SEO Audits That Change Something: A 12-Point Sequence for Indian Websites
Most SEO audits are 60-page PDFs that get read once and actioned never. The problem is not the findings; it is that they arrive as an undifferentiated list, so the client fixes the three easiest items and ignores the one that was costing them traffic.
This is the sequence I run instead, in order. Each step is a gate: if it fails, fixing anything below it is wasted work.
1. Is the site indexable at all? Check robots.txt for a stray Disallow: / left from staging, and spot-check templates for a noindex meta tag. This sounds trivial and it is the single most common cause of a site that “suddenly lost all rankings”.
2. Do canonicals point where they should? Every page should either be self-canonical or deliberately point elsewhere. Canonicals pointing at old URLs after a migration silently tell Google the current page does not matter. I have found this on eleven of the last thirty audits.
3. Is the sitemap honest? It should list only indexable, canonical, 200-status URLs, and it should be referenced in robots.txt. A sitemap full of redirects and 404s trains Google to trust it less.
4. Do redirects chain? Old URL to new URL in one hop. Chains of three or more leak signal and slow crawling. Check especially anything from the site's last redesign.
5. Is there one page per keyword? Build the keyword-to-URL map and look for two pages targeting the same term. Cannibalisation caps both pages, and no amount of link building fixes it.
6. Do titles and H1s match search intent? Not keyword stuffing. Does the title promise what the query wants, and does the page deliver it in the first screen?
7. Is internal linking doing anything? Your most important commercial pages should have the most internal links pointing at them with varied, descriptive anchors. On most sites the opposite is true: the privacy policy is the best-linked page in the footer and the money pages get one link each from the navigation.
8. Is the content actually better than what ranks? Open the top three results and ask what they contain that you do not. Usually it is specifics: prices, dates, numbers, named examples.
9. Do Core Web Vitals pass in the field? Field data, not the lab score. Fix Largest Contentful Paint first; it is nearly always an oversized hero image.
10. Is structured data present and valid? Organization, LocalBusiness where relevant, FAQPage, Breadcrumb. Validate it rather than assuming the plugin got it right.
11. Is the entity consistent? Same business name, same description, same address across the site, Google Business Profile, LinkedIn and every directory. This matters more each year as AI answer engines assemble entities from scattered sources.
12. Is there a baseline? Record average position, impressions and conversions for the target keyword set, dated, before anything changes. Without it, nothing that follows can be proven.
Run it in that order and the audit produces a work queue rather than a document. Anything below step five is premature if steps one to four are failing, which on an unaudited site they usually are.
The value of a sequence like this is that it is boring and repeatable. I use the same twelve steps on every engagement, and the first four are where roughly half of all recovered traffic comes from. More on how I work is at Deep Bhardwaj, and the companion piece on choosing between an expert and an agency covers who should be running this for you.