Bhullar Group
Taking an Australian steel fabricator from position 24.6 to 7.3 on Google.
- Client
- Bhullar Group
- Industry
- Structural steel fabrication
- Location
- Narrandera, New South Wales, Australia
- Year
- 2026 (new site launched 4 August 2026)
- Platform
- Next.js
- Services
- Apps and websites, Performance marketing
- Live site
- bhullargroup.com.au
- 7.3
- average Google position for "pile manufacturer", from 24.6 on the old site; impressions for the term rose from 68 to 207
- 615
- appearances in Google's AI features in the 28 days to 1 November 2026
- 8 of 10
- tracked buyer questions on which ChatGPT named Bhullar first or second
Sources: Google Search Console, 3 May to 2 August 2026 (old site) against 31 July to 30 October 2026 (new site); Search Console AI features report, 28 days to 1 November 2026; Marse Labs SEO report to Bhullar Group, September 2026.
About the client
Bhullar Group makes welded beams, bridge girders and pile casings in a workshop of more than 6,000 square meters in Narrandera, New South Wales, and supplies projects across Australia. The company has more than 25 years in steel fabrication, 10,000 tonnes of annual capacity and ISO 9001, 14001 and 45001 certification (source: Bhullar Group's website, November 2026).
The challenge
Bhullar's previous website showed eight product photos, a slogan and a phone number. Its product pages ran from 91 to 837 words and answered none of the questions buyers ask before they request a quote. The site had no enquiry form, no project pages and no structured data to tell search engines what the company makes, and its mobile layout blocked pinch zoom.
“Our old site was outdated. It had our phone number and a few photos, and buyers searching for welded beams or pile casings never found it.” , Bhullar Group
What we built
- One page per product: Six product pages, for welded beams, custom sections, bridge sections, pile casings, crane grillages and stopends, each with sizes, grades and standards written as text.
- Buyer questions answered on the page: 39 questions across the product pages, on sizes, grades, standards, quoting and lead times.
- Project pages with named jobs: Seven project pages, from the Sydney Metro station box at Barangaroo to the Newmont Tanami mine expansion, each naming the steel supplied.
- Open to AI search: A robots.txt file that allows GPTBot, ClaudeBot, PerplexityBot and Google-Extended by name, and an llms.txt file that summarizes the company's products, capacity and projects.
Decisions that mattered
- Facts as text, not images. We wrote sizes, tolerances and processes into the page text, where search engines and AI tools can read them. ChatGPT now quotes those details back to buyers: material traceability, automated SAW welding and custom lengths.
- Real projects, named. Each project page names the project and the steel supplied, so an AI answer about Sydney Metro steelwork has a source that mentions Bhullar.
- AI crawlers allowed by name. The robots.txt file names each AI crawler and allows it, and the llms.txt file gives AI tools Bhullar's own description of the business.
“We needed a modern website that buyers could find on Google. Marse Labs rebuilt ours around the products we make, and now we show up on Google and in ChatGPT when buyers search for them.”
Want a result like this?
Send a few lines about your project. We reply within 24 hours with questions or a time to talk.
- Reply within 24 hours
- Fixed written quote
- Your site and code are yours