The Numbers Most Independents Need to See
Let me start with some figures from Spark AI's 2026 report that sound intimidating.
WPP is spending £300 million per year on AI, with 27,000 monthly users on its Open platform and custom "Brains" fine-tuned on client data for brand strategy, audience insight, and media planning.
Publicis has invested €300 million, building its agentic AI platform in partnership with NVIDIA.
S4 Capital's Monks reports 50% reductions in delivery hours through its Monks.Flow agent ecosystem.
Those are not experiments. Those are infrastructure bets designed to lock in competitive advantage for the next decade.
If you run a 30-person independent agency, you look at those numbers and the instinct is to worry. How do you compete with infrastructure budgets of that scale?
You don't. And you don't need to.
Scale Cuts Both Ways
The numbers are real. So is the disadvantage they create.
A platform built to serve 27,000 monthly users across multiple networks, hundreds of clients, and dozens of markets has to be general-purpose. It has to handle every creative brief from a B2B SaaS launch to a luxury fragrance campaign to a global sports sponsorship activation. It has to accommodate the quality floor, the reviewer network, the compliance requirements, and the cultural adaptation of all of those markets simultaneously.
What that produces is capable, broad, useful infrastructure. What it does not produce is tight, specialist capability built around a specific kind of work.
That's the gap. That's the opening.
Tuncarp Proved It
The Spark report documents one example that should be required reading for every independent agency leader right now.
Tuncarp is a production agency. Small team. Independent. They restructured their delivery model around AI-enabled roles, rebuilt high-end video production to run at roughly a quarter of the traditional cost while maintaining 70–80% margins.
Then Publicis — the network spending €300 million on its own AI infrastructure — hired them as its dedicated AI content studio.
Think about what that means. Publicis has the internal platform, the agentic AI investment, the global scale. And they still hired a small independent as their specialist content studio. Because Tuncarp has something Publicis structurally cannot build internally: tight, focused AI-enabled operating capability for a specific kind of creative work, owned and evolved by a small team that can move in weeks.
That is the independent advantage, captured in a real commercial arrangement.
Weeks, Not Quarters
The Spark report puts the speed differential in plain terms: "A 30-person agency can move from idea to implementation in weeks rather than quarters. It can build proprietary AI capability tightly around its distinctive positioning in ways a global network structurally cannot."
Read that line twice.
Networks build for thousands of people. Decisions require alignment across markets, legal reviews across jurisdictions, compliance coordination across enterprise clients. A new AI workflow that's genuinely useful might take two quarters to become standard practice inside a global network. Sometimes longer.
A focused independent can prototype a new workflow on Monday, test it on live work by Wednesday, document it by Friday, and have the whole team using it the following week.
Over a year, the compounding effect of that speed difference is enormous. The network is ahead on scale. The independent is ahead on iteration count. Iteration count wins in a rapidly changing field.
The Strategic Opening
If you're running an independent creative team right now, the strategic move isn't to try to replicate what the networks are building. You can't afford to, and you'd be a bad version of them if you did.
The move is to be so tightly positioned, with such a clean AI-enabled operating model, that a network would rather hire you as a specialist supplier than try to build your capability internally.
That positioning depends on choosing what not to do. Spark's four agency AI archetypes make this concrete: Creative Differentiator, Operational Realist, Systems Architect, Performance Engine. Each represents a deliberate choice about where the agency competes and what AI investment reinforces.
The agencies losing right now are the ones trying to be all four at once. They're spreading AI investment across everything it can theoretically do, rather than concentrating it on what makes them genuinely distinctive. That's a trap that independents cannot afford and networks can.
The agencies winning — Tuncarp, Not Actual Size moving upstream into strategy, mark-making* building role-specific workflows across 12 people, the other independents the Spark report features — all did the same thing. They picked. Then they built hard around the pick.
What This Means This Quarter
Four practical moves for independent agency leadership:
Pick your archetype. Publicly. Not just internally — in your proposals, your case studies, your website. If clients are uncertain what makes you distinctive, they will reach for a network that has scale. Clarity on positioning is half the battle.
Concentrate your AI investment on that archetype. If you're a Creative Differentiator, AI should be freeing up time for deeper thinking and more creative territories, not trying to automate client services. If you're a Performance Engine, AI should be driving margin through output scale, not being applied to craft work where it doesn't move the needle.
Document your distinctive operating model. The advantage of being small is speed, but only if the speed produces systematised capability rather than lost workflows. Write down how you actually work. That document is what a network will eventually buy or partner with, when they need your specialism.
Move while they're still deciding. Networks are making big bets on infrastructure. Those bets will take quarters to land operationally inside client delivery. The window for independents to lead on specific-use-case capability is now.
The Counter-Intuitive Truth
The bigger the networks' bets get, the more valuable the independent opening becomes.
Every £100 million a holding company spends on general AI infrastructure is another £100 million that isn't being spent on deep specialist capability in any one niche. That's the gap a focused independent fills.
A 30-person team cannot out-invest WPP. It can out-focus WPP. In the market we're in right now, focus beats budget more often than it doesn't.
This is the independent moment. Use it.

