AI Sell-Off Hits Agencies: What UK Startups Do Next

Investor sentiment can change faster than a media plan. This week, share prices for major agency holding groups—Omnicom, WPP, Publicis Groupe and Havas—fell sharply amid a wider AI-led market sell-off (with Dentsu notably bucking the trend). The financial story is interesting on its own, but it’s more useful as a signal: the market is re-pricing what “marketing capability” looks like in 2026.

For UK startups, this isn’t a spectator sport. When big agencies wobble, budgets get scrutinised, procurement gets tougher, and clients demand clearer proof that spend creates growth. At the same time, AI is compressing the cost of execution and raising expectations for speed. That combination creates an opening for agile teams to outmanoeuvre slower incumbents—especially in climate change & net zero transition markets where credibility, measurement, and regulatory realities matter.

This post translates the agency share-price slump into practical moves you can make now: how to position your startup, how to run AI in marketing without eroding trust, and how to use sustainability and emissions measurement as a competitive advantage.

What the agency share-price drop really signals

The key signal isn’t “agencies are doomed.” It’s that markets are uncertain about how value will be created and captured as AI automates more marketing work.

When investors sell off agency stocks in an “AI moment,” they’re usually reacting to a few fears:

1. Margin compression: If AI reduces the hours needed for creative, production, research, and reporting, traditional fee models get squeezed.
2. Client in-housing acceleration: Brands can buy tools and talent, then bring more work inside.
3. Differentiation blur: If everyone has access to similar AI capabilities, the question becomes: what’s defensible?

Here’s the startup angle: incumbents often need to protect legacy revenue streams; startups can build around the new reality.

Why this matters for climate and net zero businesses

Net-zero transition marketing has extra friction compared to “normal” growth marketing:

• Claims are scrutinised (greenwashing risk is real, and regulators are watching).
• Buying cycles are longer (especially B2B energy, mobility, built environment).
• Measurement matters more (you’re selling outcomes, not vibes).

That environment rewards teams that can combine proof, speed, and precision—exactly where thoughtful AI adoption can help.

Big agencies struggle with AI for one simple reason: incentives

Most companies get this wrong: they treat AI like a tool rollout. The hard part is changing the business logic around it.

Large agency groups are built on a model where revenue often correlates with:

• headcount
• utilisation
• billable hours
• layered processes (brief → strategy → creative → production → reporting)

AI flips that. It pushes value toward:

• faster experimentation
• better targeting and personalisation
• stronger measurement
• lower production cost

If your organisation is optimised for time spent, AI is a threat. If you’re optimised for outcomes achieved, AI is a tailwind.

The opportunity for UK startups: outcome-led marketing

If you’re a startup (or a lean scale-up team), you can position your marketing around outcomes without needing to defend an old pricing model.

Outcome-led marketing for net zero transition brands typically means one of these:

• Pipeline outcomes: qualified meetings, proposal volume, conversion rate
• Revenue outcomes: CAC payback, expansion, retention
• Trust outcomes: reduced compliance risk, improved claim substantiation
• Impact outcomes: measurable reductions in emissions intensity per customer (where relevant)

Investors questioning agencies is your cue to lean into clarity: what you do, how you measure it, and why it’s credible.

Practical AI adoption for startups (without the hype)

AI is a threat when it’s used to flood channels with low-quality content. AI is an opportunity when it’s used to tighten the loop between insight → execution → measurement.

1) Use AI to speed up research, not to replace judgement

Start with structured research that improves decisions:

• Cluster customer objections from call notes, demos, and support tickets.
• Summarise competitor positioning and identify claim gaps.
• Generate sector-specific messaging variants (e.g., for housing associations vs. facilities managers).

What works in practice: set a rule that AI can produce draft thinking, but humans must validate it with primary evidence—calls, emails, data, or user testing.

2) Build an “evidence vault” to protect against greenwashing

If you market anything tied to sustainability, you need a library of proof. I’ve found that teams move faster when evidence is organised like product assets.

Include:

• LCA summaries (where available)
• methodology notes (boundaries, assumptions)
• certifications and standards alignment
• before/after case studies with numbers
• approved claim language and “do not say” lists

Then use AI to retrieve and adapt approved evidence for different channels—website, sales decks, tender responses—without improvising claims.

A strong net zero marketing system is basically: fast creative + slow claims.

3) Treat creative as a testing system, not a one-off campaign

AI lowers production cost. Don’t spend the savings on more assets; spend it on better learning.

A simple testing cadence:

1. 2 new hypotheses per week (message, offer, audience, creative format)
2. 4–6 variants each (AI-assisted)
3. 1 decision every Friday (kill, keep, scale)

For climate tech and sustainability products, the best hypotheses often combine:

• a hard business outcome (cost, compliance, downtime)
• a sustainability outcome (carbon, waste, air quality)
• proof (case study, benchmark, method)

4) Put AI into measurement and reporting where it belongs

Startups win when they can explain performance quickly and credibly.

Use AI-assisted analytics to:

• identify which segments convert (not just which ads get clicks)
• detect “false positives” (cheap leads that never progress)
• summarise weekly learning in plain English for stakeholders

If you’re selling into enterprise or public sector, this is gold. Buying committees don’t want dashboards; they want clear, defensible narrative backed by data.

Investor behaviour is a market trend you can use

A stock sell-off sounds distant from your day-to-day, but it changes how decision-makers behave.

When markets get nervous about AI disruption:

• CFOs ask agencies for tighter scopes and more accountability.
• CMOs push for performance proof, not just brand lifts.
• Procurement leans harder into rate cards and vendor consolidation.

That’s tough if your pitch is “we make content.” It’s great if your pitch is “we drive measurable growth efficiently.”

How to position your startup marketing in 2026

Use positioning that matches the moment:

• Speed: “We test weekly and roll winners into paid + sales enablement.”
• Measurement: “Every channel has a cost-per-outcome target.”
• Credibility: “Every sustainability claim maps to evidence and approval.”
• Efficiency: “AI reduces production time; we reinvest in experimentation.”

If you’re in climate change and net zero transition sectors, add a line that many teams avoid but buyers respect:

• “We won’t publish claims we can’t substantiate.”

It sounds obvious. It’s also rare.

A better way to approach AI + sustainability marketing

The reality? It’s simpler than you think: treat AI as infrastructure for decision-making, not as a content factory.

Here’s a compact operating model that works for small teams.

The 30-day AI marketing sprint (built for lean UK teams)

Week 1: Foundation

• Define one North Star metric (pipeline, revenue, retention)
• Create an evidence vault for sustainability claims
• Audit tracking: CRM stages, attribution basics, conversion events

Week 2: Messaging and offers

• Interview 5–10 customers (or review 10 call recordings)
• Use AI to cluster objections and outcomes
• Draft 3 value propositions: cost, risk, and impact

Week 3: Test and learn

• Run small-budget experiments (search, paid social, partnerships)
• Ship 6–10 landing page variants (AI-assisted, human-approved)
• Set “kill criteria” upfront (e.g., CPL ceiling, lead quality rules)

Week 4: Scale what’s real

• Roll winners into sales enablement (deck, email sequences)
• Publish one substantiated case study
• Create a simple monthly reporting pack: outcomes + learnings + next bets

If you do this with discipline, you’ll look more “enterprise-ready” than teams with 10x your headcount.

People also ask: “Will AI replace agencies?”

Agencies won’t disappear, but the ones built around hours and volume will keep getting squeezed. The agencies (and in-house teams) that win will be the ones that combine:

• strong strategy
• fast experimentation
• credible measurement
• governance for brand and sustainability claims

For startups, the play is to adopt those behaviours now, before you have complexity.

What to do next (especially if you market a net zero solution)

Agency share-price drops are a reminder that the market is re-evaluating marketing value in the AI era. That’s uncomfortable for incumbents. For UK startups, it’s a prompt to build a tighter, more measurable growth engine—one that can stand up to scrutiny from investors, enterprise buyers, and regulators.

If you’re in the climate change & net zero transition space, you’ve got an extra advantage: you can outclass bigger competitors by pairing speed with proof. Make your claims defensible, your measurement clear, and your experimentation rhythm relentless.

So here’s the forward-looking question worth sitting with: if your marketing team had to justify every sustainability claim and every pound of spend to a sceptical CFO next month, what would you change this week?

Source backdrop: share-price declines among Omnicom, WPP, Publicis Groupe and Havas amid an AI sell-off (Campaign, published 4 Feb 2026).