The Collision: A Strategic Blueprint for The Curious Executive

TL;DR — Part 1

The word ‘agency’ has two meanings — and they’re colliding.

AI systems are gaining agency (autonomous action capability) while traditional agencies (marketing services firms) face existential disruption. The agentic AI market will hit $139–199 billion by 2034.

The ‘SaaSpocalypse’ of February 2026 wiped $285 billion from software valuations in 48 hours — because per-seat pricing collapses when AI agents do the work of multiple humans.

Meanwhile, AI-native disruptors like Sierra ($100M ARR in 21 months), Artisan (AI SDRs at $999/month vs. $150K/year for humans), and Jasper (70,000+ customers with 100+ AI agents) are proving that the unit economics of traditional agencies are artefacts of a world that’s rapidly ending.

The question isn’t whether transformation will happen — it’s whether incumbents can adapt fast enough.

A Note on Agency Perspective in A&M industry.

I have been at the forefront of Digital/Mobile marketing for over a decade — through stints at Paytm (marketing solutions and ads business) and WPP (Madhouse) — and I have witnessed the evolution of platform/product/solution providers and Advertising, Creative & Media agencies as a practitioner. At Proximus Global, I've brought together cross-cultural, cross-geographic, and cross-entity teams as part of our transformation. I have spent hundreds of hours with the world’s who’s who in Marketing transformation — from product players like 6Sense, Clay, Salesforce, Adobe, and HubSpot to agencies like WPP, Publicis, Dentsu, and Accenture Song. What you read in this three-part series presents the various views on Agency, Client, Platforms, Solutions, and the Impact of AI across them in this disruptive environment.

This blueprint reflects my learning and offers a view of the Future of Work in the A&M business!

The Double Meaning That Changes Everything

Here’s an irony worth unpacking: The very technology that grants machines agency — the ability to act autonomously toward goals — is simultaneously dismantling the trillion-dollar agency (marketing services firms) business model we’ve built over decades. Two meanings. One word. An industry-defining collision.

In the boardrooms from London to Bangalore, from São Paulo to Singapore, this collision is creating what I call the Agency Paradox: The more autonomous our AI systems become, the less we need the humans who used to do that work — and yet, paradoxically, the more we need a new kind of human to orchestrate what comes next.

The numbers tell the story. The global agentic AI market hit $7.3–7.5 billion in 2025 and is projected to reach $139–199 billion by 2034, growing at a 40–44% CAGR. But here’s the figure that should keep traditional service businesses awake at night: 40% of enterprise applications will embed task-specific AI agents by the end of 2026, up from less than 5% in 2025. This isn’t a feature upgrade. This is a platform shift.

And it’s not just theory. Sierra, the AI customer service startup founded by former Salesforce co-CEO Bret Taylor, hit $100 million ARR in just 21 months — faster than even its founders expected. Artisan, an AI SDR startup founded by 23-year-old CEO Jaspar Carmichael-Jack, raised $46 million and replaced the cold outreach function that previously required teams of entry-level salespeople. Jasper, the AI marketing platform, now serves over 70,000 paying customers with 100+ specialised AI agents.

The disruptors aren’t waiting. The question is whether the incumbents can adapt fast enough.

The SaaS Reckoning and What It Signals

To understand where agency and platform businesses are headed, we need first to understand what’s happening to SaaS — because the same forces are coming for every service-based model.

The so-called ‘SaaSpocalypse’ of February 2026 wiped approximately $285 billion from software stock valuations in a single 48-hour window. Within seven days, the total market capitalisation loss expanded to roughly $1 trillion. Not because AI failed — but because it succeeded too well.

The trigger? Institutional investors reclassified SaaS companies after realising that per-seat pricing models are collapsing as AI agents begin to do the work of multiple human employees. When one user equipped with AI agents can accomplish the work of five, seat-based pricing becomes absurd.

Publicis Sapient reports actively reducing traditional SaaS licences by approximately 50% — including major platforms like Adobe — by substituting them with generative AI tools and chatbots. This isn’t a one-off; it’s a pattern.

What replaces seat-based pricing? The shift is toward three new models:

Consumption-Based: Pay for what you use. Every API call, every generation, every action has a price tag. This rewards efficiency and penalises waste.

Outcome-Based: Pay for what’s achieved. Did the campaign generate leads? Did the agent resolve the customer’s issue? Results, not activities, drive compensation.

Value-as-a-Result (VaaR): Pay for the business outcome delivered. Revenue growth, cost reduction, market share gain — the ultimate alignment of incentives.

Deloitte predicts that by 2028, 70% of software vendors will have refactored their pricing strategies to align with these new value metrics. IDC goes further: by 2028, pure seat-based pricing will be obsolete. The fundamental shift: stop charging for access, start charging for work done.

The Marketing Agency Inflexion Point

If you run marketing as a relay race between specialised teams, you will be outperformed by organisations that run it as a control room overseeing agentic AI workflows.

According to HubSpot’s 2026 State of Marketing report, 61% of marketers believe marketing is experiencing its biggest disruption in 20 years, driven by AI. And they’re right — AI is compressing the cost of content production so dramatically that a thousand ad variants now cost what five used to. That blows up the logic of charging by the unit.

The traditional agency value chain — strategy → creative → production → media → measurement — is being compressed and automated at every stage.

What’s automating now: Content volume, speed, and variation (approaching zero marginal cost). Campaign optimisation across channels. A/B testing and performance iteration. Basic creative generation and adaptation. Reporting and analytics synthesis.

What’s emerging: AI agents that don’t just report on marketing performance but actively improve it. Multi-agent orchestration handling end-to-end campaigns. Answer Engine Optimisation (AEO), replacing traditional SEO — AI-sourced traffic increased 527% from January to May 2025. Generative Engine Optimisation (GEO) — structuring content to get cited by AI systems like ChatGPT, Perplexity, and Claude.

As one Publicis executive put it: ‘AI agents are 10x faster, 100x smarter than junior staff.’ Additionally, agents don’t complain about reworks, overwork, or leaves. But here’s the nuance: Less than 10% of organisations have successfully scaled AI agents in any individual function. The technology is real. The differentiation challenge is also real.

The Disruptors: David vs. Goliath, Reimagined

Before we examine what the holding company giants are doing (in Part 2 of this series), let’s look at who’s eating their lunch. Because the most disruptive forces in this shift aren’t coming from within the industry — they’re coming from AI-native startups that have never known the billable hour.

Sierra — The Outcome-Based Pioneer

Sierra proves that outcomes-based pricing isn’t theory — it’s battle-tested at scale. With $635 million in strategic funding at a $10 billion valuation, Sierra has surged past $150+ million in ARR in just 21 months. Co-founded by Bret Taylor (former Salesforce co-CEO and OpenAI board chair) and Clay Bavor (an 18-year Google veteran), they charge for completed work, not subscriptions. Their clients include Deliveroo, Discord, Ramp, Rivian, SoFi, ADT, Cigna, and SiriusXM. What makes Sierra significant: Even ‘older businesses’ outside tech are betting their customer operations on AI agents.

Artisan — The SDR Replacement

The provocative ‘Stop Hiring Humans’ marketing campaign generated death threats — but also 250 paying customers. Artisan raised $46.1 million with just 88 employees, founded by 23-year-old Jaspar Carmichael-Jack (Y-Combinator backed). Their AI SDR ‘Ava’ handles prospecting, personalised outreach, and follow-up at scale, with access to 300 million+ B2B contacts. The economics are brutal for human SDRs: Artisan starts at ~$999/month, compared to $50,000–$150,000/year for a human SDR. The hallucination rate dropped from ‘extremely bad’ at launch to ‘maybe 1 in 10,000 emails’ through tight Anthropic integration.

Clay — The Data Intelligence Layer

Clay represents the ‘semantic layer’ that enables other AI agents to work effectively. With $210 million raised at a $3.1 billion valuation (doubled in 2025), their AI-driven lead enrichment turns a name and email into a complete prospect profile — company size, funding stage, tech stack, recent job changes, LinkedIn activity — in seconds. Research that would take 15 minutes per prospect is done in 15 seconds.

Jasper — The Enterprise AI Marketing Stack

Jasper’s evolution from ‘AI writing tool’ to ‘agent workspace’ mirrors the broader industry shift. With $131 million raised at a $1.5–1.8 billion valuation, they serve 70,000+ paying customers and 850+ enterprise clients with 100+ specialised AI agents. Their 2026 State of AI in Marketing report found that 91% of marketing teams now use AI, up from 63% the previous year. A cautionary note: Revenue is ~$88 million, down from a $120M peak in 2023 — a reminder that AI commoditisation is a real risk.

The Disruptor Pattern

What separates the disruptors from the incumbents isn’t just technology — it’s operating model:

On Pricing: Disruptors charge on outcomes or consumption. Incumbents charge billable hours and retainers.

On Data: Disruptors build proprietary enrichment integrated with LLMs. Incumbents rely on client data plus third-party tools.

On Talent: Disruptors run AI-native workforces of 50–150 people. Incumbents carry legacy talent pyramids of 50,000+ people.

On Speed: Disruptors deploy in days. Incumbents’ scope for months.

On Capital Efficiency: Artisan achieves $5M ARR with 88 employees. WPP generates $13.5B revenue with 98,000 employees. The unit economics are not even comparable.

The disruptors are proving that the unit economics of traditional agencies — the talent pyramid, the billable hour, the overhead structure — are artefacts of a world that’s rapidly coming to an end.

What’s Next

In Part 2: The Battleground, I will study how the four major holding companies — Publicis, Omnicom-IPG, WPP, and Dentsu — are responding to this collision. Each has placed a vastly different bet: data moats, scale consolidation, outcome-based transformation, or survival mode. The next 18 months will determine which strategy prevails.

The piece has been reproduced from Milind Pathak’s LinkedIn post.