The end of the web deal? The response economy is redrawing value!

AI response, zero clicks and a new valuation risk for companies

For twenty years, the web economy has been based on an implicit pact: search engines index content and return qualified traffic in exchange, monetizable in advertising, subscriptions, leads or sales. The rise of AI "answer engines" (Google AI Overviews, ChatGPT, Perplexity, Claude...) is shattering this balance: they suck up billions of pages, synthesize answers... and return fewer and fewer visitors. Matthew Prince (CEO of Cloudflare) speaks of a collapse in the "crawl-to-refer ratio" (pages sucked in / visits returned) and a "cracking web business model".

On July 1, 2025, Cloudflare launched "Content Independence Day": default blocking of unpaid AI crawlers and a new Pay-Per-Crawl model to charge for machine access to content, with onboarding that asks new customers if they authorize AI bots and under what conditions. It's a clear pivot: re-establishing an economic counterpart to the extraction of value by AI.

For managers, investors and appraisers, these changes affect revenue, acquisition costs, the value of intangible assets and therefore the valuation of their company. This paper offers a structured reading of the facts, figures, risks... and the levers that protect value.

1) What the facts say (key points and recent figures)

Before analyzing the financial and strategic impacts, it is essential to return to the observable facts. Recent data confirms a profound upheaval: the rise of the zero-click, the explosion of automated traffic, the widening gap between sucked-in pages and returned visits, not to mention the first complaints and disputes. These signals are converging to show that the historic pact between content creators and platforms is breaking down.

1.1. Zero click" explodes

Converging studies show an increase in no-click searches (the user obtains his answer on the engine page). Similarweb measures an increase to ~69% of "zero-click" after the deployment of Google AI Overviews, with a marked decline in organic traffic from news sites.

1.2. AI Overviews: complaints, investigations, litigation

Publishers (alliances of independents, pro-open web NGOs) have filed a complaint in Brussels against Google's AI Overviews, requesting interim measures under competition law. At the same time, the Commission has notified Alphabet of preliminary findings of DMA non-compliance (including self-preferencing in search), placing AI Response Blocks under heightened scrutiny. DMA penalties can amount to up to 10% of worldwide sales.

1.3. AIs refer... but return little (the "crawl-to-refer gap")

Cloudflare documents a growing gap between content crawling by AIs and visits returned. A controversy targeted Anthropic: rankings based on "crawl-to-refer" put it among the players scraping a lot and referring little. The company partially denied this, explaining that many referrals go through integrations (Slack, Notion, Poe...), hardly visible in traditional metrics. Beyond the case in point, this controversy highlights the difficulty of correctly attributing referrals and the risk of value "creeping up" to AI interface layers without tracing the source.

1.4 Booming bot traffic (and associated costs)

According to theImperva Bad Bot Report 2025, advanced bots massively target APIs: 44% of advanced bot activity targets API endpoints. The travel sector is one of the most targeted: up to 48% of traffic on travel sites is made up of bad bots. Consequences: rising infrastructure costs, analytical bias, fraud and pressure on SEO performance.

1.5. " The end of the great bargain" (Cloudflare reading)

In interviews (Wired, Business Insider) and official posts, Matthew Prince defends the "Spotifyization of the web": licensing content and charging for machine access (Pay-Per-Crawl), otherwise content production will contract and the ecosystem will be impoverished.

2) Three economic shocks for companies

The impact of response AI on businesses is not limited to marketing or SEO: it translates into real economic shocks. These upheavals affect revenues, operating costs and the value of intangible assets. Three major dynamics are taking shape: the loss of monetizable organic traffic, the rise in costs linked to bots and AI reintermediation, and finally the gradual depreciation of assets once considered strategic.

Shock #1 - Loss of monetizable organic traffic

• Media/publishers: the rise of zero clicks and AI blocks shifts attention to the results page. Similarweb/Digiday have documented significant organic declines and shares of queries where >70% of impressions result in zero clicks to source sites. Effects: CPM and affiliation under pressure, acquisition costs up.
• E-commerce & B2B:intent is intercepted by "answer layers" (AI sheets, summaries, carousels), which compress the conversion rate of landing pages and increase the cost of the acquisition mix (SEA, partnerships, retargeting).

Shock #2 - Hidden rise in digital costs

• Bots & scraping: + hosting costs (non-human peaks), false analytical signals, security/API hardening. For highly exposed sectors, we typically observe +0.5 to 1.5 points of sales in security/API OPEX (anti-bot sensors, WAF, observability, load tests) - the example of travel, with 48% of traffic in bad bots, illustrates the scale of the resources to be deployed.
• AI reintermediation: if AI captures the user upstream, margins shift to the "answer layers" (AI engines, AI browsers, aggregators), compressing the sites' value capture. This is what happened yesterday with price comparison sites / OTAs: the interface captures the premium. Tomorrow, Perplexity or AI Overviews will play this role on a web scale.

Shock #3 - Impairment of intangible assets

• SEO & content assets: the marginal utility of an article/FAQ base decreases if AI extracts the essential content without a click. A brand that used to be worth X thanks to its stable organics may see its "informational goodwill" eroded; need to update attrition rate and LT growth assumptions.
• Platform dependency: concentration of distribution power (Google, OpenAI, Anthropic, Perplexity) increases platform risk; in Europe, the DMA allows unilateral rule changes and heavy penalties. Valuation consequences: discount for SEO over-indexed profiles (historical analogy: discount for apps 100% dependent on a store).

Make an appointment with a XVAL consultant on this subject:

3) Rights, deals and litigation: towards web licensing

Faced with these abuses, the central issue becomes that of regulation and remuneration. In both Europe and the United States, regulators and publishers are seeking to redress the balance of power: DMA complaints, neighboring rights, licensing agreements and legal action. Behind these initiatives lies the future of a model where access to content can no longer be free and unlimited for AIs.

• EU - Neighboring rights & DMA: complaints against AI Overviews highlight substitution effects and the possible self-preference of a gatekeeper, at the heart of the DMA. Publishers push for precautionary measures to avoid irreparable damage during investigation.
• Licensing market: OpenAI has signed with Axel Springer (2023), Financial Times (2024), News Corp (2024 ) and AP (2023). The US press reported amounts >$250 million over 5 years for News Corp (cash + credits), foreshadowing recurring income for holders of archives and premium content.
• New Cloudflare doctrine: default blocking, Pay-Per-Crawl, Managed robots.txt, Verified Bots and Message Signatures: the aim is to charge for machine access and certify legitimate crawlers.

4) How these shocks are transmitted to valuation

The shift to the "answer economy" doesn't just mean fewer clicks: it also reshapes the mechanics of cash flow creation. Less organic, more paid and more security/API weigh on EBIT, while dependence on platforms increases the risk premium. In concrete terms, this affects your DCF valuation (growth, margins, CAPEX), your market multiples (SEO exposure discount) and the value of your intangible assets (content, data). The following section links these shocks to valuation models - with potential orders of magnitude.

4.1. DCF (discounted cash flows)

• Topline: fewer organic sessions, lower CTR → -visits, -conversion, -panier; heavier paid mix → gross margin under pressure. Integrate a "high zero-click (≈60-70%)" scenario on core queries.
• OPEX: increased security/anti-bot/API (API targets, fraud, application DDoS), content structuring costs (schema.org), marking/content signatures, monitoring crawler/licensing. Empirically, forecast +0.5 to 1.5 pt sales, depending on sector exposure.
• CAPEX: 0.5-1.0% ofinitial data/IA sales (schematization, signatures, observability, Pay-Per-Crawl tools). This envelope is consistent with the investments advocated by infrastructure players (Cloudflare) to regain control of machine access.

4.2 Market multiples (comps)

Over-indexed SEO models should trade at a higher risk premium (EV/EBITDA spread) vs. diversified peers (subscriptions, proprietary data, direct channels). Investors are now scrutinizing dependence on 1-2 platforms and the ability to monetize AI access (licensing).

4.3. Intangible assets: impairment testing & re-scoring

• CGU "SEO content": revise the recoverable value with a lower SERP CTR and an implicit IA substitution rate.
• Media brand" goodwill: review g and WACC ("platform risk" premium) in the light of DMA / litigation / click share trends.

5) Three strategic levers to manage the Answer Economy

Beyond this observation, the key for managers is to know how to manage this new situation. The "answer economy" requires them to reassess their exposure, measure their ability to monetize content differently, and anticipate transition costs.

Three strategic levers can be used to guide this reflection and integrate these issues directly into the management and development of the company.

1. AI risk exposure
   What percentage of traffic / leads / sales still depends on traditional SEO? Example: an e-merchant with 65% SEO acquisition is highly exposed to interface flips (AI Overviews), so higher risk premium in valuation. Market benchmark: zero clicks ≈ 69% post-Overviews.
2. Indirect monetization capacity
   Do you have assets that can be monetized through licensing (archives, proprietary databases, specialized content)? The wave of OpenAI deals (AP, Axel Springer, FT, News Corp) proves that recurring licensing income can stabilize margins and support terminal value.
3. Operational and financial resilience
   Have you budgeted for 0.5-1.0% of sales in data/IA CAPEX (schematization, signatures, monitoring) and 0.5-1.5 pt of sales insecurity/API OPEX (anti-bot, WAF, observability, anti-fraud)? Companies that anticipate these costs inspire greater investor confidence (controlled exposure, crawler governance, ability to price access).

6) Defensive & offensive strategies (and their effects on value)

Adapting to the "answer economy" isn't just about recognizing losses: it's about building a proactive strategy. Some measures are aimed at protecting assets and limiting value erosion, others at capturing new sources of revenue through AI. Whether defensive or offensive, these approaches redefine how companies can preserve - and even strengthen - their value in this new ecosystem.

Defensive

• Crawler governance: managed robots.txt, Verified Bots, Message Signatures, default blocking of unauthorized bots, Pay-Per-Crawl. Objective: reduce dark crawling and charge for access.
• Hardening APIs: rate-limits, token binding, anomaly detection; advanced bots now target APIs first and foremost (44%).
• SERP-/Answer-first design: structured data (schema.org), extracts readable by AI but marked/licensed (content signatures) to negotiate remuneration rather than "free aspiration".

Offensive

• Licensing: contract with OpenAI/Anthropic/Perplexity/Google or via intermediaries; package your assets (metadata, rights, traceability) to sell access.
• Proprietary AI products: RAG on first-party data, customer co-pilots, "business knowledge" APIs; shift part of the value to direct experience (apps, community).
• Pivot "data": enrich your content with exclusive measurements (tests, benchmarks), which are harder to paraphrase and more monetizable.

Valuation effect: these levers lower the risk premium, stabilize margins (licensing + direct channels) and support terminal value by reducing dependence on a single interface.

7) What Matthew Prince (really) says - and why it's structuring

Prince puts words to a systemic transformation:

1. AIs strip-mining the web (aggressive extraction, few referrals).
2. The historic deal collapses (Google moves from a "links" to a "responses" paradigm, the crawl/visitor gap widens).
3. We need prices & licenses (Pay-Per-Crawl), otherwise content supply will shrink. His point is based on Cloudflare's network position (infrastructure, security, ~"a big chunk" of the web), which sees real flows and increasing botification.

8) XVAL recommendations for managers and investors

At the end of this diagnosis, a crucial question remains: how can these findings be translated into decisions for managers and investors? The upheavals linked to the AI response must be integrated into financial models, due diligence and strategic company valuations. The recommendations that follow offer concrete guidelines for adapting your valuations and securing value creation in this changing context.

• Integrate an "Answer Risk Add-On" into your models: simulate -10% / -20% / -35% organic growth over 12-24 months depending on exposure, with CPA +15-30% and OPEX security/API +0.5 to 1.5 pt of sales; recalibrate terminal value (platform-dependency discount).
• Enhance the value of licensing flows: build an "IA licensing" P&L (price per 1,000 crawls/requests, guaranteed floors, reporting); VT sensitivity to the rise in these recurring revenues.
• Re-score intangibles: premium on unique content and proprietary datasets (difficult to paraphrase), discount on assets whose use is easily substituted by AI (generic FAQs, commoditized content).
• Reduce dependency: develop direct channels (apps, newsletters, community), in-house AI products, reverse affiliate programs (your AI sends... home), and DMA governance (compliance monitoring & scenarios).

Conclusion - Is the response economy redrawing value towards ?

The web "of the visit" is sliding towards a web "of the response". In this new order, traffic is no longer the finished product: it's a by-product among others (licensing, APIs, embedded AI). Companies that measure their exposure, monetize machine access (Pay-Per-Crawl, deals) and reinternalize customer relations (AI products, direct channels) preserve their margins... and their valuation.

XVAL can help you quantify these risks/opportunities, re-score your intangible assets and integrate credible scenarios into your DCF/comps - in order to shift the value curve from clicks... to licensing or new models to be defined...

Book an appointment with a XVAL consultant :

Main sources

• Cloudflare / Matthew Prince: post " Content Independence Day: no AI crawl without compensation! ", Pay-Per-Crawl ads, interviews (Wired, Business Insider). Business Insider+4The Cloudflare Blog+4The Verge+4
• Zero-click & traffic impact: Similarweb (up to ~69%), sector analyses. Search Engine Roundtable+1
• Bots & APIs: Imperva Bad Bot Report (44% of advanced bot activity targeting APIs; 48% of bad bots on travel). cpl.thalesgroup.com+1
• DMA / Antitrust (EU): complaint against AI Overviews; preliminary findings of Alphabet's non-compliance with the DMA (including self-preferencing). Reuters+1
• Anthropic & "crawl-to-refer" controversy: review articles and partial denials (integrations not visible). Business Insider+1
• AI licensing: OpenAI deals with AP, Axel Springer, Financial Times, News Corp (order of magnitude >$250 M / 5 years reported by the press).