greensurance-braucht-struktur-warum-nachhaltige-versicherung-bei-normierten-datenfl-ssen-beginnt

{ "translated_markdown": "10.11.2026 Sasha Justmann\n# Greensurance Needs Structure: Why Sustainable Insurance Begins with Standardised Data Flows\nData flows between customer, broker and insurer as a lever for sustainability in BiPRO standardisation.\n\n## Sustainability is New on the BiPRO Agenda\n\nFor twenty years, sustainability played no significant role in BiPRO standardisation. The standards governed how policy, document and claims data flowed between insurers and distributors – not whether a product possessed ecological or social characteristics. That is currently changing: within BiPRO e.V., a working group is now addressing the question of how sustainability and greensurance can be represented in the world of standards.\n\nTwo main strands are emerging. The first concerns green IT: what contribution does standardised data transmission itself make – for instance because normalised, deterministic data streams render compute-intensive AI interpretation of unstructured documents redundant? The second concerns green insurance: how can product sustainability characteristics, customer preferences and processes be standardised so that they reach all connected distribution partners?\n\nIn my article Green Interfaces, I discussed the core proposition that sustainability in insurance IT begins with data structure. What remained open was which data these are specifically. That is what this article is about – alongside an observation that crystallized early in the discussion: for traditional sustainability reporting, standardised attributes are surprisingly unhelpful. So what actually needs to be standardised?\n\n## Taking Stock: Plenty of Regulation, Little Shared Data\n\nRegulators, rating agencies and customers are increasingly asking for sustainability – yet outside of capital investments, a shared, machine-readable data foundation barely exists.\n\nRegulation is intensifying, but remains fragmented: revised ESRS since 10 November 2026¹, a planned “Adaptation Underwriting KPI” for non-life business², the natural hazards extension for residential building insurance³, the UWG ban on unsubstantiated environmental claims since 27 September 2026⁴, and three new SFDR product categories⁵.\n\nThe classes of business are unequally prepared. For funds and insurance-based investment products, the European ESG Template (EET) provides an established exchange standard⁶. In property and casualty insurance, standardised sustainable tariff features are still rare⁷; Franke und Bornberg includes the topic on the watchlist in its contents insurance rating because reliable criteria are lacking⁸. According to EIOPA, only 16 out of 46 manufacturers had classified sustainable non-life products at all⁹.\n\nIt is a similar picture for customers and distribution: high intent, low implementation. 80 per cent of customers want to consider at least one ESG criterion during their next contract conclusion¹⁰, yet only a quarter know what sustainability actually means in insurance¹¹. In advice, the topic is missing in four out of five consultations¹².\n\n## Where the Discussion is Headed: Data Crossing Corporate Boundaries\n\nAn insurer's own sustainability report is an internal aggregation. It brings together premiums, investments and emissions from its own systems – an additional attribute in the transmission to the broker changes little about that. The obvious conclusion: standardisation creates value where sustainability data crosses corporate boundaries – between customer, intermediary, insurer and service provider.\n\nThree such data flows stand out in particular:\n\n- Preference matching in advice. The preference originates with the broker, the product feature with the insurer. Matching only works if both sides are standardised in the same format.\n- Prevention and geodata from application to tariff. For the taxonomy under the explicit approach, only premiums whose terms contain climate-related criteria with a pricing impact count¹³. Examples include the geocoordinates of the building for flood and heavy rainfall zones, installed protective measures such as backflow valves or mobile flood protection, roof condition and construction type against storm and hail, or energy-efficiency retrofits. All of this is established at the point of sale – which usually means at the broker.\n- Lifecycle events in the in-force portfolio. Sustainability is not a static state at contract conclusion. Energy retrofits, adding natural hazard cover or reclassifying a fund alter the status of a policy – and the broker needs to be informed.\n\nEven reporting benefits indirectly from these flows. The PCAF standard calculates commercial insurance emissions via the ratio of premium to customer revenue – yet this revenue is not always collected during application submission¹⁴. The data gap in reporting is often the result of a gap in the application – or at the very least when updating in-force portfolio information.\n\n## Lessons from Investments and Life Insurance\n\nThe investment sector created obligations first and supplied the data later. Since August 2022, intermediaries dealing with insurance-based investment products have had to ask for sustainability preferences – with a sobering outcome:\n\n- Obligation before product classification: In an EY test, 95 per cent of intermediaries omitted the query¹⁵. Without appropriately classified products in their tools, intermediaries lack the foundation for a systemic recommendation.\n- Overwhelmed by complexity: Taxonomy, Sustainable Finance Disclosure Regulation and PAI in a single query – customer willingness to engage dropped from 53 to 22 per cent according to the AfW¹⁶; the association considers the query to have failed¹⁷.\n- Correction via categories: With SFDR 2.0, preferences under MiFID II and IDD are likely to be aligned with the new product categories – presumably not before 2028¹⁸.\n\nIn non-life insurance, there is no obligation to query preferences – in 30 studied consultations, no one asked about sustainability preferences¹². No supply without demand, no demand without supply.\n\nThe lesson: Non-life insurance can do it the other way around – first establish a shared data basis, then advice and comparison, and regulation that builds on an existing standard.\n\n## How Sustainability Could be Standardised in the BiPRO Data Model\n\nThe following ideas are conceptual models, not decided standards. They follow a core principle: anyone creating a boolean field “sustainable” in their product or pricing model has already lost the domain complexity. Sustainability is not a yes/no value, but a bundle of multiple characteristics.\n\n### An ESG Data Object Instead of Free Text\n\nToday, sustainability characteristics in pricing, quote and application processes are mostly represented via free text, special agreements or proprietary data extensions. This does not scale. A future standard could define a dedicated ESG data object referenced via standardised lists. It would be structured across three dimensions:\n\n- Circular economy: repair before replacement, reconditioned parts, drying instead of replacement, ecological reinstatement values.\n- Climate: physical risks and adaptation measures, i.e. geodata, risk classes and prevention measures.\n- Social: accessibility of the application flow, understandable terms, affordability in exposed locations, non-discriminatory attribute combinations.\n\n### Two Levels: Category and Attribute\n\nInvestments show that advisory processes need simple classification, whereas pricing and proof require details. The data object should therefore separate two levels: a coarse category for advice and comparison, and detailed attributes underneath. In life insurance, the category can align with SFDR 2.0 categories. For non-life insurance, there is no concrete specification yet – an open list could follow the framework of the Versicherungsforen market monitor, which categorises non-life concepts into climate change adaptation, ecological added value, social added value and sustainable lifestyle¹⁹.\n\n### Line-Specific Attributes in the TAA Standards\n\nThe TAA standards 421 to 426 are ideal for line-specific implementation:\n\n- Property & Casualty (Standard 421 Personal Lines Composite, Standard 426 Commercial): energy efficiency class, construction standard, PV system, heat pump and storage as insured property, prevention measures and coverage modules with limits. Such modules already exist, but inconsistently – such as additional costs for ecological restoration and energy consulting²⁰ or a dedicated sustainability module in residential building insurance²¹.\n- Motor (Standard 423): drive type, battery health, charging infrastructure, telematics and claims settlement using reconditioned parts.\n- Life and Health (Standards 422 and 424): taxonomy-compliant investment criteria and preventive health factors, structured as cover or endorsement. For the investment side, BiPRO should build a bridge to the EET, which already comprises 600+ data fields²².\n\n### References\n\nMuch sustainability data originates outside insurance and is maintained there. The data model should be able to reuse these sources.\n\n- Buildings: Article 22 of the EU Energy Performance of Buildings Directive obliges every member state to establish a national database for the overall energy performance of buildings²³; the transposition deadline ended on 29 May 2026. In Germany, it is still being built²⁴. A reference to building ID or energy performance certificate will save re-entry in future.\n- Batteries: From 18 February 2027, the digital battery passport becomes mandatory for EV batteries, e-bikes and industrial batteries over 2 kWh²⁵. The passport ID can link motor, contents and building policies with verified data.\n- Policy Terms: GDV clauses already exist for many modules, such as PK 7367 on extra costs for energetic modernisations not required by law²⁶. They can be referenced directly as codelist entries.\n- Claims Remediation: GDV and VdS have formulated sustainability goals for property damage remediation as a non-binding industry standard with VdS 6038²⁷. A standard can also reference this instead of inventing its own criteria.\n\n### Versioned and Verifiable\n\nTwo properties make the data model future-proof.\n\nFirst, every category carries a reference to the regulatory framework under which it was assigned – SFDR in its old or new version, future requirements for non-life. Regulation changes; the data model must not break every time.\n\nSecond, every attribute carries a proof reference. Since the new UWG, blanket environmental claims without proof are banned. A tariff that calls itself “sustainable” therefore requires verifiable individual attributes – precisely what a typed data model provides.\n\n## What Needs to Change Between Broker and Insurer\n\nProcess follows the data model, not vice versa. Once the ESG data object is established, steps along the entire chain change:\n\n1. Advice. The customer's preference becomes its own structured object: category, topics, weighting, timestamp. The default value is “neutral” – as EIOPA already foresees for insurance-based investment products when a customer expresses no sustainability preference²⁹. If no product matches the expressed preference, the discrepancy is documented with reasons.\n\n2. Quote and Application (TAA). The outcome of the matching is transmitted with the application. This embeds the advice record directly into the application flow – a point EY identified as a prerequisite for distribution implementation shortly after the IDD query was introduced³⁰.\n\n3. In-force Portfolio (Standard 430.4). The indicators fixed in the application must flow consistently back into broker management programs and distribution systems – not just once, but with every event that changes the status: an energy retrofit, switching to natural hazard cover, reclassifying a fund under SFDR 2.0. Only then can the broker see whether a contract still suits their customer's preference and react accordingly in sales.\n\n4. Claims. The mode of settlement becomes a verifiable attribute: repair, reconditioned part, drying, “build back better”. The industry is further ahead here than with products – according to VdS 6038, 81 per cent of the market considers sustainability criteria in their processes, 71 per cent rely on repair over new purchase²⁷. What is missing is the structured feedback of this information into the policy.\n\n5. Reporting as a By-product. If prevention attributes, climate coverages and settlement modes are available in a structured form, the taxonomy share of premiums and the data for PCAF calculations can be derived directly from the in-force portfolio, rather than compiled manually every year.\n\n## Outlook: Is the Green Cycle Coming?\n\nA cycle initiated by end customers out of conviction is rather unlikely in property and casualty insurance.\n\nThe intent is there, the implementation is not. According to Q-Perior, 58 per cent of insurance customers would pay more for sustainable products, parents under 40 an average of 15 per cent more³¹. At the same time, according to a study by FH Dortmund, 40 per cent put sustainability on hold as long as prices rise³². Market researchers estimate that only 30 to 50 per cent of expressed intentions actually lead to buying behavior³³.\n\nOther markets show what works. Organic food sales reached a record 18.23 billion euros in 2025, but the market share stands at 6.5 per cent³⁴. Growth drivers were the retail sector's organic own-brands³⁵ – sustainability on the normal shelf at the normal price. Bain also observes that people act sustainably without labelling it as such: 48 per cent of AI users have already used AI to live more sustainably, 41 per cent use product information apps³⁶.\n\nA vision for property and casualty insurance. Sustainability becomes a by-product of good data. Premium automatically rewards behaviour – based on verified data from energy performance certificates, battery passports, prevention measures and sensors instead of an ideology query. The customer's AI agent queries tariff attributes digitally and filters according to their preference; whoever fails to provide structured attributes remains invisible to it. The broker becomes a resilience coach. And claims handling closes the loop: repair or sustainable restoration is managed, and proof flows back into the policy.\n\nSo a green cycle can emerge: structured data makes sustainability visible in comparisons, at equal prices it drives selection, volume encourages insurers to offer further modules, raters turn it into a rating criterion – and regulation builds upon an existing standard. What remains at the end is something like a product passport for insurance tariffs, rendered transparent via BiPRO standards and processes.\n\n## Conclusion\n\nThe investment sector created obligations first and supplied the data later – with familiar consequences in advisory practice. Non-life insurance can turn this around and get ahead of the curve using BiPRO standards.\n\nThe real lever does not lie in the sustainability report, but in the data flows between customer, broker, insurer and service provider: preferences, prevention and geodata, lifecycle events and settlement modes.\n\nA typed, versioned and verifiable data model makes sustainability audit-proof – for brokers, comparison engines, AI agents and regulators. That BiPRO e.V. is now addressing this issue is an important signal. Those who standardise the data first will set the tempo before obligations arrive." }