Thinks 1584

Ashu Garg: “jensen’s [AI factory]…phrase isn’t just metaphoric. It’s meant to reframe how we think about compute infrastructure – shifting it from a cost center to a production system. In manufacturing, a factory transforms raw materials into finished goods through a structured, optimized process. The economic value lies in throughput and efficiency: how much output can be generated from each unit of input, whether labor, energy, or capital. Jensen argues that the same logic now applies to AI infrastructure. The inputs to these modern factories are electricity and data. The outputs are tokens – the atomic units of prediction, reasoning, and generation that power AI systems. The new unit of productivity is tokens per second per watt: a measure not only of how fast chips are, but of how efficiently intelligence can be produced at scale. This is more than a branding exercise. It’s a redefinition of what a data center is for. Traditional data centers are designed to retrieve pre-written software and execute it deterministically. AI factories are built to generate software on the fly. The logic is not fixed but emergent. As Jensen put it: “In the past, we wrote the software and ran it on computers. In the future, the computer is going to generate the tokens for the software.””

Wired on Anthropic: “Dario Amodei [is] trying to create an artificial general intelligence—or as he calls it, “powerful AI”—that will never go rogue. It’ll be a good guy, an usher of utopia.”

Arnold Kling: “In Austrian economics, the concept of capital is replaced by “roundabout production.” We can knead bread by hand, or we can produce kneaded bread in a roundabout way by obtaining a bread machine and using that to knead bread. Modern economies use increasingly roundabout production. The processes for producing automobiles or smart phones involve countless steps and use parts that come from many places, including foreign countries. As production becomes more roundabout, it becomes more complex. We become richer as a result of increasingly roundabout production. But the complexity changes our society in ways that are often unpleasant.”

FT (early April): “Roughly half of Musk’s $314bn net worth is now tied to SpaceX — which was valued at $350bn in a recent private tender offer — while his 20 per cent stake in Tesla has dwindled to $100bn. Morgan Stanley estimates Starlink will generate $16.3bn in revenue this year, up 74 per cent from an estimated $9.3bn in 2024, with subscribers almost doubling to 7.8mn from 4.65mn, according to a January report. By comparison, SpaceX’s rocket launch business is forecast to make $5.8bn in 2025 revenue, up 20 per cent from $4.9bn last year. “While it all began with market-leading launch capability, we believe Starlink will be the primary driver of SpaceX’s growth and profitability,” said Adam Jonas, an analyst at Morgan Stanley.  Starlink has said its more than 7,000 low-orbit satellites provide service to 118 countries and territories.”

McKinsey: “B2B sellers often struggle with oversimplified rules, manual customer research, a lack of data integration, or inadequate training on sales tools. AI can help lead them to their “next-best opportunity.” It can process multiple disparate data sources to prioritize possibilities. Gen AI can parse significant amounts of unstructured data (for example, PDFs, flat files, or photographs) to provide advanced recommendations and instructions. Gen AI can also synthesize relevant information about leads onto a consolidated battlecard, allowing sellers to chase their next-best opportunity based on clear, critical information. This use case can significantly accelerate the time-consuming process of conducting account research, mapping relationships, and identifying additional stakeholders.”

Progency: The AI-First Agency of the Future (Part 10)

My Take – 2

Three Core Services

Progency will likely specialise in three interconnected service areas that specifically target different customer segments based on their engagement levels and value potential.

  1. NeoN: Reacquisition Reimagined

For “Test” customers—those who have gone dormant or inactive (90+ days)—the traditional approach has been expensive retargeting through Google and Meta. NeoN offers a revolutionary alternative:

  • An email-based ad network leveraging authenticated identity
  • Brand-to-brand collaboration that eliminates platform intermediaries
  • Precision targeting through PII-based matching
  • Interactive ActionAds that enable frictionless transactions directly within email
  • Performance-based pricing that aligns all parties’ incentives

NeoN transforms reacquisition economics by enabling brands to reach their dormant customers through the active engagement channels of non-competing brands. This creates a dual benefit: one brand “prints money” by monetising their engaged audience while another “saves money” by efficiently reactivating their dormant customers.

  1. NeoMails: Reactivation Through Interactive Engagement

For “Rest” customers—those showing declining engagement (30-90 days inactive)—NeoMails provides a proactive intervention to prevent them from becoming fully dormant:

  • AMP-powered interactive email experiences
  • Atomic Rewards (Mu) that gamify engagement and drive consistent opens
  • Microns that deliver 15-60 second “brain gain” experiences
  • SmartBlocks for zero-party data collection
  • In-email transactions that eliminate the “click-through penalty”

NeoMails transforms traditional email from static broadcasts into engaging, interactive experiences that command attention and drive action—solving the “No Hotline” problem that forces brands into expensive reacquisition cycles.

  1. Neo360: End-to-End Customer Growth

For brands seeking comprehensive optimisation of their customer relationships, Neo360 offers an integrated solution combining platform capabilities with expert services and AI orchestration:

  • Complete ownership of customer journeys across touchpoints
  • N=1 personalisation at scale through AI Twins
  • Automated content creation and optimisation
  • Cross-channel orchestration with real-time adaptation
  • Continuous testing and refinement of engagement strategies

Neo360 typically begins by focusing on maximising value from Rest customers before expanding to Best customer retention and Next customer conversion—creating a seamless ecosystem for sustainable growth.

The Economics of Progency

Progency creates a fundamentally different economic relationship with clients:

Zero Base Fee + Performance

  • Minimal upfront costs for clients
  • Agency compensation directly tied to measurable outcomes
  • Shared risk and reward creates perfect alignment

Measurable ROI

  • Clear attribution through control/test methodologies
  • Transparent performance reporting
  • Accountability for business impact, not just marketing metrics

Long-Term Partnerships

  • Multi-year agreements focused on customer lifetime value
  • Deeper integration with client operations
  • Shared data infrastructure and knowledge building

This economic model transforms agencies from vendors selling services to partners invested in long-term business success—a dramatic shift from today’s project-based, procurement-driven agency relationships.

The Competitive Landscape

Progency’s transformative model will face different competitive threats than traditional agencies:

Traditional Agency Holdcos

  • Struggling with legacy cost structures and business models
  • Attempting digital transformation but constrained by quarterly earnings pressure
  • Unable to make the technology investments needed for true AI-native operations

Tech Platforms

  • Building increasingly sophisticated self-service tools
  • Limited by focus on their own ecosystems rather than cross-channel orchestration
  • Facing privacy headwinds that challenge their targeting models

Management Consultancies

  • Strong in strategy but weaker in execution
  • Growing creative capabilities but lacking proprietary martech
  • Typically using performance-based pricing but not fully retention-focused

In-House Teams

  • Developing deeper customer understanding
  • Challenged by talent acquisition and technology integration
  • Often still acquisition-focused due to organisational structures

Progency will need to navigate this complex landscape by offering a compelling alternative that delivers measurable business results through retention-focused strategies, proprietary technology, and performance-based economics.

Thinks 1583

FT on Microsoft at 50: “Adaptability has been one of the hallmarks of Nadella’s leadership at Microsoft — something he has shown again as the generative AI market goes through convulsions. Despite the tie with OpenAI, the Microsoft CEO was quick to see how models like these might commoditise even the most advanced AI technology. Microsoft was one of the first to back Meta’s open model, Llama, and was offering access to DeepSeek’s R1 reasoning system less than two weeks after its launch.”

Andrew Chen: “In today’s tech landscape, where the product culture has turned so metrics-driven, the biggest opportunities might actually lie in areas that require intuition to discover. Taking an intuitive, qualitative approach can be faster than relying on A/B testing – especially for new products where low user numbers make data collection painfully slow. After all, your initial product direction requires exceptional judgment. Pick the right starting point, and you’ll be miles ahead of someone who chose poorly and tried to iterate frantically to success. The ubiquity of metrics-oriented thinking means that breakthrough opportunities often exist precisely where data-driven product leaders won’t look. This typically means entering mature markets or categories requiring significant upfront investment. After all, if it were simple and immediately measurable, big tech companies would have already pursued it.”

Crime Reads: “Let me be very obvious at the start and say: a murder victim can’t tell you who the killer is. Locked-room mysteries are puzzling because the only person you’re sure was in the room is the one person you can’t ask for testimony. That’s also what makes locked-room plots such challenging things to read or write — a baffling impossibility turns out to be an illusion with a material explanation. An airgun from the empty house across the street, a serpent in the ventilator, a disguise or an accomplice or a clock hand nudged to display a fraudulent time.”

NYTimes: “We’re seeing a meaningful transition in the way work is done in the white-collar world,” said Carl Tannenbaum, the chief economist of Northern Trust. “I tell people a wave is coming.”…A recent academic paper found that software developers who used an A.I. coding assistant improved a key measure of productivity by more than 25 percent, and that the productivity gains appeared to be largest among the least experienced developers. The result suggested that adopting A.I. could reduce the wage premium enjoyed by more experienced coders, since it would erode their productivity advantages over novices.”

WSJ: “After years of building up my ability to articulate nuanced ideas in French, AI had made this work optional. I felt my brain get a little rusty. I was surprised to find myself grasping for the right words to ask a friend for a favor over text. But life is busy. Why not choose the easy path? AI developers have promised their tools will liberate humans from the drudgery of repetitive brain labor. It will unshackle our minds to think big. It will give us space to be more creative.  But what if freeing our minds actually ends up making them lazy and weak? “With creativity, if you don’t use it, it starts to go away,” Robert Sternberg, a Cornell University professor of psychology, told me. Sternberg, who studies human creativity and intelligence, argues that AI has already taken a toll on both.”

Progency: The AI-First Agency of the Future (Part 9)

My Take – 1

As can be expected, I have a very different view on the agency of the future in an AI-first world. (This section builds on my previous writings with assistance from Claude.)

Beyond Consolidation: A New Model Emerges

Reiterating what I wrote earlier: “What’s abundantly clear is that traditional agencies face an existential inflection point—consolidation merely addresses symptoms while ignoring the underlying disease. The Omnicom-IPG merger represents a defensive manoeuvre in an industry fundamentally threatened by both technological disruption and structural inefficiency. The path forward demands more than scale; it requires dismantling the acquisition-centric paradigm that has dominated marketing for two decades and replacing it with a retention-first model that eliminates the systemic waste of repeatedly paying to reach existing customers. The agency of the future will emerge not from incremental adaptation but from radical reinvention—one that transforms marketing from a cost centre driven by ephemeral impressions into a profit engine powered by sustainable customer relationships.”

The Retention Revolution: Agency of the Future

While holding companies struggle with consolidation to combat shrinking margins and technological threats, a fundamentally different agency model is emerging—one built not around creative services or media buying but on the principle of maximising customer lifetime value through retention.

Retention-First Paradigm

The agency of the future – Progency – inverts the traditional marketing funnel by focusing primarily on keeping and growing existing customer relationships rather than the endless (and increasingly expensive) pursuit of new ones. This retention-first approach acknowledges a fundamental truth that adtech-oriented agencies have long ignored: acquiring a customer once and nurturing that relationship delivers far greater ROI than repeatedly paying platforms to reacquire the same customers.

This shift means redeploying resources from wasteful reacquisition to relationship deepening through:

  • Daily engagement with existing customers through owned channels
  • True 1:1 personalisation enabled by AI and zero-party data
  • Frictionless conversion experiences within owned media
  • Customer advocacy programmes that drive referrals (essentially zero-cost acquisition)

For client brands, this transformation promises a dramatic improvement in marketing efficiency, cutting the 70% of budgets currently wasted on reaching customers already in their databases while simultaneously increasing lifetime value from each relationship.

The PEAK Framework: Building the Next-Generation “Progency”

At the core of this new model lies the PEAK framework—a comprehensive approach combining Platform, Experts, AI Agents, and Kaizen continuous improvement methodology to deliver performance-based outcomes rather than just services or software.

Platform (Martech)

Progency doesn’t merely use technologies—it owns or deeply integrates proprietary martech systems that enable:

  • Complete customer data unification across touchpoints
  • Seamless multi-channel orchestration
  • Interactive engagement capabilities like AMP in email (NeoMails)
  • First-party data collection and activation
  • Authenticated identity solutions for precision targeting (NeoN)

Unlike traditional agencies that buy third-party tools or consultancies that implement others’ software, these next-generation agencies build their competitive advantage on proprietary technology stacks.

Experts

Human expertise remains critical—not for routine execution but for strategy, creativity, and relationship management:

  • Vertical industry specialists who understand sector-specific challenges
  • Customer journey architects who design frictionless experiences
  • Data scientists who translate insights into actionable strategies
  • Creative directors who establish brand guidelines and messaging frameworks

The key distinction: these experts focus on high-level strategy and innovation rather than day-to-day campaign management or production tasks, which become increasingly automated.

AI Agents

Perhaps the most transformative element, AI agents form a “Department of One” that handles complex marketing operations at scale without proportional staffing increases:

  • Content agents that generate personalised messaging across channels
  • Journey agents that orchestrate multi-step customer experiences
  • Analytics agents that continuously monitor performance and identify optimisation opportunities
  • Testing agents that autonomously experiment with variations to improve results
  • Segment agents that identify micro-audiences for targeted engagement

These agents work collaboratively under the supervision of an AI Co-Marketer that ensures all activities align with brand guidelines and strategic objectives—effectively performing tasks that would otherwise require dozens of human specialists.

Kaizen (Continuous Improvement)

The final element is a methodology of relentless optimisation:

  • Rigorous A/B testing across all touchpoints
  • Incremental enhancement of customer experiences
  • Data-driven refinement of targeting models
  • Streamlined workflows and processes
  • Constant knowledge sharing across client engagements

This systematic approach ensures that performance improves continuously over time, compounding the value delivered to clients.

Performance-Based Economics

Progency fundamentally reimagines compensation through performance-based models that align agency incentives directly with client business outcomes. Rather than billing for time, media commission, or flat retainers, the agency earns a percentage of the measurable uplift it creates:

  • Revenue share on incremental sales generated
  • Performance fees tied to customer retention metrics
  • Cost reduction bonuses from eliminating AdWaste
  • Value-based pricing linked to lifetime value improvements

This revolutionary model transforms marketing from a cost centre into an accountable investment with measurable returns—much like adtech platforms but focused on retention rather than acquisition.

Thinks 1582

Tom McAllister: “Don’t outsource your life story to A.I…The only thing under my control as a teacher is what I do in my classroom. I will continue to teach students that, whether they go on to write a best-selling memoir or simply scribble in their journals occasionally, we can try to do the work as honestly and earnestly as possible, bringing our full obsessive selves to the page. The act of writing itself can be an act of self-preservation, even one of defiance. That spark of rebellion is our greatest strength, and it’s found nowhere else but within us.”

Tech Buzz China: “Now home to more than half of the world’s humanoid robotics firms, China has seen nearly 100 new companies emerge in the past year alone, driven by rising investment and technological advances. Recent headlines might suggest China is catching up—or even surpassing—the West in this field, but the reality is more complex. While its growth is undeniable, China still lags behind in key areas such as AI sophistication, high-precision manufacturing, and software-hardware integration. That said, the field is still in its infancy, making it difficult to draw clear lines of progress. Humanoid robots, in particular, may not even be the most practical pursuit in robotics. With no clear market leader and no definitive benchmarks for success, it remains uncertain whether China’s surge in humanoid robotics will lead to long-term dominance—especially when it’s still unclear whether humanoid robots are the most viable path for large-scale automation.”

Yann LeCun: ““I am not interested anymore in LLMs. They are just token generators and those are limited because tokens are in discrete space. I am more interested in next-gen model architectures, that should be able to do 4 things: understand physical world, have persistent memory and ultimately be more capable to plan and Reason.”

ET: “Digital creators are evolving from mere brand endorsers to crucial advisors in narrative building and engagement strategies for brands. Both mainstream and niche influencers are securing lucrative consulting roles, reflecting a shift in recognizing the comprehensive value creators bring to brand storytelling and strategy.”

FT talks to Rolls Royce CEO Tufan Erginbilgiç and discusses how he restored growth at the 119-year-old British engineering group: “Rolls-Royce’s transformation is underpinned by a carefully worked out framework of four “pillars” — an approach he has used before. The first of these, Erginbilgiç describes as “holding up a mirror” to make clear to staff the reality of the company’s position..His second pillar is to set a clear, granular strategy that engages employees…He describes his third pillar as “performance management” — which flows from the strategy and is about how you manage the business, with very clear targets. The fourth is to do all three with “pace and intensity” to show you are delivering…If you don’t score quickly, you lose people.”

Jim O’Shaughnessy: “There may not be a secret, easy path to success. But there is still a path. The REAL path is one most people don’t want to take (why do you think so many people play the lottery?) The real path is emotionally hard. The real path requires you to often hold uncomfortable opinions and challenge conventional thought. The real path requires you to learn, learn and then learn some more, without an end date. It requires you to seek out wisdom and the latest thinking continually in a never-ending commitment to gaining knowledge. Note I say knowledge, not information. That’s another part of the real path–after all that reading, listening and watching, you have to critically separate the wheat from the chaff. And given the information tsunami we have today, that’s a task that gets harder and harder. The real path requires you to have a burning desire to not only *know* but to then also *act* on that knowledge.” The Infinite Loops Canon. Also: Writing in the 21st Century

Progency: The AI-First Agency of the Future (Part 8)

Deep Research

I asked ChatGPT Deep Research to perform an in-depth analysis of how agencies are evolving in response to technological advancements, particularly AI, and shifting client demands. Here’s a summary of the key insights. (The full research document is approximately 19,000 words.)

Historical Evolution of Agencies

Traditional agency models began with creative agencies focusing on campaign concepts, media agencies handling placement, and later digital agencies emerging for online marketing. Compensation evolved from the standard 15% commission system to fee-based models, project fees, retainers, and increasingly, performance-based pricing tied to measurable outcomes. Major disruptions included digital transformation, programmatic advertising, data-driven marketing, the shift from mass marketing to personalization, and data privacy regulations.

Current Agency Landscape

Today’s agency ecosystem includes global holding companies (WPP, Omnicom, Publicis, etc.) alongside independent and boutique firms. Agencies face significant challenges:

  • Talent attraction and retention as skilled employees are lured to tech companies or start their own firms
  • Margin pressure from clients negotiating aggressive fees
  • Shorter client relationships as brands frequently review accounts or parcel projects to multiple agencies
  • Competition from consulting firms and in-house agencies (82% of major advertisers have some form of in-house agency)
  • Keeping pace with rapidly evolving technology

Traditional business models are being tested as clients move away from retainers toward project-based or performance-driven engagements. Agencies are responding by specializing in niches, integrating more deeply with client teams, or developing hybrid compensation models.

AI-Driven Future

The “AI-first agency” will likely have a leaner structure with AI handling production and optimization while humans focus on strategy, oversight, and creative direction. Key technologies reshaping agencies include:

  • Generative AI creating text, images, and video
  • Predictive analytics forecasting outcomes and audience behavior
  • AI-driven media buying and optimization
  • Multi-agent systems with specialized AI agents collaborating on different functions
  • AI tools for project management and operations

AI will transform core agency functions: creative teams will use AI as a partner for ideation and production; media planning will become more data-driven with algorithms managing placements in real-time; performance marketing will shift from reporting the past to predicting the future; and strategy roles will focus on decision-making with AI inputs.

Insourcing vs. Outsourcing

AI tools becoming more accessible could accelerate the in-housing trend, as brands find they can achieve similar results with fewer people. Agencies will need to reposition themselves as innovation partners and strategic advisors rather than execution teams. The balance between insourced and outsourced work will vary by company size and region, but agencies must prove their value in an environment where baseline capabilities are higher for everyone.

Future Business Models

Several forward-looking models are emerging:

  • Performance-based pricing aligning agency compensation with client results
  • Subscription-based services offering predictable costs and recurring revenue
  • Hybrid SaaS-agency blends combining proprietary platforms with consultative services
  • Consulting models charging for strategic advice while execution is automated
  • Marketplace platforms connecting clients with talent and AI solutions

Potential Builders of the Future Agency

Multiple players are competing to define the agency of the future:

  • Holding companies have scale, global reach, and client relationships but face legacy structures and talent drain
  • Tech platforms (Google, Meta, Adobe, etc.) possess cutting-edge AI and vast data but lack service DNA and creative heritage
  • Consulting firms (Accenture, Deloitte) offer strategic rigor and C-suite access but may struggle with creative culture
  • AI-native startups bring innovation and agility but face challenges with credibility and resource constraints

Additional Considerations

The essay also highlights several important factors that will shape future agencies:

  • Ethical challenges including data privacy, algorithmic bias, transparency, IP rights, and workforce impact
  • Disruptions to the advertising supply chain as AI streamlines production and buying processes
  • The need for agencies to navigate varied global privacy laws and develop privacy-first strategies
  • A potential shift in focus from customer acquisition to brand building and retention as AI commoditizes performance marketing

Strategic Takeaways

The agency industry is at an inflection point, creating opportunities for innovators. AI will be the centerpiece of future agencies, but success will come from human-AI collaboration rather than replacement. Regional differences will persist, with emerging markets potentially leapfrogging with new models. Ethics and trust will become key differentiators as agencies navigate AI responsibly.

The “Agency of the Future” won’t follow a single model but will share common characteristics: AI-first operations, outcome-based mindsets, integrated services, global yet local capabilities, and ethical frameworks. Those who invest in technology and talent today will become the indispensable marketing partners of tomorrow.

Thinks 1581

Jamin Ball: “M&A has followed an interesting arc over the last ~20 years. A while back companies companies built platforms through M&A. They build organically, but also acquired their way into new products and markets. [Salesforce] is a great example. They built out their core CRM capabilities, but expanded into analytics when they acquired Tableau, expanded into integration when they acquired Mulesoft, and expanded into communication when they acquired Slack. With more recent cloud software companies, there was a lot of organic expansion – ie expanding into new markets and capabilities through internal product development. Datadog is a great example of this. They started with infrastructure monitoring, and then organically built out application performance monitoring (APM), logging, cloud security solutions, ITSM offerings, etc. I do wonder if the pendulum will start to swing back to inorganic expansion as the cloud platform companies who were founded 10-15 years ago start to see growth slow and look at inorganic opportunities to continue their expansion. OR – if earlier stage companies (not startups, but companies in the $5-$20b valuation range) look to inorganic expansion earlier than they would have otherwise in a more M&A friendly environment. There are also a lot of “zombie” point solutions stuck in the ~$100-$300m ARR range who probably don’t have much of an independent future, but would be quite valuable as part of a broader platform. For private companies in this bucket – this may be your window. Better to be on the front end of it, otherwise you risk missing the wave. ”

Jeff Bezos: “What are the inputs to a higher stock price? Okay, well, free cash flow and return on invested capital are inputs to a higher stock price. Let’s keep working backwards. What are the inputs to free cash flow? And you keep working backwards until you get to something that’s controllable.” [via Shane Parrish]

Siddharth Chowdhury: “Every man has in his memory a few songs which facilitate time travel.”

Arnold Kling: “Human society is dynamic. We are always trying out new ideas for products, services, processes, and rules. We can think of innovation as taking place according to three e’s: experimentation, evaluation, and evolution. Experimentation means trying out the new idea. Evaluation means assessing how well the new idea works. Evolution means keeping good ideas and discarding ideas that no longer work or never worked in the first place. People trying to innovate from within the government have to persuade a sluggish bureaucracy to attempt something that might seem risky. In the market, all it takes is willpower and the ability to mobilize resources. This makes experimentation easier in the market, unless we are talking about a project that requires a massive undertaking, such as the Manhattan Project or the first manned moon landing. The market has a natural mechanism for evaluating new ideas and for undertaking evolution. That is the profit and loss system. Ideas that yield more benefits than costs will generate profits and survive. Those that fail to do so will generate losses and be unsustainable.”

Astasia Myers: “We think web AI agents are the next frontier that really helps with the end-to-end automation of human tasks. [W]eb AI agents are this dynamic bridge between static pre-trained models that are mostly text-focused in the ever-changing digital landscape.”

Progency: The AI-First Agency of the Future (Part 7)

Who – 2

ChatGPT

If I had to bet on one entity to successfully build the Agency of the Future, it would be Adobe.

Why Adobe?

  1. AI and Creative Fusion: Adobe has deep domain expertise in creativity and is already pioneering generative AI through Adobe Firefly, Adobe Sensei, and Creative Cloud products. It uniquely blends creative intuition with technological sophistication.
  2. Martech Dominance: Adobe Experience Cloud (Marketo, Adobe Analytics, Adobe Campaign, Adobe Real-Time CDP) gives it unrivalled access to customer data, analytics, and the capability to scale hyper-personalisation seamlessly.
  3. Existing Trust and Brand Credibility: Adobe has global brand recognition, strong existing relationships with Fortune 500 CMOs, and a trusted position as an enterprise partner—crucial for becoming the strategic agency partner of the future.
  4. Ability to Transition from Tools to Agency: Adobe already offers extensive consulting, managed services, and strategic advisory. Transitioning from tool-provider to agency-provider can be a natural evolution, allowing Adobe to own the full creative-to-consumer value chain.
  5. Financial Muscle and Strategic Vision: Adobe is financially strong, has successfully navigated previous technological shifts (print → digital → mobile → AI), and has consistently demonstrated willingness to invest and pivot aggressively towards market leadership.

Close Competitors:

  • Salesforce (strong martech/data presence, less creative DNA)
  • Accenture Interactive (deep consulting capabilities, weaker native AI tech stack)
  • Google/Meta (powerful tech stacks, but limited agency-style personalised customer journeys)

However, Adobe’s unique intersection of creative mastery, powerful martech stack, and advanced AI capabilities makes it the most compelling bet to build the definitive Agency of the Future.

DeepSeek

If forced to pick a single contender most likely to build the Agency of the Future, Publicis Groupe stands out as the best bet—not because it’s flawless, but because it’s aggressively reengineering itself to balance legacy strengths with future-proof agility. Here’s why:

  1. Strategic Acquisitions & Tech Investments

Publicis has spent the past decade acquiring capabilities that legacy peers are still scrambling to replicate:

  • Epsilon ($4.4B in 2019): A data powerhouse with 250M+ consumer profiles, enabling hyper-personalised AI-driven campaigns.
  • Sapient ($3.7B in 2015): A digital transformation consultancy, bridging marketing and enterprise tech.
  • CoreAI (2024): Its proprietary AI platform, trained on Publicis’ 2.5 petabytes of data, automates everything from media planning to creative ideation.

These moves position Publicis not just as a “marketing agency,” but as a data-driven tech integrator.

  1. Decentralised Agility

While rivals like WPP remain siloed, Publicis has restructured into the “Power of One” model:

  • Agencies (Leo Burnett, Saatchi & Saatchi) collaborate under a single P&L, breaking down internal competition.
  • Marcel, its AI-powered talent platform, connects 100,000+ employees globally, enabling fluid team-building for client projects.
  • Micro-agencies like Publicis Sapient and Publicis Spine act as nimble innovation labs, bypassing holding-company bureaucracy.

This structure mirrors the “network of micro-agencies” future—big enough to scale, agile enough to pivot.

  1. AI Fluency Without Losing Creativity

Publicis avoids the “soulless AI” trap:

  • Its Creative at Scale division pairs generative AI tools with human creatives to produce 10x more content (e.g., personalised Coca-Cola ads).
  • Partnerships with Adobe and Microsoft integrate Firefly and Azure AI into workflows, but campaigns retain emotional punch (e.g., award-winning work for Heineken, Samsung).
  • Ethical guardrails: Publicis openly audits AI models for bias and trains staff to use AI responsibly—a trust advantage over startups and tech giants.
  1. Consulting-Like Client Access

Publicis mimics consultancies’ C-suite influence:

  • Sapient advises CEOs on digital transformation, embedding Publicis deeper into client operations.
  • Dynamic Client Solutions offers AI-powered performance guarantees (e.g., “We’ll cut customer acquisition costs by 20% or refund the difference”).

This shifts Publicis from a vendor to a growth partner, competing directly with Accenture Song.

  1. First-Mover in the AI Talent War

Publicis is poaching talent from both tech and creative worlds:

  • AI engineers from Google and Meta to refine CoreAI.
  • Storytellers from Hollywood and TikTok to humanise AI outputs.
  • Upskilling programs: 100,000 employees trained in AI prompt engineering and data ethics by 2024.

The Risks

Publicis isn’t invincible:

  • Integration challenges: Melding Epsilon’s data with creative teams remains messy.
  • Overpromising on AI: Clients may revolt if campaigns feel robotic.
  • Consultancy competition: Accenture Song’s $18B revenue dwarfs Publicis’ consulting arm.

Why Publicis Over Others?

  • Tech Giants (Google/Meta): Lack creative credibility and face brand distrust.
  • Boutiques: Too niche to handle global, integrated campaigns.
  • WPP/Dentsu: Still shackled by legacy structures and slower AI adoption.
  • Accenture: Strong on data, weak on creative emotional

Publicis uniquely merges enterprise-grade tech, creative firepower, and operational agility—the trifecta needed to dominate an AI-first world.

Final Take: The Agency of the Future must be part-creative studio, part-consultancy, part-tech lab. Publicis Groupe is the only legacy player actively becoming all three. If it continues to decentralise, invest in ethical AI, and prioritise outcomes over billable hours, it could render the “holding company vs. boutique” debate obsolete—and redefine marketing itself.

Thinks 1580

WSJ: “Ciara Greene and Gillian Murphy argue in “Memory Lane” that our power of recollection is something of a marvel, its flaws simply the downsides of its strengths. Consider that the primary purpose of memory is to retain information likely to be of value in the future. Given our limited neurological resources, that requires us to forget anything inessential. A good memory is by necessity a bad one, in the sense that it must fail to recall most of what we have experienced. Forgetting has other benefits. We say that time heals, but it is the fading of memories, not the passing of days, that allows us to get over heartache and trauma.

WaPo: “A great hobby takes simple ingredients — a skein of yarn, a piece of lumber, a group of friendly strangers — and alchemizes them into something that makes day-to-day life better. But the relentless push toward self-optimization, the pressure to monetize every passion, and the pull of consumerism can tarnish the joy of hobbies.”

Sam Altman: “My favorite historical analog is the transistor for what AGI is going to be like. There’s going to be a lot of it, it’s going to diffuse into everything, it’s going to be cheap, it’s an emerging property of physics and it on its own will not be a differentiator…Where I think there’s strategic edges, there’s building the giant Internet company. I think that should be a combination of several different key services. There’s probably three or four things on the order of ChatGPT, and you’ll want to buy one bundled subscription of all of those. You’ll want to be able to sign in with your personal AI that’s gotten to know you over your life, over your years to other services and use it there. There will be, I think, amazing new kinds of devices that are optimized for how you use an AGI. There will be new kinds of web browsers, there’ll be that whole cluster, someone is just going to build the valuable products around AI. So that’s one thing. There’s another thing, which is the inference stack, so how you make the cheapest, most abundant inference. Chips, data centers, energy, there’ll be some interesting financial engineering to do, there’s all of that. And then the third thing is there will be just actually doing the best research and producing the best models. I think that is the triumvirate of value, but most models except the very, very leading edge, I think will commoditize pretty quickly.”

Gulzar: “The middle-class squeeze is being felt in India too. India’s fundamental economic problem of a narrow consumption base is compounded by an economy which is not creating the number of good jobs required to quickly expand the middle-class base to support sustained high growth rates…The vast majority of job creation is in gig work and the likes of construction, security guards, and housemaids, all of which generate monthly incomes in the range of Rs 15000-20000 and have limited productivity improvement opportunities and occupational mobility.”

Tyler Cowen on The Great Forgetting: “One example would be a lot of individuals today—some of them even economists—think rent control is maybe a decent idea. Several decades ago, virtually everyone would have realized that rent control creates shortages, quality problems, corruption. But bad ideas have been making a comeback. The use of excess inflation to try to stimulate the economy. Or the notion that high crime is not so bad for your city’s economy. We learned that lesson, but we seem to be forgetting that again.”

Progency: The AI-First Agency of the Future (Part 6)

Who – 1

To a specific question about who can build the agency of the future, I got varying answers from the AIs. Three AIs give three different answers!

Claude

If I had to select just one entity best positioned to build the agency of the future, I would choose Accenture.

Accenture represents the optimal blend of capabilities needed to create the AI-first agency:

  1. Resource Advantage: Their $3 billion investment in AI through 2030 and development of 50+ industry-specific AI agent solutions demonstrates both commitment and capacity to lead this transformation.
  2. Hybrid Business Model: They’ve already bridged the consulting-creative divide through Accenture Song (formerly Accenture Interactive), acquiring creative agencies while maintaining their strategic consulting foundation.
  3. Enterprise AI Expertise: Their AI Refinery platform and specialised agent builder position them at the forefront of enterprise-grade AI implementation across industries.
  4. C-Suite Relationships: Unlike pure creative agencies or tech startups, Accenture has established trust at the highest levels of global organisations, giving them unique influence over AI transformation roadmaps.
  5. Scale and Talent: They have both the global reach and the ability to attract specialised AI talent that smaller players cannot match.
  6. Practical Implementation Track Record: With 7% market share in generative AI services and hundreds of successful AI implementations, they’ve demonstrated ability to move beyond theory to practical application.

What Accenture lacks in creative DNA compared to traditional agencies, they can acquire through strategic purchases. What they lack in foundational AI model ownership compared to tech giants, they compensate for with implementation expertise across multiple platforms.

The agency of the future requires blending strategy, technology implementation, creative execution, and ethical governance – precisely the combination Accenture has been assembling through its evolution from pure consulting into a hybrid strategic-creative-technical powerhouse.

I then asked for additional info on Accenture Song.

Accenture’s existing agency business operates primarily through Accenture Song (formerly Accenture Interactive), which has grown to become one of the world’s largest digital agencies through a strategic acquisition strategy and integrated service model.

Accenture Song: Evolution and Scale

Accenture launched Accenture Interactive in 2009 as a digital marketing services division. In April 2022, they rebranded it as Accenture Song to better reflect its expanded creative capabilities. The unit now generates approximately $15-16 billion in annual revenue, making it comparable in size to major advertising holding companies.

Acquisition Strategy

Accenture has built its agency capabilities through an aggressive acquisition strategy, spending billions to acquire over 40 creative, design, and marketing agencies globally. Key acquisitions include:

  • Droga5 (2019) – One of their most significant acquisitions, bringing high-profile creative talent and award-winning campaigns
  • Karmarama (2016) – A major UK creative agency
  • The Monkeys (2017) – A leading Australian creative agency
  • Fjord (2013) – A global design and innovation consultancy
  • MediaHound (2022) – A content intelligence platform
  • Adaptly (2018) – A digital media buying agency
  • Rothco (2017) – An Irish creative agency

Integration Model

Unlike traditional holding companies that maintain separate agency brands, Accenture has generally integrated acquisitions under the Song umbrella. Their model connects:

  1. Creative capabilities (campaign development, content creation)
  2. Experience design (UX/UI, customer journeys)
  3. Technology implementation (platforms, data systems)
  4. Business strategy (leveraging Accenture’s consulting expertise)

Differentiated Positioning

Accenture Song has positioned itself differently from traditional agencies by:

  1. End-to-end services: Offering everything from strategy through implementation and ongoing optimisation
  2. Technology integration: Connecting creative work directly to commerce, data, and enterprise systems
  3. Business transformation: Focusing on business outcomes rather than just campaign metrics
  4. Data capabilities: Leveraging Accenture’s analytics expertise for personalisation and measurement

AI Integration

Accenture has been integrating AI into Song’s offerings through:

  • Content production: Using generative AI to scale creative outputs
  • Experience optimisation: AI-driven personalisation engines
  • Media automation: Advanced algorithmic approaches to media buying
  • Creative analytics: AI tools to predict campaign performance

Challenges and Criticisms

The Accenture agency model has faced some criticism around:

  1. Creative culture: Challenges in maintaining creative excellence within a consulting-driven organisation
  2. Integration complexity: Difficulties in fully integrating acquired agencies
  3. Talent retention: Some creative talent departures after acquisition
  4. Identity tensions: Balancing consulting methodologies with creative processes

Despite these challenges, Accenture Song has established itself as a legitimate competitor to traditional agency networks, particularly for clients seeking integrated technology, business strategy, and creative services as they navigate digital transformation and now AI adoption.