How to Do SEO: A Tactical Playbook You Can Run This Week
To do SEO, you remove what stops search engines from reading your site, map each keyword to one page that can win it, then publish and connect content until those pages earn links and clicks. The work is repetitive, measurable and intentionally slow. Most small teams see their first real movement in 8 to 12 weeks, and compounding results after two quarters.
No call required. Send your domain and we will reply with the first three fixes we would make.
How to Do SEO in Seven Moves?
You do SEO in seven moves: audit crawl and index health, build a keyword map, fix existing pages, publish for demand you can serve, add internal links, earn a few genuine backlinks, and review one dashboard monthly. Run them in that order, once, then repeat the strongest two every month. Nothing here needs a large budget.
Why the order matters more than the effort
Most teams start with content because content feels like progress. Then they publish twenty articles onto a site that search engines crawl badly, index slowly, and rank for nothing. Six months later the folder is full and the traffic graph is flat.
The sequence below is not decoration. Each chapter assumes the previous one is finished, because a keyword map built on a broken crawl produces pages that cannot be indexed, and backlinks pointed at a thin page convert nothing. If you are new to the vocabulary first, read Whats Is SEO? before you continue. Otherwise, work top to bottom.
Fix access first
A page that cannot be crawled cannot rank, no matter how good the copy is.
Then map demand
One keyword, one page, one job. Decide this before writing a word.
Then compound
Links and updates stack. Six months of dull consistency beats one sprint.
The seven chapters, in running order
Find what is blocking you before you add anything new
Open Search Console and read three reports: page indexing, Core Web Vitals, and the query list. Look for pages marked crawled but not indexed, duplicate titles, and soft 404s. On the site itself, run a crawler and sort by status code. This is the cheapest hour in the whole playbook, because every fix here lifts everything you do later.
The index sweep
- Use when
- You have never audited the site, or traffic dropped without an obvious cause.
- Do this
- Export all URLs, tag each as indexed, excluded or error, then fix the errors in order of internal links pointing at them. Re-check in 14 days.
Build a map, not a list
A keyword list is useless on its own. A map assigns each query cluster to one URL that already exists or should exist. Group by intent, not by wording: a buyer searching for a supplier and a student searching for a definition will never be served by the same page. Cap yourself at twenty clusters for the first quarter. Fewer, finished, beats many, half-built.
One page, one job
- Use when
- Two pages compete for the same query, or a page targets five unrelated terms.
- Do this
- Write the primary query at the top of the brief. If a section does not support that query, it belongs on another URL or nowhere.
Fix what you already own
Existing pages usually respond faster than new ones. Rewrite the title tag so it states the query and the benefit, open the body with a direct answer in the first two sentences, add the missing subtopics competitors cover, and make sure one clear action sits above the fold. If on-page work is not your strength, a specialist team can carry it faster than you can learn it mid-quarter; that is exactly what On-Page SEO Services covers.
The 40-minute page rescue
- Use when
- A page ranks on page two for a query you actually sell.
- Do this
- Answer the query in sentence one, add the three subtopics the top results all include, link two relevant pages into it, then request re-indexing.
Publish against demand you can actually serve
One useful page a week is a realistic cadence for a small team, and it beats four thin posts. Write the answer first, then the evidence: specifications, process, lead times, failure modes. Buyers reward specificity and search engines reward coverage of the subtopics that a complete answer requires. Do not publish a page you would not send to a customer who asked the same question by email.
Brief before draft
- Use when
- Anyone other than you writes the page.
- Do this
- Lock the query, the reader, the promised answer, five must-cover subtopics and one internal link target. Writers who get a brief deliver usable drafts.
Connect pages so equity and readers can travel
Internal links are the only ranking lever you fully control. Every new page should receive at least two links from older, stronger pages, and each link should describe where it goes. Anchor text is a label, not decoration. Build topic clusters with one hub page and supporting pages that link up to it and across to each other.
The hub pass
- Use when
- A cluster has four or more pages.
- Do this
- Link every child page to the hub with the hub's exact topic, and link the hub out to each child once. Never leave an orphan page published.
Earn a small number of links that mean something
You do not need hundreds. Ten links from pages that real people read will outperform a thousand directory placements. Supplier pages, trade associations, local press, partner blogs and genuinely useful resources all work. The pattern is always the same: find a page that already links to similar things, then give the editor a reason to add yours.
Ten-reply outreach
- Use when
- You have one page worth citing and no links yet.
- Do this
- Send ten short personal emails a week to pages that cite similar resources. Reference the exact paragraph you would improve. Expect one yes in thirty.
Watch one dashboard, monthly
Track four numbers: indexed pages, impressions, clicks and inquiries from organic sessions. Position is a diagnostic, not a trophy. If impressions rise but clicks do not, your titles are weak. If clicks rise but inquiries do not, the page promises the wrong thing. Review monthly so you can separate a trend from a bad week.
The monthly four
- Use when
- The first of every month.
- Do this
- Record the four numbers in one sheet, write one sentence on what changed, and pick one page to rescue before adding anything new.
Your week, in five slots
SEO fails when it is a project. It works when it is a rhythm. This is the cadence we run for most small accounts, and it fits inside one working day spread across five mornings.
| Slot | Task | Output | Time |
|---|---|---|---|
| Monday | Read Search Console: new errors, new queries, pages losing impressions | One fix list | 30 min |
| Tuesday | Publish or rewrite one page against its brief | One live URL | 90 min |
| Wednesday | Internal link pass on the newest page | Two inbound links | 45 min |
| Thursday | Outreach: ten emails to pages that cite similar resources | Ten sends | 60 min |
| Friday | Check that everything published this week is indexed | Clean index | 20 min |
Sample cadence for a site with fewer than 200 URLs. Scale the Tuesday slot, not the Monday one.
The 90-day route
Three phases, one promise each. Do not skip to phase three because phase one feels boring; the boring phase is what makes the rest compound.
Five mistakes that cost the most time
Publishing before indexing works
New URLs on a site with 400 soft 404s get crawled less, ranked later, and quietly waste the writing budget.
Chasing volume instead of intent
A high-volume informational query rarely converts. A low-volume supplier query often pays for the quarter.
Rewriting the whole site at once
You lose the ability to tell which change worked. Change one variable, wait two crawl cycles, then judge.
Buying links that no one reads
Placements on pages with no traffic pass almost nothing and create cleanup work later.
Reporting position instead of inquiries
Rankings are an input. A stakeholder who sees inquiries every month keeps funding the work.
Want the playbook run for you?
We audit, map, fix and report while your team keeps shipping. Send a domain and get a first take in two working days.
How we would run it
Access audit
Crawl, index and speed review with a ranked fix list.
Keyword map
Clusters assigned to URLs, with priority and effort.
Build cycle
Page fixes, new pages, internal links, outreach.
Monthly review
Four numbers, one decision, next month's queue.
Every engagement starts with the same question: what is the smallest set of changes that would make this site earn one more inquiry a month? We answer that in writing before we propose anything larger.
Questions we hear every week
How long does SEO take to work?
Plan for eight to twelve weeks before you can call a movement real, and two quarters before the curve bends. Sites with clean indexing, few competitors and strong intent move faster; new domains in crowded categories move slower. Anyone promising page one in thirty days is describing paid ads.
Do I have to publish new content every week?
No. Rewriting pages that already have impressions is usually the faster win. A team that ships one useful new page a month and rescues two old ones will outperform a team dumping four thin posts a month.
How many keywords should one page target?
One primary query and its close variants. If you write down two primary queries and they would not appear in the same sentence, they belong on two pages.
Do backlinks still matter?
They matter less than relevance, more than most people hope. A handful of links from pages that real buyers read will help a good page rank. A thousand links from pages nobody visits will not rescue a thin one.
What if my site has fewer than 50 pages?
Then you can skip most of chapter one and spend the time on the map and on-page rewrites. Small sites win by depth on a narrow topic, not by volume.
Can I do this without a tool subscription?
Yes for the first quarter. Search Console plus a free crawler covers diagnosis, index status and queries. Paid tools save time on competitor research, not on execution.