility for marketing teams to establish guardrails that create trust and transparency with consumers. In practice, this means establishing internal systems that give consumers agency over data collection and usage: users must know what data is being collected, when it is gathered, how it will be used, and who to contact for help. Brands that excel in creating this balance between personalization and privacy are likely to be those that develop not just actionable insights that benefit campaigns and marketing mixes but also improved customer loyalty and satisfaction.
The Current Reality
In addition to concerns that AI-enabled systems have the potential to overstep privacy boundaries, marketers should keep in mind that AI is only as good as the training data it uses. Any biases inherent in that training data can mean perpetuating inequality and harming, rather than enhancing the customer experience. To cite just two well-publicized examples: AI systems have favored white male job applicants, and promoted discriminatory lending practices in mortgage applications. For marketing, relying on algorithm inputs that are inadvertently biased could mean alienating and harming the very audiences you hope to connect with on a deeper level.
The Takeaway
Discussions around ethical AI and enhanced data governance are gaining traction precisely because AI and machine learning is reaching a tipping point. Excitement surrounding AI’s impact on marketing continues to grow—and rightly so. But the vast promises of AI do not mean it is a panacea. AI should never be seen as a set-it-and-forget-it system. Evaluation of the data sources used, and algorithms applied, should take place early and be supplemented by subsequent monitoring and evaluation to ensure responsible privacy and non-discriminatory practices are being followed. More broadly, marketers must create transparent messaging to consumers about the information that is being collected, establish clear pathways for consumers to opt-out, and provide systems that help them retain agency over how their data is used and provide remedy for any potential issues that arise.
Theme
Theme
What is it?
Marketers are all too familiar with charges that the discipline’s unique combination of art and science makes it difficult to assess attribution, assign ROI, and forecast effectively. Outputs from AI promise to give marketers greater data-backed insights to evaluate what is working well and where marketing mixes should be adjusted to have the most impact.
The Current Reality
Adjusting to rapid shifts in consumer behavior or platform volatility is likely to remain a key concern for marketers in 2023. In addition to continued price and inflationary pressures, you need look no further than the ongoing transformation of Twitter since Elon Musk’s purchase to know that the best laid marketing plans can go out the window at any time. From offering the previously coveted Twitter blue check mark for sale, to fake brand accounts then arising, to removing guardrails around what can and cannot be said on the platform, recent Twitter changes have created a fluid situation for marketers. In the immediate wake of this chaos, many brands paused their ad spend with Twitter or moved their marketing mix away from the platform entirely. Against this backdrop, marketers who use AI for enhancing predictive modeling, creating faster time to insight, and supporting the ability to adjust to rapidly changing conditions, might have a secret weapon to improve their marketing mixes.
03
02
01
03
02
01
BACK TO TOP
"
In our constantly evolving digital world, maintaining brand relevance is an always on challenge for marketing teams. Advancements in AI have provided direct access to the habits and behaviors of target audiences that is helping marketers stay ahead. The ability to power campaigns using automated data collection and analytical tools allows teams to make real time pivots on campaigns. For PR pros and content creators, Generative AI has enabled teams to quickly react to current trends and a rapidly changing news cycle by producing instant copy
and visuals.
Technology can be empowering, but a healthy balance is needed for AI in marketing to be successful with audiences. I’m most interested in how developers will be addressing biases in AI to ensure the resources and tools being implemented by marketers are equitable for all communities. In the end, brand communications
will always require a human touch.”
Suzanne Matulay,
Founder & Managing Partner, Mat+Lo Creative
"
HEAR
FROM
OUR EXPERTS
This push-pull tension between personalization and privacy is likely to continue to be a dominant storyline throughout 2023.
On the one hand, AI is accelerating long-established marketing efforts to get closer to the consumer and refine audience understanding; on the other hand, AI is a catalyst for rethinking traditional marketing boundaries and data practices.
Product Price Place Promotion
If you’d like to keep up to date on all things AI, then be sure to subscribe to our AI Business Newsletter here.
Gilbert Hill, Privacy
Technologist &
Entrepreneur
AI will play a key role in the post-cookie world, but not how people imagine,
at least initially.
I see lots of vendors pitching AI software which helps companies dynamically ‘roll their own’ adaptable policies. I think the first real use of
AI will be to scan existing policies for intelligence on the structure and operational efficiency of a company’s data governance, whether it’s done by a regulator, competitor or potential investor. This will show who is ready to move beyond compliance to real accountability and transparency. With
trust baked in, and
as platforms open their silos, the potential for AI to create de-risked, serendipitous insights for consumers based
on their data is
truly exciting.”
"
AI is being leveraged to improve outcomes in a wide range of use cases across marketing and advertising. We, at Omdia – the technology research arm of Informa Tech – are tracking and forecasting spend for more than 20 of those use cases, with the top nine featured in the chart below.
AI is already improving outcomes in Marketing & Advertising
The Expert’s View –
Mark Beccue, Principal Analyst, AI and NLP, Omdia
AI’s Impact on Creative
(and the ‘Creators’)
AI and Privacy
Enhancing Your Marketing Mix with AI
01
02
03
In this special report, we will explore how personalization and privacy are likely to manifest themselves across three key themes in 2023
AI has been quietly at work improving outcomes for marketing and advertising professionals for the past few years through more than a dozen use cases. Spending on marketing and advertising AI will grow from $4.3 billion in 2023 to $6.9 billion in 2025.”
"
Top AI use cases for Marketing
and Advertising
Overall, the two biggest use cases for marketing and advertising are image recognition/tagging and digital experience marketing. Image recognition fueled by computer vision technology is accelerating use cases in retail for visual search and automated content management.
Visual search is widely associated with the notion of consumers leveraging images or their smartphone camera lens (Pinterest Lens, Google Lens) to find products they want. How a consumer describes a product is wildly varied, and nuanced. The problem visual search solves is not one of seeing a product per se; rather, it is one of reckoning natural language, text-based search inquiry with discovering exactly what a shopper wants. “Find products like this photo.”
For content management, Adobe is using image recognition to help retail teams create content and cut costs for catalog production. Some of Adobe’s e-commerce merchant customers no longer use photo sessions to create their catalogs, instead moving completely to computer-generated images. Other vendors have created content tagging solutions for e-commerce merchants that use AI to automatically assign meta tags to merchandise and categorize it for the search function, replacing the time-consuming and expensive human-powered approach.
Digital experience marketing leverages customer usage analytics, such as engagement, time onsite, click-through rates, search patterns, real-time activity, product preferences, purchase history, past queries, or help tickets to create a profile specific to each customer. These models can then predict customer wants and needs in a granular fashion. For retail, CRM efforts have increasingly focused on how to deliver personalized customer experiences as the best way to grow top-line revenue. AI-driven solutions tap the power of deep learning to pull data from wide-ranging domains, including social media, demographics, and personal behaviors, such as previous purchases, abandoned shopping carts, and website navigation, to create highly personalized product recommendations, promotions, and marketing. The scalable analysis of data from disparate domains and predictive analytics are key to this rapidly expanding use case.
2023
2024
2025
Image recognition, classification, & tagging
Ad insertions into images & video
Intelligent CRM systems
Ad targeting & media buying
Digital experience marketing
Video content analysis
Querying image content
Human emotion analysis
Creative workflow intelligence
Mark Beccue, Principal Analyst, AI and NLP, Omdia
BACK TO TOP
Omdia’s AI research portfolio extends from processors and software to practical deployments across industries and applications, empowering our customers to make smart business decisions.
Discover how you could benefit >
$2
$1
$1
$1
$1
$1
$0
$0
$0
(billions)
$2
$1
$1
$1
$1
$1
$0
$0
2024
(billions)
$2
$1
$1
$1
$1
$1
$0
$0
2025
(billions)
