Short answer
For energy storage and lithium battery makers, GEO comes down to evidence AI can check. Publish UN 38.3, IEC 62619, UL 1973 and UL 9540 / 9540A on HTML pages with certificate numbers and the models covered, state the test conditions behind cycle life, and support it with structured data and third-party sources. "Long life, high safety" without a source does not get cited.
What overseas buyers ask AI
Storage procurement involves many people: developers, EPC contractors, system integrators, the owner's technical adviser and the insurer. The questions they put to AI almost always name a certification, an application or a region:
- "UL 9540A tested battery energy storage system manufacturers in China"
- "LFP battery cabinet suppliers for commercial and industrial storage in Germany"
- "Which BESS suppliers comply with the EU Battery Regulation?"
- "Compare 280Ah LFP cell manufacturers by cycle life and warranty"
To answer, AI retrieves the public material it can read and picks the sources that show certifications, parameters and project experience. The mechanism is explained in how ChatGPT and Gemini pick suppliers. A site that only says "industry-leading and safe" gives AI nothing to cite.
The evidence AI and buyers check
| Evidence | What buyers use it for | How to present it |
|---|---|---|
| UN 38.3 test report | Whether cells and modules can be shipped legally | List the models covered and the report number, with a summary |
| IEC 62619 | Safety requirements for industrial lithium batteries | Name the certification body, the certificate number and the models |
| UL 1973 / UL 9540 | North American safety certification for batteries and storage systems | Say whether it is battery-level or system-level, with a verifiable number |
| UL 9540A test | Thermal runaway propagation results, often required by fire officials and insurers | State the test level (cell, module, unit, installation) and the lab |
| NFPA 855 notes | US installation requirements for stationary storage | Explain how the design helps installers comply; do not write "certified to NFPA 855" |
| EU Battery Regulation (EU) 2023/1542 | Readiness for the EU market | Describe progress on the carbon footprint declaration, due diligence and battery passport, with the correct dates |
Two rules apply to every certification: say exactly which model and which level it covers, and give a number or body that can be checked, so buyers and AI can cross-reference it. UL certifications can be looked up in UL's public certification database, so an accurate number lets a third party verify it.
How to implement it
1. One page per product, with parameters in HTML tables
If the datasheet only exists as a PDF, AI crawlers and search engines often miss most of it. Give every cell, module, cabinet or container system its own page, with the key parameters in an HTML table: chemistry (LFP / NMC), rated capacity and energy, voltage range, C-rate, cycle life and its test conditions, operating temperature, ingress protection, cooling, communication protocols and certifications. Keep the PDF as a download.
2. Cycle life and efficiency need test conditions
"6,000 cycles" means little without the depth of discharge, C-rate, temperature and end-of-life criterion (70% or 80% retained capacity). Buyers cannot compare it and AI treats it as unverifiable. Add a "test conditions" column to the table, and say whether each figure is cell data or system data.
3. Structured data
Output Schema.org Product on each product page, put the key parameters in additionalProperty (PropertyValue with name, value and unit), point manufacturer at your Organization, and list certifications with hasCertification or in-page links. Without published prices you will not get Google's product rich results, but the structured parameters still help search engines and AI understand the product. See structured data and Google's Product structured data documentation.
4. Project pages
Storage buyers care most about what you have installed and how long it has run. Give each publishable project a page: location (country or region is enough), application (C&I, utility-scale, solar plus storage), power and capacity, commissioning date, grid code and operating record. Name the customer only with written permission; otherwise write "a commercial user in Europe".
5. Make it clear to AI who you are
Use the same English company name, address and product lines on your website, LinkedIn, trade-fair exhibitor pages (such as The smarter E Europe or RE+) and association member listings. If the names differ, AI may treat you as several companies. See entity.
6. Third-party sources
Project coverage in trade media (such as Energy-Storage.news or PV Magazine's storage section), public certificates from certification bodies and partners' case pages are the sources AI looks for when it cross-checks. Record the URL, date and buyer question for each one, so re-tests can be compared.
Pitfalls
- Do not present a component certification as a system certification. A cell listed to UL 1973 does not mean the cabinet is certified to UL 9540, and the gap shows up in due diligence.
- Keep safety claims factual. Avoid "never catches fire" or "zero thermal runaway". Say what was tested, what the result was and who issued the report.
- Get the EU Battery Regulation right. Regulation (EU) 2023/1542 applies in stages; the battery passport applies from 18 February 2027 to LMT batteries, industrial batteries above 2 kWh and EV batteries. Describe your readiness rather than claiming full compliance. The text is on EUR-Lex.
- Match the warranty to the datasheet. Warranty years and retained capacity must agree with the cycle life in the table. Contradictory numbers are questioned by buyers and AI alike.
- Cover transport and dangerous goods separately. UN 38.3, the SDS and the transport class are often treated as a precondition for delivery.
Launch checklist
- One page per product, parameters in HTML tables, key figures with test conditions
- Each certification with model, level, number and issuing body
- Product structured data with parameters in additionalProperty
- At least three publishable project pages
- Consistent company details on the website, LinkedIn and trade-fair listings
- robots.txt allows AI crawlers such as GPTBot, OAI-SearchBot and PerplexityBot
To see whether AI can read your site today, run a free audit. Our export GEO service follows this method, and the general product page rules are in the six-point product page checklist.
Frequently asked questions
- What should a storage company do first for GEO?
- Turn certifications and parameters into checkable HTML pages: model, level and number for every certification, and test conditions for figures such as cycle life. That is the precondition for AI to cite you.
- We only have cell-level certification. How should the site describe it?
- State the level accurately, for example "cells listed to UL 1973", and describe the plan for system-level certification. Do not imply the whole system is certified; it will be checked in due diligence.
- Does the EU Battery Regulation affect website content?
- Yes. European buyers ask about your readiness for the carbon footprint declaration, due diligence and the battery passport. Describe your progress and who owns it, without claiming full compliance.
- What if customers have not approved their names for project pages?
- Leave the name out and give the country or region, application, size and commissioning date. Real but anonymous project details are still useful evidence.
Further reading
- Service: Export GEO
- Service: Overseas website
- Service: Export SEO
- Glossary: What is an entity, and why must AI recognise your brand first?
- Glossary: What is structured data (Schema and JSON-LD), and what should an export site add?
- Glossary: What are AI crawlers, and should you allow GPTBot, ClaudeBot and PerplexityBot?
- Lesson: 1.2 What GEO is: how it differs from SEO and how they connect
- Lesson: 3.1 Overview of the six-step loop: diagnosis, modelling, content, sources, reputation and monitoring
See how AI describes you today
The free AI visibility audit asks ChatGPT, Perplexity and Gemini questions as an overseas buyer would, and checks your website’s technical foundation.