How rebuilding Welding & Welder's ecommerce foundations grew organic revenue by 97%

Most case studies are written from the outside. Welding & Welder is different.

I own a share in the business. At the end of 2022, four of us bought Welding & Welder from one of the directors’ fathers, who had spent years as a welder, run a bricks-and-mortar welding shop, and later moved the business online as he moved towards retirement.

This case study shows what ecommerce SEO looks like when technical fixes, catalogue quality, product data and commercial operations all have to work together.

The business had real trade roots, a specialist product range, supplier relationships and customers who already trusted it. The missing piece was a mature ecommerce setup. The site had grown around the business rather than being designed around how customers search, compare and buy welding supplies online.

We knew very little about welding when we bought it. What we did understand was ecommerce, SEO, development, marketing and the opportunity to improve a strong but under-optimised business by fixing the foundations.

Buying into a real trade business with weak digital foundations

The starting point was commercially interesting but operationally messy. The business had history and demand, but the site made too much of the buying journey harder than it needed to be.

Templates were dated. Mobile usability was weak. Navigation was cluttered, and important categories could sit several levels deep. Customers who knew exactly what they wanted could sometimes find it, but anyone comparing products, browsing by use case or narrowing a technical choice had to work too hard.

The catalogue was the bigger problem. Technical products often vary by size, voltage, output, fitting type, capacity, process, material or compatibility. Product variations and structured attributes were handled poorly, so many near-identical items existed as separate product listings. That made the catalogue bigger, harder to manage and harder to shop.

The same issue affected search and filtering. Without consistent attributes, customers could not easily narrow a category by the specification that mattered to them. Product listing pages were less useful than they should have been, and site search had limited structured data to work with.

Product data quality varied as well. Some pages had missing images, thin descriptions, weak technical specifications or inconsistent information. There was little useful supporting content beyond the product catalogue, and many categories lacked the buying guidance a technical customer would expect.

The problem was wider than SEO. The site needed better ecommerce foundations before organic traffic could become more commercially useful.

Fixing templates, mobile journeys and navigation without a full-time team

This was a bootstrapped side project rather than a large transformation programme with a full delivery team behind it. None of the owners could work on the business full time, and hiring dedicated merchandising resource was difficult to justify at that stage.

That constraint shaped the work. We had to focus on changes that improved customer experience, organic visibility and commercial performance at the same time.

The first priority was the template and journey work. Product pages and listing pages needed to be easier to use. The mobile experience needed to feel less like an old desktop site squeezed onto a smaller screen. Browsing paths had to become simpler, especially for customers without the exact SKU in mind.

Navigation and category structure came next. We reduced unnecessary depth, brought important categories closer to the surface and made the structure more intuitive. The aim was simple: help customers find the right part of the catalogue faster, while giving search engines a clearer view of the site’s commercial structure.

The template, mobile and navigation changes moved the site forward. Organic visibility improved because the site became easier to understand, easier to crawl and easier to use.

The early uplift came from making the ecommerce experience work better. Hundreds of articles were never the first move. The priority was fixing the paths customers actually used.

When catalogue quality became the limiting factor

The first round of improvements helped, but deeper quality issues remained inside the catalogue.

As Google continued to reward stronger ecommerce experiences and better product information, those weaknesses became harder to ignore. The site started losing ground across several core updates. The core update decline was uncomfortable, but useful. It pointed straight at the next layer of work.

The issue went beyond content. It was ecommerce quality at scale.

Thin product descriptions, inconsistent specifications, weak product imagery, missing attributes and poor variation handling all affected how useful the site was. They also affected how confidently search engines could understand the catalogue.

Manual clean-up would have taken a huge amount of time. Every product family needed different judgement. Welders, welding wire, regulators, clamps, PPE and consumables all have different buying criteria. A generic content clean-up would not have solved that.

We needed a way to improve the catalogue steadily without pretending we had a full merchandising department.

Using AI to improve product data without removing human judgement

AI became useful because the bottleneck was practical. The business needed better product information, better category content, better attributes and cleaner variation handling, but the work was too large to handle manually at pace.

We used AI to support product description improvements, category copy, product image work, attribute creation and variation grouping. AI stayed in a support role. It helped turn a slow manual problem into a process the ownership team could work through with review.

For technical products, that mattered. A welding customer might need to compare duty cycle, current range, voltage, process, wire diameter, material, pack size or compatibility. If those details are missing or inconsistent, the page is less useful. If they are captured as attributes, they can support filters, comparison, internal search and cleaner product management.

Variation work helped reduce bloat. Instead of treating every near-identical product as a separate listing, the catalogue could start moving towards grouped product options where that made sense. That improved the shopping experience and made the site easier to manage.

The product data work had to reflect how people actually choose welding supplies. Welding machines could be grouped by process, output, voltage and use case. Consumables needed clearer compatibility, size and material data. PPE needed fit, safety and application details to be easier to compare.

The same attribute-led approach applied to product listing pages. Better attributes and better category guidance made listing pages more useful. Customers could understand the category faster, narrow choices more confidently and reach relevant products with less friction.

AI stayed secondary to ecommerce judgement. It helped us make steady progress on catalogue quality when the manual version of the job would have been too slow and too expensive.

Building commercial resilience while organic visibility recovered

Organic search was still important, but the business could not rely on one channel while catalogue quality work took time to compound.

Shopping feeds became a priority. Better product data, better titles and stronger attribute coverage supported paid and free shopping visibility. We also worked towards Google Top Quality Store status, which helped support trust and shopping performance without treating it as an organic ranking shortcut.

Email and affiliate activity helped diversify revenue. Those channels gave the business more ways to reach customers while organic visibility was more volatile.

Margin visibility also became part of the work. Revenue alone can hide weak ecommerce economics, especially when product ranges, suppliers, delivery costs and paid channels all vary. A custom backend view showed margin against completed orders, which made it easier to understand real profitability rather than only top-line sales.

At that point, the project became more than an SEO rebuild. The site needed traffic, but the business also needed better channel mix, cleaner product data and a clearer view of which orders were worth chasing.

Organic growth from the pre-acquisition baseline

The results are measured against 2021 as a clean pre-acquisition benchmark. Ownership changed at the end of 2022, so the comparison avoids a simple “since acquisition” claim.

From that baseline, organic sessions grew 232%, organic orders grew 72%, and organic revenue grew 97%.

Organic sessions grew fastest, which is worth saying plainly. Not every extra visit turned into revenue at the same rate. Some of the growth came from broader visibility, better category coverage and stronger discovery across the catalogue.

The analytics data also understates the commercial influence of search. Welding & Welder has a B2B buying pattern where discovery and purchase can happen in different places. A customer may find the product through organic or paid search, send the link internally, and the final order is then placed later by a procurement team or another department as direct traffic.

Direct revenue supports that reading. Direct share of revenue rose from 22.6% in the baseline year to 37.1%, peaking at 43.6% during the period. Some of that is genuine brand and repeat demand. Some of it is also likely to be earlier organic or paid discovery finishing through a direct product link once the purchase moves into an internal buying process.

Last-click organic revenue is useful, but it only captures part of search’s commercial influence. The larger opportunity was to make the site better at turning product demand into confident buying journeys, cleaner product management and more profitable sales.

Sustainable ecommerce SEO came from improving the business behind the rankings: structure, catalogue quality, customer journeys, feeds, margins and the way product information was managed.