Until about a year ago, you would type a search question into Google. You’d get a list of roughly 10 results (usually out of some incredible number like seventy-four point eight million – whether that number could ever be verified is a moot point). Today however, we are regularly served up a sage-like text summary, prepared for us by someone (actually something) called Gemini.

Changes like this are having profound economic effects for advertisers. Digital ads are being displayed in different ways, to different people, in different places and at different times than what we have come to understand and expect over the last two decades. Our invaluable, predictive advertising models and forecasts are being torn to pieces.

Today’s digital media owners (principally the oligopoly of Google, Meta and Amazon, which represent over 50% of all global ad revenues) are also going to be challenged. In the way that Nvidia has dethroned Intel, might SearchGPT or DeepSeek slay Google? (You could argue Intel was committing hara-kiri anyway). TikTok, with a staggering 1.9bn users, is taking megabytes [sic] out of Meta’s lunch, breakfast and dinner (ByteDance is the owner of TikTok – at least for now). And Amazon’s core e-commerce experience is under threat from new, AI driven shopping platforms like Shopify, WooCommerce and Wix.

In this article we explore how these changes will affect the performance of your brand. We predict it’ll be the biggest change in brand and product marketing since search engines invented paid media advertising in the mid-nineties and social media created interest-targeted advertising platforms in the early noughties.

Costs will increase and your brand visibility will reduce as a result of AI controlling advertising platforms

And as the implications of DeepSeek’s incomprehensibly efficient content generator and search engine become better understood, we’ve included a screen video of its response to our brief for this article, instead of a ChatGPT version: in our view, it’s a step up.

(Warning – beware using DeepSeek: its terms and conditions give it the right to hoover up all of your personal data from your phone and store it in China. Excerpts from their terms and conditions are included at the end of this article).*

The Gemini is out of the bottle

Since we last wrote about ChatGPT taking the professional world by storm (please click here for full article), 39% of Americans say they use generative AI (28% for work, and 11% every day) thanks, the Economist. Goldman Sachs forecast that 300m white collar jobs are at risk from AI. Despite this, companies are reporting limited success:

But corporate demand for generative AI skills outstrips supply. A sales rep with AI skills can earn $45,000 a year more than one who lacks them. (Thanks Lightfoot).

There has been a flurry of new generative AI entrants, product announcements, technological advances and eye-popping investments. And tech stock highs and lows. Flash thinking, image, animation and motion video generation, code generation, agentic AI and agent bots are all coming thick and fast, and in many forms. The disruption to social media, search, e-commerce and traditional and digital advertising will be significant.

SearchGPT, the new search engine from Open AI that uses Microsoft’s search engine data and technology (Bing) is for sure going to challenge Google for its search dominance: it arguably offers a simpler, more joined-up and efficient search experience to the user. ChatGPT’s brand (now 2 years old) already attracts approximately 5% of the traffic of Google (26 years old) (thanks Similarweb).

And DeepSeek, launched just a month ago, is now threatening to dethrone ChatGPT. It reportedly had more app downloads than ChatGPT in Apple’s App Store last month. Llama from Meta looks left behind in the dust.

Google is having to respond to the threat to its core advertising business, which in turn is going to cause severe disruption to businesses that rely on Google for new customers and growth.

For brand advertisers, Gemini from Google will have a profound impact on the economics of search advertising.

Gemini is not granting advertisers their wishes

Google’s genius was to charge advertisers for clicks on ads that were matched to the individual searcher’s questions. Advertisers would bid for these clicks: the higher they bid, the higher up the page their ad appeared and the more clicks they received. And the higher up the page your ad appears, the higher the quality of the engagement is with those ads. Various search engines joined the party, but only Microsoft’s Bing search engine captured any meaningful share (about 10% today).

ChatGPT, on the other hand, charges a subscription to users of its more advanced generation and search tools: the revenue model is customer driven, as opposed to Google’s advertiser driven model. DeepSeek (so far) is free. At some point, it will need to start earning revenue, despite its reported development costs being one sixth of ChatGPT’s.

Last month OpenAI launched Operator, an agent that helps users make purchases online. It also launched a new AI “agent-bot” that can help users on its ChatGPT platform generate in-depth research reports on complex topics. Deep Research, as the service is called, scours the internet for information, including images and PDFs and provides in-depth analysis of the data it finds.

Investors like SoftBank are making big bets on huge data centres that will power this new generation of AI. It’s going to be a volatile time (witness Nvidia’s 18% share price drop the day DeepSeek announced its new AI generation app).

To respond to all of this, in early 2024, Google started introducing descriptive results at the top of the search pages, generated by Gemini, Google’s generative AI technology. Gemini aggregates all of the information and data that Google has “scraped” from the web and intelligently transposes it into understandable sentences which it then presents to a search query in its search engine results. Clever stuff.

However, by adding this aggregated information to the top of its results pages, Google is disrupting the fundamentals of its advertisers’ economics. Pushing advertisers ads (sponsored links) and natural listings further down the search pages reduces the number of clicks. Much like ads being placed in print titles, the position of an ad in Google’s results massively affects the ad’s performance (along with creativity, targeting and a host of other factors).

The traditional search page, when disrupted by Gemini reduces the quality and volume of clicks, messes up forecasting models and will increase advertising costs

FirstPageSage reports that the Gemini AI results at the top of the search page attract approximately 39% of all of the clicks, when it appears. This means hypothetically 39% less traffic for advertisers. Each position that an advertiser’s ad is “bumped” down the page, leads to a significant reduction in the number of clicks (up to 40% for example from position 1 to 2), much like adverts in newspapers or magazines seeing engagement drop off as readers reach the latter sections of their publications.

This demotion of ads and hard-earned positions means as advertisers, we have to increase the frequency of our ads (which are now further down the page) and therefore, to spend more in order to maintain the same number of clicks. And although never proven, many of us in ad land have long suspected that Google controls the price it charges for those clicks and that it’s not a pure, advertiser-led, auction model.

Another brick in the wall

Google has also introduced an AI based ad and campaign management tool called PMAX (short for Performance Max). PMAX decides which ads to place where, and to whom, across Google’s media (Google Search, YouTube, Gmail, Google Maps, as well as its network of millions of other websites which display banner ads).

The new happy couple, Gemini and (P)Max, are promising to transform the economics of digital advertising. But given some of the reported “glitches” in their results, there is much turbulence and disruption ahead, as addmustard’s performance team can attest to.

Other digital channels and platforms are rapidly evolving new, generative AI technologies too, particularly social media.

A “slop”in the face for brands

With ads, content and even user accounts now being generated and targeted en masse by generative AI tools, brands need to be really concerned about authenticity.

“Slop” is the term for AI generated text and images (and other content) which are patently not true, poor quality and misrepresentations. Such as Google Gemini’s top image result of Shakespeare, which seemingly accurately depicts the bard…with six fingers on each hand.

Or Labour’s disastrous AI generated TikTok video, promoting its public service agenda. It includes a soundtrack, with lyrics encouraging men to give drugs to “naughty” young girls. It’s been withdrawn, but left lasting damage to the Labour Party’s brand.

The end of Meta’s fact-checking will lead to more “fake” content. Good news for lawyers. Bad news for brands.

What will the effects of the increasing amount of slop, fake information, misleading data and plain rubbish clogging up social media be? And how will your brand appear authentically amongst all this detritus?

Across all digital media platforms, media owners like Google, Meta and TikTok are building models to create a deeper understanding and context for what you are trying to do on the web and to match it with more relevant results.

All of these models are based on historic data. So how will they create new ideas, innovate and develop our communications in distinctive directions?

Despite the recent commentary about generative AI “hallucinating”, it runs the risk of homogenising brand and other forms of marketing communications. Whilst AI can significantly reduce the time and cost of creating and producing ads and content, and even enhance some aspects of it, there is an increasing body of evidence that AI is generating bland and ineffective creative output.

Standout brand communications require diligence, skill, process and craft: rigorous audience analysis, understanding and insight; a thorough and diligent process for exploration, creative development and thinking; experience, talent and breadth of expertise; testing, data and analysis. And time.

If you’ve been fortunate enough to see the ABBA Voyage show, in London’s East End, you’ll have seen the power and potential of diligent, authentic AI generation. An investment of £120m has relaunched the ABBA brand back to global fame. What you may not know is that the human effort, talent and skill involved in designing, scripting, directing, producing and editing the avatars of Agnetha, Frida, Benny and Bjorn involved over 1,000 professionals and not just the rumoured billions of hours of AI computer processing.

Human talent, craft and creativity is core to the extraordinary show’s screenplay, the choreography and its staggering success. The extraordinary show demanded obsessive attention to detail and required “people who know what they are doing”, like the Royal Ballet’s resident choreographer and a live, ten-piece band of world-class musicians.

The problem with using the current form of generative AI models is that they are all based on what has gone before. The very nature of their structure and modus operandi limits them from “thinking” laterally or creating something new. AI cannot judge if the ideas, the work or the output is original.

Joining the dots

There is real excitement about this rapidly developing technology. There’s no doubt it will change many aspects of our professional and private lives. The ways in which brand and product communications are devised, created, produced, targeted, displayed and measured can be significantly enhanced by the use of these tools. They are comprehensive and efficient.

But they are tools. None of them are a panacea.

And as we said in our original analysis of OpenAI’s early version of ChatGPT “A fool with a tool is still a fool”.

The brand basics of awareness, interest, desire and action are as relevant in a Generative AI world as they have always been.

Devising a great brand is one thing. Driving awareness and visibility is another. Making your brand relevant and appealing to Generative AI requires investment in unique and original content; authenticity; subject authority; depth and breadth; and relevance.

And in a world where our customers are bombarded with communications for hours every day and which are often homogenised by AI, having something distinctive and compelling to say that resonates with your audience becomes even more important to creating brand value. This new breed of tools will reward unique and visible brands even more than search engines and social media did before.

Intuition, instinct, lateral thinking, humour, curiosity, emotional intelligence, abstraction, accuracy, context, inference, ethical and moral judgment, critical thinking, collaboration and most importantly of all, creativity are remarkable characteristics of humans and should also be celebrated. Frequently.

They are not in the capabilities of the current generative AI tools.

They are, however, essential to building lasting value in your brand.

Get in touch

Copy of DeepSeek’s screen video, output and excerpts from their terms and conditions are included below:

Our brief to DeepSeek:

DeepSeek Response:

DeepSeek’s text transcript is here:

Generative AI and AI-driven tools like Google’s **Gemini**, **Performance Max (PMAX)**, **SearchGPT**, and **Operator** are transforming digital marketing by enhancing efficiency, personalization, and scalability. Below is an analysis of their likely economic, financial, and effectiveness impacts:

—

### **1. Economic Impacts**

#### **Positive Impacts:**

– **Increased Productivity**: AI tools automate repetitive tasks like ad creation, keyword research, and campaign optimization, freeing up marketers to focus on strategy and creativity.

– **Cost Efficiency**: By optimizing ad spend and reducing manual labor, businesses can achieve better ROI with lower operational costs.

– **Market Expansion**: AI enables hyper-targeted campaigns, allowing smaller businesses to compete with larger players and enter new markets more effectively.

– **Job Creation in AI and Data Science**: While some traditional marketing roles may decline, demand for AI specialists, data analysts, and prompt engineers will grow.

 

#### **Negative Impacts:**

– **Job Displacement**: Routine tasks in digital marketing (e.g., copywriting, basic ad management) may be automated, potentially reducing demand for entry-level marketing roles.

– **Concentration of Power**: Large tech companies like Google may dominate the AI marketing space, creating barriers to entry for smaller players and reducing competition.

– **Increased Costs for AI Adoption**: Smaller businesses may struggle to afford advanced AI tools, creating a divide between AI-enabled and non-AI-enabled companies.

—

### **2. Financial Impacts**

#### **For Businesses:**

– **Higher ROI on Ad Spend**: Tools like **PMAX** and **Gemini** optimize campaigns in real-time, improving conversion rates and reducing wasted ad spend.

– **Reduced Operational Costs**: Automation reduces the need for large marketing teams, lowering labor costs.

– **Scalability**: AI allows businesses to scale campaigns globally with minimal additional resources.

 

#### **For Google and Tech Companies:**

– **Revenue Growth**: AI-driven tools like **PMAX** and **SearchGPT** will likely increase Google’s ad revenue by attracting more advertisers and improving campaign performance.

– **Market Dominance**: By integrating AI into its ecosystem, Google can strengthen its position as the leading digital advertising platform.

– **Subscription and Licensing Revenue**: Google may monetize advanced AI features through premium subscriptions or licensing fees.

 

#### **For Consumers:**

– **Higher Advertising Costs**: As AI improves targeting and effectiveness, competition for ad space may drive up costs, particularly in high-demand sectors.

– **Privacy Concerns**: AI’s reliance on data may lead to stricter regulations, increasing compliance costs for businesses.

—

### **3. Effectiveness Impacts**

#### **Enhanced Targeting and Personalization:**

– **Hyper-Personalized Ads**: AI tools like **Gemini** and **SearchGPT** can analyze user behavior and preferences to deliver highly relevant ads, improving engagement and conversion rates.

– **Dynamic Creative Optimization**: Generative AI can create thousands of ad variations tailored to different audiences, increasing campaign effectiveness.

 

#### **Improved Campaign Performance:**

– **Real-Time Optimization**: Tools like **PMAX** use AI to adjust bids, targeting, and creatives in real-time, maximizing performance across channels.

– **Cross-Channel Integration**: AI enables seamless integration of campaigns across search, display, video, and social media, ensuring consistent messaging and better results.

 

#### **Data-Driven Decision Making:**

– **Predictive Analytics**: AI can forecast trends, customer behavior, and campaign outcomes, helping marketers make data-driven decisions.

– **A/B Testing at Scale**: AI can automate and scale A/B testing, quickly identifying the best-performing strategies.

 

#### **Challenges:**

– **Over-Reliance on AI**: Marketers may become overly dependent on AI, potentially neglecting creative and strategic thinking.

– **Ad Fatigue**: Highly personalized ads may lead to ad fatigue if not managed carefully, reducing their effectiveness over time.

– **Ethical Concerns**: AI-generated content and targeting may raise ethical issues, such as bias or manipulation, which could harm brand reputation.

—

### **4. Specific Impacts of Key Tools**

#### **Google’s Performance Max (PMAX):**

– **Efficiency**: PMAX automates campaign management across Google’s ecosystem, improving efficiency and reducing manual effort.

– **Scalability**: Businesses can scale campaigns across multiple channels with minimal additional resources.

 

#### **Google’s Gemini:**

– **Creative Enhancement**: Gemini’s generative AI capabilities can create compelling ad copy and visuals, improving engagement.

– **Cost Savings**: Automating creative production reduces the need for expensive design and copywriting resources.

 

#### **SearchGPT (Hypothetical Tool):**

– **Enhanced Search Ads**: If integrated into Google Ads, SearchGPT could generate highly relevant search ad copy and landing pages, improving click-through and conversion rates.

– **Natural Language Understanding**: Improved understanding of user intent could lead to better targeting and ad relevance.

 

#### **Operator (Hypothetical Tool):**

– **Automated Customer Interactions**: If Operator is an AI-driven customer service tool, it could enhance lead nurturing and customer retention through personalized interactions.

– **Cost Reduction**: Automating customer service and lead management could significantly reduce operational costs.

—

### **5. Long-Term Implications**

– **Shift in Marketing Roles**: Marketers will need to upskill in AI and data analytics to remain competitive.

– **Increased Competition**: As AI levels the playing field, businesses will need to focus on creativity and strategy to differentiate themselves.

– **Regulatory Changes**: Governments may introduce stricter regulations on AI-driven advertising, particularly around data privacy and transparency.

—

### **Conclusion**

Generative AI and AI-driven tools like Google’s **Gemini**, **PMAX**, and others are poised to revolutionize digital marketing by improving efficiency, personalization, and scalability. While they offer significant economic and financial benefits, challenges such as job displacement, ethical concerns, and increased competition must be addressed. Businesses that embrace these technologies while maintaining a focus on creativity and strategy will likely gain a competitive edge in the evolving digital landscape.

DeepSeeks data privacy laws:

*We searched DeepSeek’s terms and conditions to see how it stores users data (like name, data of birth, email address – using ChatGPT – and these are the extracts to consider before using it.

*“DeepSeek’s Terms of Use and Privacy Policy outline the company’s rights to access and store users’ personal data.”

“Data provided during account creation, such as date of birth, username, email address, telephone number, and password”.“We store the information we collect in secure servers located in the People’s Republic of China.”

“We may collect your text or audio input, prompt, uploaded files, feedback, chat history, or other content that you provide to our model and Services.”

Follow us