SEO blueprint: what goes in one, and how we built ours

What is an SEO blueprint?

An SEO blueprint is the written plan that decides which searches you want to win, which single page answers each one, what happens to the URLs you already have, and how you will know it worked. An audit tells you where the site stands. The blueprint turns that into decisions, in order, with an owner for each.

This guide starts where the audit ends. For what the audit should check and how to rank its findings, read what an SEO audit should cover.

Each step gives the method, then how we applied it when we rebuilt pixelcrayons.com, which went live on 29 September 2026. The figures come from our own planning records and carry their dates.

What should you decide before researching keywords?

Decide the goal, the offers that matter most and the buyer before you open a keyword tool. Those three choices tell you which searches to ignore, however large they look. Without them, research produces long lists of phrases that nobody acts on, and the page map fills with topics you do not sell.

• Pick one goal to lead, such as qualified leads, online sales or a new market.
• Rank the services or products you most want to sell through search. They get pages and effort first.
• Write down who searches, what sends them to Google and what they ask before they buy. Your sales team hears those questions every week.

For our own site, scope came first. When we sized our rankings in October 2026, searches for industries we do not serve, such as fintech marketing, were left out whatever their demand.

How do you size keywords and decide what to build?

Group close variants into one topic, then judge each topic on three things: demand, difficulty and who ranks today. Volume alone misleads. A topic is worth a page when buyers search for it, when it matches what you sell and when the current results are weak enough for a new page to beat them.

Intent comes before volume: questions suit guides, comparisons suit lists and case studies, and searches ready to buy suit service pages. Our post on the keyword that’s costing you money shows how to find rankings whose intent never pays.

How we sized it for pixelcrayons.com

On 12 October 2026 we exported 428 search terms from Ahrefs Keywords Explorer across the US, the UK, India, Australia and the UAE. Of those, 266 had measurable demand. We grouped variants such as “best”, “top” and “in India” into one cluster each, so that every cluster could become one page.

Demand did not decide alone. For each leading cluster we read the US top 10: who ranks, how strong their sites are and how many sites link to them. Where Reddit threads, forum posts and thin lists held the top 10, a sourced page had room to win.

That check removed topics too. Branding agencies dropped out, because searchers wanted a local agency. Country editions are built only where that country’s buyers search for their own firms. Google’s spam policies name pages aimed at regions or cities that funnel users to one page as doorway abuse.

The output was a build order, weakest results pages first. It became the PixelCrayons rankings, with 18 scored lists in the first edition, in October 2026.

How do you give each search intent one page?

Write the page map as a table with one row per search intent and one URL in each row. Before adding a page, check whether an existing one already targets that intent; if it does, improve or merge it instead. Two pages chasing one search split the signals, and both tend to rank worse.

Our standard writes this as one primary search intent per page, stated in the title, the H1 and the opening answer. A hub page sells the capability, a spoke sells the specialised variant and a cost guide prices it. None restates a sibling’s pitch; each links to it and stays in its own lane.

The map also decides when to build. In September 2026 our migration plan added one page, for eCommerce PPC, because Ahrefs groups those searches as a topic separate from general PPC. A dedicated page would not compete with the main PPC page.

Placing keyphrases without stuffing them

On 26 September 2026 we checked a sheet of 541 target phrases against the built HTML of 157 pages. Only 55 appeared as written, and 105 pages carried none of their phrases. Every URL in the sheet existed, so the gap was wording, not missing pages.

The plan ranked the missing phrases instead of pasting them in. Each page takes its primary phrase in the H1 or first sentence and in the title, then two or three secondary phrases in H2s or FAQ questions. The rest stay optional.

Fifteen phrases used words our style rules ban, and 18 were superlative claims that would need evidence, so those went for a decision before any copy was written. Turning the finished map into briefs and dates is the job of SEO content strategy.

Should URLs be flat or nested under hubs?

Keep URLs short and stable, and build hierarchy through navigation: hub pages, breadcrumbs and links between related pages. A URL you nest today is a redirect you may need tomorrow. Name each page after the search it should win, and let the parent hub give the context.

When we grouped our rankings into categories on 22 October 2026, every list kept its flat address under /rankings/, unchanged and with no redirects. The hierarchy came from five category hubs, a breadcrumb on each list and an “Also in” line for lists that belong to more than one hub.

One rule keeps it consistent. Every list has exactly one primary hub, which sets its breadcrumb, and the site build fails if a list has none or two.

Which old URLs should you keep, merge, retire or redirect?

Give every old URL exactly one decision and one destination. Keep pages that earn visits or links and fit what you sell, merge pages that answer the same question, and redirect the rest to the closest page that serves the same need. A page with no closer match and no links can return 404.

Google’s site move guidance asks you to map each old URL to a new one and redirect straight to the final destination. It warns that sending many old URLs to one irrelevant page, such as the home page, may be treated as a soft 404. Permanent server-side redirects, 301 or 308, also tell Google the target should be canonical.

Our legacy blog, in numbers

Our legacy blog register, as generated on 11 October 2026, held 242 old posts. Each got one of four decisions.

The redirect map gives every legacy post exactly one entry, and nothing chains. When we exported the live site’s own redirect table on 28 September 2026, all 417 rules were carried across in one hop to a page on the new site, because those rules lived in WordPress and would stop working at cutover.

Links set a second threshold. On 3 October 2026 we found 67 old blog URLs that had links from three or more sites but returned 404, and redirected each to the same post at its new path or to the nearest page. Old blog URLs with fewer linking sites were left as 404s.

Leaving a page gone is sometimes the right call. Of 20 dead legacy URLs found on 6 October 2026, three were redirected to a page serving the same intent. The other 17 stayed gone, because a redirect to a general hub would have been a soft 404.

We also corrected our own first pass. A review on 22 October 2026 found retired posts redirecting to pages that did not answer them, including a general “why us” page. Each now points at its closest page or matching ranking, and no post redirects to the “why us” page any more.

Redirects, canonicals and indexing after a move are the core of a technical SEO review. Most sites fresh from a migration need one.

What should the blueprint say about launch day and the weeks after?

Write the launch checklist and the review dates into the blueprint before anything moves. Say which sitemap you will submit, which pages you will ask Google to index first, when you will compare results with the baseline and what would make you reverse a change. Decide those rules in advance, while nobody is anxious about a dip.

Our plan of 28 September 2026 covered 19 old ranking URLs. Each one that moved went to exactly one new URL with a 301, with no chains, no redirects to the home page, and both the trailing-slash and the bare form covered. No page, menu or sitemap on the new site could link to an old URL.

The new XML sitemap was to list only new URLs and be submitted on launch day, with indexing requested for the ten highest-impression new pages. The old sitemap was to stay reachable for four weeks, so Google would recrawl the old URLs and see the redirects.

We set checkpoints at day 7, 14, 28 and 56, and agreed not to act before day 28, since impressions dip and wobble while Google processes redirects. A reversal would be a per-page switch, not a site rollback. A page with almost no baseline clicks would be fixed rather than reverted.

Google says a medium-sized site can take a few weeks or more to move, and redirects should stay for at least a year. In the page indexing report, moved URLs should show as “Page with redirect”, not as soft 404s.

Our free SEO audit checklist runs 77 checks in ten areas, including crawl and indexing. The same list works before a move and after it.

How does answer-first writing fit into the blueprint?

Make every important page answer its question near the top, in words that still make sense when quoted alone. Google says AI Overviews and AI Mode have no extra requirements: a page must be indexed and eligible for a snippet. So the blueprint needs no separate AI plan, only pages that state their answers plainly.

We wrote that into our own standard: each H2 section’s first sentence answers its heading, and answer blocks run 40 to 60 words. On 10 October 2026 only 20 of the 185 answer blocks on the site were in range. Two days later, after a rollout across every page in the staging build, all 194 were.

Google’s helpful content guidance asks whether a page offers original research or analysis and first-hand expertise. A worked example from your own records meets that test most plainly.

Search Console counts AI feature traffic inside the Web search type, with no separate figure. To know whether AI answers mention you, keep your own baseline of fixed buyer questions, rerun monthly, as our AI visibility audit describes. AEO vs SEO covers what else changes.

How do you measure whether the blueprint is working?

Record a baseline before anything changes, then compare at fixed dates. Search Console shows impressions, clicks and position by page and query; your analytics shows enquiries and sales. Judge the blueprint by leads from organic search, not visits alone, and review it each quarter or after any big change such as a migration.

Check where every number comes from. While planning our migration, a tool’s traffic estimates for our old service pages proved far higher than the clicks Search Console measured, so the plan used Search Console as its baseline.

Search is one channel among several. For a budget decision across channels, the digital marketing audit sets the order of checks.

How can PixelCrayons help?

If you would rather have the blueprint built with you, our SEO team starts from an audit and works through the steps in this guide: goals, sizing, the page map, URL decisions, launch checks and review dates. The difference is that every decision is made on your site, your searches and your records.

Sources

• Google Search Central: Site moves with URL changes
• Google Search Central: Redirects and Google Search
• Search Console Help: Page indexing report
• Google Search Central: Spam policies for Google web search
• Google Search Central: Creating helpful, reliable, people-first content
• Google Search Central: AI features and your website