Maksim Barodzich
Master of Economics, Co-Founder of Fluxy.One
The long arc of demand drivers
Industrial Revolution → early 20th century — Availability & price.
What sold: mass‑produced goods at affordable prices.
Why it changed: the moving assembly line (1913) drove unit costs and cycle time down, so access and price led demand.
1920s–1950s — Quality & standardization.
What sold: products with reliable, comparable quality.
Why it changed: expanding industrial standards and early information‑economics insights (“search” vs “experience” attributes) made quality differences observable before purchase.
1930s–1980s — Brand & mass advertising.
What sold: brands acting as trust proxies.
Why it changed: brand management (e.g., P&G’s 1931 “Brand Man”) and costly signals (ad spend, premium pricing) reduced uncertainty when quality wasn’t directly observable.
1980s–2000s — Service & experience.
What sold: superior service, convenience, and post‑purchase experience.
Why it changed: convergence of product features shifted competition toward service quality and customer experience.
2010s — Labels & registries.
What sold: products with visible, standardized performance (e.g., energy efficiency).
Why it changed: mandatory labels and public registries (like EPREL) made “credence” attributes salient and comparable.
2020s → Machine‑readable data (DPP, Schema.org, GS1 Digital Link/EPCIS).
What sells now: products with complete, verifiable, machine‑readable attributes that AI agents can filter and buy against user constraints.
Why it’s changing: structured data plus EU rails (ESPR; Battery Passport from 18 Feb 2027) shift the edge from brand stories to verifiable product truth.
Why brand premiums shrink when quality is observable
- Information economics. When buyers can’t observe quality before purchase, markets pay a reputation premium to sellers with credible signals (Shapiro). As information improves, that rent compresses.
- Signals. Price, advertising, or umbrella branding can act as costly signals of unobservable quality (Milgrom & Roberts; Wernerfelt). DPPs and verified labels introduce cheaper, truthful signals, eroding the payoff to noisy ones.
- From experience to credence to machine‑observable. Nelson split goods into search vs experience; Darby–Karni added credence (quality hard to evaluate even after use). DPPs shift parts of credence (origin, carbon, repairability) into ex‑ante, machine‑verifiable form.
AI agents are the new retail gatekeepers
Recommendation engines already change market concentration depending on design (can push either blockbusters or the long tail). Now extend that to shopping agents optimizing on constraints you set: price ceiling, delivery date, materials, certifications. Whoever provides clean, verifiable, and accessible product data will be ranked, filtered, and bought.
Evidence: what happens when labels make attributes visible
- A cointegrated VAR analysis of the EU energy label redesign found +55% sales for high‑efficiency cold appliances at announcement and another +42% at implementation in Denmark (2005–2017). When information gets salient, the market moves.
- Newer work documents consumer inattention and producer responses around mandatory eco‑labels (washing machines), confirming that clear, credible signals shift both supply and demand.
- EPREL—the EU’s public registry behind the energy label QR—shows how open, queryable product data underpins informed choice and enforcement. Expect the DPP ecosystem to follow a similar pattern.
Europe’s fast track to DPP
- ESPR (Regulation (EU) 2024/1781)—in force since 18 July 2024—creates the legal backbone for Digital Product Passports across many categories via delegated acts.
- 2025–2030 Working Plan prioritizes steel & aluminium, textiles (apparel), furniture, tyres, mattresses, plus horizontal measures on repairability—and explicitly routes product information via DPP or EPREL.
- Battery Passport (Regulation (EU) 2023/1542). From 18 Feb 2027, every EV battery, every industrial battery >2 kWh, and every LMT battery must have a battery passport, accessible via a QR code; the QR marking requirement itself also applies from that date. Access rights split public and restricted data. This is DPP’s first at‑scale real‑world deployment.
- Anti‑greenwashing Directive (EU) 2024/825. Member States must transpose by 27 Mar 2026 and apply from 27 Sep 2026. It bans generic environmental claims without proof and clamps down on unverifiable sustainability labels. DPP‑style evidence becomes not just smart strategy but legal hygiene.
A concrete use‑case: the picnic tablecloth
A consumer asks an AI agent: “Find a white picnic tablecloth, non‑absorbent, eco‑materials, ≤$25, delivered by Friday.”
How the agent buys in seconds:
- It reads structured web data (Schema.org Product/Offer/Review) and Merchant Center feeds for price/availability.
- It follows a GS1 Digital Link (from a QR or URI) to the product’s passport (composition, coatings, certifications, durability tests).
- It cross‑checks delivery SLAs and user context (calendar, color preferences) and places the order.
Who wins? The merchant whose machine‑readable fields are complete, correct, and verifiable—because agents privilege trusted, structured signals over brand aesthetics.
CEO playbook
Do in the next 90 days
- Close the structured‑data gap. Add/repair Schema.org Product + Offer + AggregateRating; validate in Google’s Rich Results Test.
- Stand up GS1 rails. Map GTINs to GS1 Digital Link URIs; choose or deploy a resolver; plan EPCIS 2.0 for event data (origin, custody, transformations).
- Greenwashing audit. Remove generic “eco‑friendly” language that lacks substantiation ahead of 2026 enforcement.
Build in 12 months
- Minimal viable DPP profile per priority category: identity, materials, durability/repair, certificates, date/version, public vs restricted fields, access rights. Align with ESPR/Working Plan.
- Data governance. Owner, SLAs for freshness, change logs; align public data with claims policies to avoid 2024/825 exposure.
- Go‑to‑agent strategy. Treat AI agents like a new acquisition channel; instrument logs to learn which attributes drive selection.
KPIs to watch
- Conversion uplift on pages with valid Product/Offer markup vs control.
- Return rate deltas after adding precise usage/care fields to DPP.
- Brand premium compression in categories where DPP adoption is high (track list vs national brands).
- Share of “agent‑originated” traffic and time‑to‑basket for agent sessions.
- Green claims incidents (alerts, takedowns, fines).
8) Risks, counter‑moves, governance
- “Brand is dead.” Not quite. Brand remains a trust layer—but increasingly as a trust in the data you publish (accuracy, timeliness, compliance) rather than a narrative cloak. Theory predicts shrinking rents when credible information is cheap.
- Data asymmetry & access. Battery Passport shows a tiered‑access model (public vs legitimate interest). Expect similar patterns in other DPPs; design your stack for selective disclosure.
- Algorithmic concentration. Recommenders can intensify blockbusters or amplify the long tail; mitigate by diversifying surfaces and exposing the right attributes (don’t rely on a single platform’s defaults).
9) FAQ
What exactly is a Digital Product Passport?
A machine‑readable identity and data bundle for a product (and sometimes an individual unit), typically resolvable via GS1 Digital Link (URI/QR) and populated with attributes like materials, origin, durability/repairability, certifications, and audit trails. Think “EPREL for many more categories,” but broader than energy.
When will DPPs be mandatory beyond batteries?
ESPR empowers the Commission to set product‑specific rules via delegated acts. The 2025–2030 Working Plan lists priority categories (steel/aluminium, textiles, furniture, tyres, mattresses), with information to flow via DPP/EPREL. Timelines will be set per act.
What’s the first large‑scale DPP in production?
The EU Battery Passport—mandatory for EV, industrial (>2 kWh) and LMT batteries from 18 Feb 2027, accessible via QR; QR marking requirement also from that date.
How do I make my products “agent‑readable” today?
Implement Schema.org Product/Offer and keep Google Merchant Center attributes complete; map GTINs to GS1 Digital Link; plan EPCIS 2.0 for supply‑chain events; keep public claims aligned with 2024/825.
Sources & further reading
Industrialization & demand: Britannica on the Industrial Revolution; Ford’s moving assembly line (1913).
Information economics: Nelson (1970); Shapiro (1983); Milgrom & Roberts (1986); Wernerfelt (1988).
Labels & empirical effects: Energy‑label redesign studies (Denmark CVAR; newer work on attention/supply responses).
EU policy: ESPR page (in force 18 Jul 2024); 2025–2030 Working Plan; Battery Regulation (passport & QR from 18 Feb 2027); Anti‑greenwashing Directive (transpose by 27 Mar 2026, apply from 27 Sep 2026).
DPP plumbing: GS1 Digital Link & EPCIS; Schema.org Product; Google structured‑data & Merchant Center guidance.
Recommenders: Fleder & Hosanagar (2009) on how algorithms shape the sales distribution.
One last executive nudge
If an AI agent can’t see your product’s facts, it can’t select it. Treat DPP + structured data as the new shelf space. The sooner you make your products machine‑legible, the sooner you’ll stop paying a tax to story‑telling that can’t be verified.
