SEO Howto: Six Structured Steps With Purpose and Output

WebSEOGrow Howto Card Chat on WhatsApp

Howto Card · Six Steps

SEO Howto: Six Structured Steps With Purpose and Output

An SEO howto is a sequence, not a pile of tips. Six steps, and each step has a purpose, an action, and an output you keep. Work them in order and every step hands the next one its input. Skip one and the following steps quietly lose their footing.

6 steps · 3 cells each · 1 card you can keep

Step 3 · Fix structure Cell sample

Purpose

Make the site easy to crawl so new writing is not wasted on pages nobody reaches.

Action

Flatten depth to three clicks, add links from high traffic pages, clear duplicate titles, check canonicals, sitemap, robots, and mobile load.

Output

A crawl report with the fixes applied and one page template the rest of the site can copy.

Every step on this page fits the same three cells. No output, no step.

1Diagnoseblocker list
2Select demandquery map
3Structurecrawl fixes
4Write answerlive page
5Mentionslink sheet
6Measurechange log

Diagnose, demand, structure, copy, mentions, measure. Each step hands the next one its input, and each leaves an object behind.

What is an SEO howto in one sentence?

An SEO howto is a fixed order of work: find the constraint, choose queries you can serve, fix structure, write the answer, earn a few mentions, then measure and keep what worked. The order matters because each step hands the next one its input, so skipping a step does not save time, it borrows it.

If the vocabulary is still unfamiliar, read Whats Is SEO? first, then come back to the card.

Why every step needs three cells, not one

Most SEO instructions fail for a boring reason: they describe an action and stop. You are told to audit titles, so you audit titles. Nobody says what the audit was for or what should exist afterwards, so nothing finishes and the same audit comes back next quarter with a new name.

Three cells close that loop. Purpose keeps you from doing work that does not matter this month. Action makes the step small enough to finish in one sitting. Output turns the step into an object a person can hold: a list, a map, a report, a published page. That object is what makes the next step possible, and it is what you show a client or a boss when they ask what happened last month.

If a step has no output, it is not a step. It is a habit, and habits are hard to audit.

The six steps, one card each

Read the whole card once before you start. The steps are short, but they are ordered for a reason: two through six all consume something that step one produces.

Find the constraint

Diagnose before you touch anything

Purpose

Know which problem is worth fixing before you spend a month on the wrong one.

Action

Export ninety days of queries from Search Console, crawl the site, and mark pages with impressions but no clicks, plus pages nothing links to.

Output

A blocker list of three to seven items, each with a URL and the number it affects.

Write the list on one page. If it does not fit on one page, you have not chosen yet.

Select demand

Choose queries you can actually serve

Purpose

Stop competing for terms you cannot win and start owning a small set you can.

Action

Group the export by intent, open the results for each group, and note the format that wins: guide, comparison, service page, or local page.

Output

A query map with one row per cluster, a target URL, and the format the results reward.

A cluster with no page is a gap. A cluster with two pages is a conflict. Both go on the map.

Structure

Fix the frame before you fill it

Purpose

Make the site easy to crawl so new writing is not wasted on pages nobody reaches.

Action

Flatten depth to three clicks, add links from high traffic pages, clear duplicate titles, and check canonicals, sitemap, robots, and mobile load.

Output

A crawl report with the fixes applied and one page template the rest of the site can copy.

Structure work is boring and it is the step most often postponed. Postpone it and steps four and five underperform for reasons you cannot see.

Write the answer

Publish pages that answer on the first screen

Purpose

Give the searcher the answer immediately, then the detail that earns a click and a reply.

Action

One H1, one intent, an answer in the first sixty words, headings phrased as the questions buyers ask, and links to the next step in their decision.

Output

A published page with title, meta, H1, and an answer block, plus a note of what it replaced.

This is where most of the hours go. Our On-Page SEO Services team takes the query map and turns it into pages that answer before they sell.

Mentions

Earn a few relevant mentions

Purpose

Give the page a reason to be trusted beyond its own claims.

Action

List twenty places your buyers already read, then send a short note about one specific page to each. Trade directories and association lists beat cold outreach.

Output

A link sheet with contact, page, status, and date. Three to five live mentions a quarter is a real result.

Five relevant mentions from places your buyers recognise do more than five hundred from a marketplace.

Measure and hold

Measure, then lock the result in

Purpose

Turn the work into a decision instead of a mood.

Action

Record a baseline the day you change something, then check the same query set at twenty eight and fifty six days.

Output

A change log with before, after, and one word: keep, revert, or repeat.

Without the log, next year you will run the same experiment and pay for it twice.

The six steps on one card

A one page summary you can copy into a document and fill in as you go.
Step	Purpose	Action	Output
1 Diagnose	Find the constraint	Query export, crawl, dead link scan	Blocker list, 3 to 7 items
2 Demand	Pick winnable queries	Cluster by intent, read the results	Query map with target URLs
3 Structure	Make the site crawlable	Depth, links, titles, canonicals, speed	Crawl report and page template
4 Write	Answer the query	H1, answer block, question headings	Published page
5 Mentions	Earn trust	Twenty targets, one page per note	Link sheet with status
6 Measure	Decide and hold	Baseline, day 28 and day 56 checks	Change log with a verdict

What skipping a step actually costs

None of these failures announce themselves. The work still feels busy, which is why people keep doing it.

01No diagnosis

You rewrite pages that were never the bottleneck while the real block stays exactly where it was.

02No query map

You publish pages that answer questions nobody in your market is asking, then wonder why the traffic is cold.

03No structure fix

New pages sit four clicks deep, get crawled once a month, and compete with three older pages for the same words.

04No measurement

You cannot separate a good change from a lucky week, so you keep everything and learn nothing.

Four habits that break the card

Running step four first

Writing before structure is the most common order error. The copy is fine, the page is invisible, and everyone concludes the writing failed.

Keeping the output in your head

If the blocker list lives in memory, it changes every time you remember it. Write it down and date it.

Judging at day three

Early wobble is noise. Set the check date before you ship, then honour it.

Copying a competitor's card

Their step two was built for their catalogue and their margin. Borrow the structure, not the rows.

How we run this card for clients

The card is deliberately small. Running it well is mostly a matter of refusing to skip cells when a deadline arrives.

Week one: access and diagnosis

We take read access to Search Console and analytics, crawl the site twice, and deliver the blocker list with the numbers behind each item.

Week two: query map in writing

Clusters, intent, target URL, and the format each result page rewards. You approve the map before any copy is written, so scope is agreed early.

Weeks three to six: fixes and pages

Structure work ships first, then pages in batches of five. Each batch has its own baseline line so a bad week cannot hide inside a good month.

Week seven onward: mentions and reads

We run the mention sheet, then read the same query set at day twenty eight and day fifty six, and write the keep, revert, or repeat decision into the log.

One card, one quarter: a worked example

An industrial parts supplier with about four hundred pages asked for a review after a year of steady but flat traffic. Step one took two days and found the real constraint: sixty percent of site traffic landed on nine pages, and the product pages that carried the margin were four clicks deep with no internal links from those nine.

Step two produced a map of forty clusters. Eleven already had a page, nine had two competing pages, and twenty had nothing. Step three flattened the catalogue and added links from the nine high traffic pages. Step four rewrote the eleven existing pages and built four new ones for the clusters with real demand.

Step five brought six mentions from trade directories and one association page. Step six recorded the baseline and set the read dates. The result was not a spike. It was a slow, uneven rise in impressions on the mapped clusters, and a written rule that no new product page ships without an owner and an internal link.

Want this card run on your site?

Send the domain and one sentence about what has gone flat. We reply with the first three blockers we can see, not a slide deck.

SEO howto questions we hear

Do I have to follow the six steps in order?

Yes for the first pass. Each step produces the input the next one needs, so writing before structure usually means rewriting. After one full pass you can loop back to any step as new data arrives.

How long does one pass of the card take?

For a site under five hundred pages, six to nine weeks at a few hours a week. The slow part is usually step five, because it depends on other people replying.

What if I only have two hours a week?

Run one step at a time. Two hours a week is enough for steps one and two if you stop when the output exists rather than when the research feels complete.

Can I skip link building if my pages are good?

You can skip mass link building. You should not skip relevant mentions entirely, because a page with no external reference has a harder time being trusted for competitive queries.

How do I know the card is working?

The change log answers that. If three of five shipped items moved the query set at day fifty six, the card is working, and you can name which cells did the work.

Is this the same as an SEO audit?

No. An audit is step one done thoroughly. The card continues into query selection, structure, copy, mentions, and measurement, and each of those produces its own output.