Technology
We build our own tools for the repetitive parts of search and site operations work, so that more of the time on each project goes into decisions and fixes.
YellowArc Site Audit
The crawler and reporting tool we use at the start of every search engine optimisation project.
What it checks
- Status codes and redirects
- Titles and meta descriptions: missing, length and duplicates
- H1 headings
- Canonical tags and robots directives
- Structured data and Open Graph tags
- Image alt text
- Amount of main text on each page
- HTML response time
Drafted titles and descriptions, checked against the page
For pages with missing, weak or duplicate titles and descriptions, the tool drafts replacements with Anthropic’s Claude API. Each draft is written only from the text on that page. It is then checked automatically against length limits and the site’s brand format, and for any number or name that does not appear on the page. Drafts that fail a check are marked for review, and nothing is published until a person has approved it.
Polite crawling
The crawler follows robots.txt, reads XML sitemaps, identifies itself with its own user agent and waits one second between requests by default.
Output
An HTML report for the site owner, a CSV file for developers with every page, issue and suggested change, and a JSON file we use to compare audits over time.
Planned next
Tools we intend to build on the same approach: real data in, drafts checked automatically, a person approving every change.
Search and revenue brief
A weekly summary that combines Search Console, analytics and ad-network data into a short list of prioritised actions.
Operations assistant
Help with reading server logs, triaging incidents and reviewing configuration changes before they are applied.
Hosted audits
Site Audit offered as a hosted service, so clients can run checks on their own websites between engagements.
Discuss a project
Tell us about your website and what you would like to improve. We reply by email to arrange a short call.