SEO Strategy: Four Decisions and a 90/180/365 Plan
The planning document, not another definition — the audit you start from, the four choices that make a strategy more than a task list, and how to tell at each checkpoint whether it is working or you are simply early.
Which signal moves first. Judging month three on enquiries is how good strategies get killed.
A task list is not a strategy. Fifty blog posts, a plugin and a monthly report is a task list.
A strategy is the shorter document behind it — the one that says what you are not doing, and why the order is this order.
This page assumes the decision is already made. If you are still working out what organic search is and how Google's ranking systems fit together, that is what our SEO guide is for. Nothing below redefines SEO. This is the plan you write once you have read that.
Start with the audit that tells you where you actually are
You cannot choose keywords before you know what already gets seen. Four facts come first, and all four are free to get.
| What you must know | Where it comes from | The decision it feeds |
|---|---|---|
| How many pages are indexed, and which are not | Search Console Page indexing report | Whether you have a publishing problem or a crawling problem |
| Which queries already return you at all | Performance report, queries view, last 3 months | Which clusters you are already partway into |
| Which pages get impressions but no clicks | Performance report, pages view | Your cheapest wins — the page exists and is seen |
| Which two or three pages produce enquiries today | CRM or form submissions, matched to landing page | Which pages must not be touched carelessly |
Indexing first, because it is binary. Google's Page indexing report documentation sets the standard: "You should not expect all URLs on your site to be indexed, only the canonical pages", but "all of your important pages should be indexed" (Search Console Help; no revision date published on the page, checked 28 September 2026). If a money page sits under "Crawled — currently not indexed", keyword planning will not save it.
Then impressions, because an impression is the earliest honest signal you have — it means someone saw, or could have seen, a link to your site in Search, Discover or News (Search Console Help, no revision date published, checked 28 September 2026). A page with 900 impressions and four clicks is not a failure. It is a title and intent mismatch: a week of work, not a quarter.
If the tool is new to you, our Search Console guide covers the reports and filters. Treat the last few days as provisional — recent data "can be preliminary, meaning it's still being collected and might change in the next few hours" (Performance report Help, no revision date published, checked 28 September 2026).
The four decisions that make it a strategy
A strategy is defined by its exclusions. If your document could have been written for any competitor in your category, it is a task list. These four choices are what make it yours.
Decision 1: which queries you are choosing not to chase
Write the excluded list before the target list. It is shorter and it settles more arguments.
The test is intent, not difficulty. Our search intent guide covers how to classify a query from the results page it produces; sourcing and volume work sits in keyword research. Strategy is only the choosing.
- Exclude queries where every top result is a directory, a marketplace or a "top 10" list you would have to become to compete.
- Exclude high-volume informational queries with no next step on your site, however tempting the number.
- Exclude queries that describe your product in words your buyers never use — which is most internal jargon.
- Keep lower-volume queries that name a problem you get paid to solve, and any query where you already have impressions. You are partly there.
Decision 2: which pages carry the commercial intent
Name them. Usually three to eight — the service pages, the category pages, a location page or two. Everything else exists to send attention and links to those.
This is the decision most in-house plans skip, and it is why blog traffic so often arrives without enquiries. If no page is nominated to convert, the blog becomes the destination, and a blog post is a poor closing argument.
Write one sentence per page: "this page ranks for these queries, and its job is to produce an enquiry." If a page cannot carry that sentence, it is a support page, and it should link to one that can.
Decision 3: what the architecture will be
Decide the shape now, because retrofitting it later means redirects. One commercial page per intent, supported by pages that answer the questions around that intent and link inward.
Two pages chasing the same intent is the most common self-inflicted wound in Indian SEO, and it comes from a content calendar written without an architecture. Keyword cannibalisation covers the diagnosis and the merge; internal linking covers how support pages actually pass signal to the commercial one.
Google's own argument for consolidating, useful when a stakeholder wants both pages: it "helps search engines to be able to consolidate the signals they have for the individual URLs (such as links to them) into a single, preferred URL", and avoids Googlebot "spending crawling time on duplicate pages" (Google Search Central, last updated 10 July 2026).
Decision 4: what you will stop doing
Every SEO plan is added to an existing workload. If nothing comes off, the plan gets done at 60% and reported at 100%.
The usual candidates: publishing weekly for its own sake, chasing directory listings, updating pages nobody reads, and producing a twelve-tab monthly report nobody opens. Decide which one dies and write it in the document.
Why the first quarter is repair, not publishing
Fixing and consolidating what exists beats publishing new pages for the first three months, for a reason that has nothing to do with effort: existing pages already have history, links and impressions, and a new page has none.
A page Google has crawled, indexed and shown 900 times is one whose title, intro and internal links you can change this week. A new page starts at zero and has to be discovered before it can be judged. Google's framing on timing is loose on purpose: "Some changes might take effect in a few hours, others could take several months" (SEO Starter Guide, last updated 10 December 2025).
So the first quarter is a triage pass, not a calendar. Keep, merge, rewrite, delete — those mechanics live in our content audit guide, the right starting point for anyone whose instinct is "we should publish more". The technical half sits in technical SEO.
- Merge pages competing for the same intent into one, and redirect the losers.
- Rewrite titles and opening 100 words on pages with impressions and almost no clicks.
- Fix anything the Page indexing report flags on a page you care about.
- Link existing support pages into the commercial pages from Decision 2.
- Delete or consolidate pages that exist only because somebody had a publishing target.
Publishing starts in month four, against the architecture, not against a spreadsheet of keywords.
What 90, 180 and 365 days realistically look like
Organic work takes roughly three to six months to produce results you would call meaningful, and it compounds after that. Nobody can compress the first part, and Google states plainly that "No one can guarantee a #1 ranking on Google" (Do you need an SEO?, last updated 5 June 2026).
| Window | What the work is | What you should be able to show |
|---|---|---|
| 0–90 days | Audit, indexing fixes, merges and redirects, rewritten commercial pages, internal links built | Indexing status changed on the pages you named; more distinct queries returning impressions |
| 90–180 days | Publishing against the architecture — roughly 8 to 15 pages answering the questions around your commercial intents | Positions moving on the clusters you chose; clicks rising on repaired pages; first non-brand enquiries |
| 180–365 days | Refresh what half-worked, extend the clusters that moved, earn links, add depth where competitors are thin | Non-brand organic as a stated share of enquiries; earlier pages still climbing |
The compounding is the reason to do this at all. A page published in month five is still earning in month twenty, which is not true of a campaign you switched off.
One planning note for the back half: budget for refreshing, not only producing. Most plans give the whole content budget to new pages and have nothing left for the ones that nearly worked. From month six, roughly two-thirds new and one-third refresh is our working rule of thumb — not a benchmark, and we know of no published figure for it.
How to tell whether it is working or you are just early
Signals move in a fixed order, and judging the wrong one too early is how sound strategies get cancelled in month three. Impressions and query coverage move first, positions second, clicks third, enquiries last.
| Signal | Typically moves | Reads as working if… | The trap |
|---|---|---|---|
| Distinct queries returning impressions | Weeks 3–8 | The count of queries you appear for is widening, even at low positions | Ignoring it because positions are 40-plus. Coverage before rank is the normal order |
| Impressions on named pages | Weeks 4–10 | Repaired pages are being shown more often | Reading the last three days, which Google says may still be preliminary |
| Average position | Months 2–5 | The trend over a quarter tilts up on your chosen clusters | Comparing month to month. This number drifts |
| Clicks from non-brand queries | Months 3–6 | Clicks rise while brand-name queries stay flat | Counting brand searches as SEO working |
| Enquiries attributable to organic | Months 4–9 | Enquiries land on the commercial pages from Decision 2 | Expecting this in month two and killing the plan |
On average position, treat it as drifting rather than as a measurement. Search Console reports "the topmost position occupied by a link to your property or page in search results, averaged across all queries", and adds that the heuristics behind it — including the position numbering — "are subject to change" (Search Console Help, no revision date published, checked 28 September 2026). A move from 14.2 to 12.8 in a month can come entirely from which queries you happened to appear for. Read it over a quarter on a filtered set of queries you chose, or do not read it.
The honest month-three review is two sentences: coverage is widening or it is not, and the pages we repaired are being shown more or they are not. Both yes with flat enquiries means you are early. Coverage flat after a quarter of shipped work means the strategy is wrong, usually at Decision 1. Cadence and dashboards sit in SEO reporting.
What one person can run, and what needs outside help
One competent in-house marketer, at roughly a day and a half a week, can genuinely run the first 90 days and a modest publishing programme after it. What they cannot run is anything needing three skills at once in the same week.
- A one-person in-house team can run: the Search Console audit, the three-list inventory, title and intro rewrites, internal linking, the excluded-query decision, two to four pages a month, and the monthly read.
- Needs a developer either way: redirects at scale, canonical and indexing fixes, page speed, anything touching templates. Budget those hours or the plan stalls in week three.
- Usually needs outside help: a site migration, a multi-hundred-page consolidation, technical work on a platform nobody in-house understands, competitive clusters where the content has to beat a well-funded incumbent's, and link earning.
- Rarely worth outsourcing: the four decisions. Nobody outside your business knows which enquiries you actually want.
On fees, as a directional market range and not a rate card: SEO retainers in India run roughly ₹25,000–₹1,50,000 a month depending on site size, competition and how much content production is included, and those fees sit outside tooling and any media spend. The bottom of that band buys advice and a few hours of execution, not a publishing programme. Ours is one of the options — our SEO services — and a business that already has a developer and a writer on payroll often does better keeping the work in-house and buying two days of strategy instead.
If the plan depends on steady content output, settle who owns that before you start. Briefs, calendar and distribution are a separate discipline, covered in our content marketing guide.
Where to go from here
Write the document this week, on one page: four decisions, a 90-day repair list drawn from the three-list inventory, and the two signals you will judge month three on.
Then leave it alone for a quarter. The common failure in SEO is not a wrong strategy — it is three strategies in nine months, none given long enough to read.
Frequently asked questions
What is an SEO strategy?
An SEO strategy is a written set of choices about organic search: which queries you will chase and which you will deliberately ignore, which pages carry the commercial intent and have to convert, what the site architecture will be, and what current activity you will stop to make room. It begins with an audit of what already indexes and already collects impressions, spends its first quarter fixing and consolidating existing pages rather than publishing new ones, and is judged in a fixed order — query coverage first, positions second, enquiries last.
How long does an SEO strategy take to show results?
Plan for roughly three to six months before organic work produces results you would call meaningful, and expect it to compound after that. Google's own guidance is deliberately loose: "Some changes might take effect in a few hours, others could take several months" (SEO Starter Guide, last updated 10 December 2025). Query coverage and impressions usually move within the first two months, average position over a quarter, and enquiries last. Google also states plainly that "No one can guarantee a #1 ranking on Google."
What should the first 90 days of an SEO plan contain?
Repair, not publishing. Audit indexing in Search Console, build an inventory of which URLs are submitted, which are indexed and which have had at least one impression in 90 days, then work the gaps: fix indexing on pages you care about, merge pages competing for the same intent and redirect the losers, rewrite titles and opening paragraphs on pages that get impressions but almost no clicks, and build internal links into the commercial pages you nominated. Existing pages already have history and impressions; new pages start at zero.
How do I know if my SEO strategy is working or if it is just early?
Check the signals in the order they move. The count of distinct queries returning impressions widens first, usually in weeks three to eight. Impressions on the pages you repaired follow. Average position tilts over a quarter, never month to month — Search Console reports the topmost position averaged across all queries and notes that the heuristics behind it are subject to change. Non-brand clicks come next, and enquiries last, often months four to nine. If coverage is widening but enquiries are flat at month three, you are early. If coverage is flat after a quarter of shipped work, the query choices are wrong.
Can one person run an SEO strategy in-house?
Yes, for the first 90 days and a modest publishing programme after it, at roughly a day and a half a week. One in-house marketer can run the Search Console audit, the URL inventory, title and intro rewrites, internal linking and two to four new pages a month. They will still need developer hours for redirects, canonicals, speed and template changes. Outside help earns its fee on migrations, large consolidations, unfamiliar platforms, genuinely competitive clusters and link earning — but not on the four strategic decisions, which need someone who knows which enquiries the business actually wants.
We will tell you what to stop doing first
An audit of what already indexes and ranks, the four decisions written down, and a 90-day repair list you could hand to a developer tomorrow.
Related guides
Make Digital Hangover a preferred source
One tap tells Google to show more of our SEO and marketing coverage in your Top Stories.