Forty URLs a month, and what forty honestly means
Most on-page proposals commit to a vague "ongoing optimisation". That's unfalsifiable, which is precisely why it's popular. We publish a number instead: 40 URLs a month, listed by URL in the monthly report, with the date each one changed.
Here's the uncomfortable part of that number. Forty full rewrites at 45-odd delivered hours a month would be about an hour per page, and an hour per page is not a rewrite — it's a haircut. So the forty splits. Around ten to twelve are deep rewrites: the page is re-researched, restructured and largely rewritten against the query it's meant to own. The remaining twenty-eight to thirty are structured passes: title, meta description, H1, opening answer block, heading hierarchy, internal links in and out, alt text, schema.
Both are worth doing. A structured pass on a page that already ranks at position 6 with a lazy title routinely earns more clicks than a deep rewrite of a page nobody finds. But you should know which one your page got, so we label it.
The per-URL checklist
Every URL, deep rewrite or pass, goes through the same nine checks. It's deliberately boring. Boring is what survives staff turnover.
- Title tag — the target query near the front, a reason to click after it, and under about 60 characters so Google doesn't rewrite it for you. Pound for pound it's the most consequential string on the page.
- Meta description — not a ranking factor, plainly a click factor. Written as a promise about what's on the page, because Google will replace it with page text if it doesn't match.
- H1 — one per page, matching the search intent rather than your internal product name. The number of Indian sites whose H1 is their own logo alt text is remarkable.
- Opening answer block — the first 40 to 60 words answer the query directly. This is what a featured snippet and an AI overview lift, and burying it under a paragraph of preamble costs you both.
- Heading hierarchy — H2s that are real sub-questions, in the order a reader asks them. If your H2s read as "Introduction / Benefits / Conclusion", the page has no structure to extract.
- Body copy — the specific claim, the number, the caveat. Thin sections get merged or deleted rather than padded.
- Internal links — in and out, with descriptive anchors. Covered properly in the next section.
- Images and alt text — descriptive alt, sensible filenames, correct dimensions and format. Also the fastest way to fix a bad LCP score on a content page.
- Schema — Article, FAQPage, Product, BreadcrumbList or Organization as appropriate, validated, and only ever describing content that's visibly on the page.
Internal linking: how many, and from where
Internal linking is the only ranking lever you fully control, it costs nothing but attention, and it's the first thing dropped when a retainer is under-resourced. It's also invisible on an invoice, which is why it goes missing quietly.
Our rule on every rewritten URL: it leaves with at least three inbound in-content links from relevant existing pages, and at least two outbound to related pages. In-content, not footer — a link inside a sentence somebody might read carries more than the same URL repeated on every page of the site.
Once a quarter we run a full pass over the link graph rather than page by page: orphan pages that receive nothing, pages that sit five clicks from the homepage and shouldn't, hub pages that link down to a cluster but never get linked back up, and anchor text that says "click here" 200 times. On older Indian sites that have accumulated a few hundred blog posts, this pass alone tends to be the highest-return work in the quarter — no new content, no new links, just redistribution of authority the site already earned.
What's included and what isn't
On-page work sits inside the SEO retainer rather than being sold separately, for the honest reason that on-page changes to pages nobody links to and nobody can crawl don't do much.
| Item | Included | Not included |
|---|---|---|
| Rewrites and passes | 40 URLs a month, labelled deep rewrite or structured pass, listed by URL with the date changed | Unlimited revisions of the same page because a stakeholder prefers different adjectives |
| Copywriting | Rewriting existing pages, including new sections where a page is missing an obvious answer | Brand voice development, taglines, ad copy, or writing net-new articles — that's the content workstream |
| Internal linking | Per-URL links in and out, plus a quarterly full link-graph pass and orphan fix | Rebuilding your navigation or information architecture. We'll spec it; implementing it is a project |
| Schema | Design, implementation and validation for the templates we touch | Schema for content that doesn't exist on the page. That's a manual action waiting to be issued |
| Implementation | Direct edits in WordPress, Shopify, Webflow or Framer, or pull requests against your repo | Design changes, new page templates, or anything needing a front-end build |
| Measurement | Per-URL CTR and average position before and after, from Search Console, in every monthly report | Third-party "on-page score" or "SEO grade" numbers. They measure a tool's opinion, not a searcher's |
₹75,000 and ₹40,000, and what happens on a 5,000-page site
The forty-URL commitment sits inside the ₹75,000/mo SEO retainer, ex-GST, alongside technical work, content and links. On the ₹40,000/mo small-site tier the number drops to around 15 URLs a month, with roughly four deep rewrites — which is the right shape for a 40-to-60-page site, where forty a month would mean rewriting your whole site every six weeks for no reason.
Then the other extreme. On a 5,000-page ecommerce catalogue or a publisher archive, 40 URLs a month is arithmetic nobody should accept — at that rate you'd finish in about ten years and need to start again. Page-by-page work is simply the wrong tool.
Instead we fix the templates and the rules, because one change to a category template rewrites 900 titles at once, then spend the manual forty on the pages that warrant individual attention: top revenue URLs, top-impression underperformers, and the handful a human actually reads before spending money. That's a different scope and we quote it as one.
Ex-GST throughout, month-to-month after the first quarter, 30 days' notice, every asset yours. Rate card on our pricing page.
- Under 60 pages — ₹40,000/mo tier, roughly 15 URLs a month. You'll run out of pages worth optimising before you run out of months, and that's the point at which the budget should move to new content.
- 60 to 800 pages — the ₹75,000/mo retainer with 40 URLs a month is a sensible fit. Full coverage inside two years, priority URLs inside the first quarter.
- Over 2,000 pages — template and rule changes first, manual work reserved for the commercially critical tail. Quoted separately, because pretending otherwise would just be a slower way of failing.
How we grade a rewrite, and what the guarantee covers
Every rewritten URL gets tagged with the date it changed. Twenty-eight days later we pull that URL's clicks, impressions, CTR and average position from Search Console and compare against the 28 days before the change.
One honest caveat, because the number is easy to abuse: CTR rises naturally as average position improves, so a rewrite that coincided with a position gain hasn't necessarily earned its CTR change. We report both side by side and say which we think moved first. An agency showing you CTR without position is showing you half a story, usually the flattering half.
The wider guarantee is unchanged. On day one we freeze your trailing-90-day count of qualified leads from organic search, in writing, before anything is touched. If we haven't beaten it in 90 days, we keep working free until we do. For on-page work we additionally freeze the CTR and average position of the URL set we're committing to rewrite, and report against that frozen list rather than a blended site average. We never promise a specific ranking position for a specific keyword — nobody controls Google's index, and a promised position is a coin flip with a contract attached.
On-page work has a ceiling worth naming, too. If the page is unindexable, technical fixes come first. If the query needs a page you don't have, that's content, not a rewrite. And if the category is owned by sites with ten times your referring domains, better titles will not close it.