I’ve Spent Years Refining a 10-Step SEO System. Here’s How to Use It.

When founders ask me how to start with SEO, they usually expect a tool recommendation. The honest answer is that the tool isn’t the problem. The sequence is.

For the past few years, my agency has been refining the same end-to-end SEO process — the one I now use with every new client and break down across an eight-module curriculum we run for our team. The steps work because they’re ordered. Most founders fail at SEO not because they skip steps but because they do them in the wrong order: writing content before researching keywords, building links before fixing crawl errors and chasing traffic before defining what kind of traffic moves their business.

Here’s the 10-step sequence I follow, in the order I follow it. You can run all of it yourself for the cost of two free tools and a few weekends.

1. Start with niche research, not keyword research

Before you type a single seed keyword, define what your business actually has the right to win on. A skincare brand selling to dermatologists shouldn’t try to rank for best moisturizer. A SaaS for restaurant owners shouldn’t compete on small business software.

Write down the three or four sub-categories your business owns. Sanity-check each one against the competition in Google’s results. If the first page is dominated by Wikipedia, government sites and major publications, narrow further. Niche before keywords. Always.

Take 30 minutes and write down every phrase your customers actually use to describe their problem, in their own words. Look at your inbound emails, sales call transcripts and product reviews. The vocabulary your customers use rarely matches the vocabulary you use internally — and the customer vocabulary is what ranks.

Once you have a list of 30 to 50 seed terms, open Ahrefs or Google’s Keyword Planner. Tools are for expanding what you already know; they’re terrible at telling you what to know in the first place.

3. Layer in the specialty keyword types your competitors miss

There are at least nine specialty keyword formats most agencies ignore: geographic, seasonal, event-based, question-format, service-based, commercial, comparison, best-of and alternative-to. Each one maps to a different stage of the buying journey, and each one is usually less competitive than the obvious head terms.

In a recent niche project, layering comparison and alternative-to keywords on top of a head-keyword strategy roughly tripled the addressable search volume — without touching a single competitive primary term. The same pattern shows up in nearly every site I audit.

4. Tag every keyword by search intent before writing a word

Every keyword falls into one of four intent buckets: informational, navigational, transactional or commercial investigation. The same phrase can mean different things to different searchers, and the only way to know is to look at what’s currently ranking on page one. Google’s own guidance on understanding user intent comes down to the same principle: match your page to what the searcher actually wants.

If page one is full of blog posts, the intent is informational. If it’s full of product pages, the intent is transactional. Match your page type to the intent before you decide what to write. Mismatched intent is the single most common reason good content fails to rank.

5. Fix your technical foundation before publishing anything new

Before adding new pages, run your site through a free Screaming Frog crawl and Google’s PageSpeed Insights. Look for four things: crawl errors, broken internal links, slow Largest Contentful Paint on mobile and any redirect chains longer than one hop. Each one is silently capping the ceiling on every page you publish.

This is unglamorous work. It’s also the work that determines whether the next six months of effort compound or evaporate.

6. Standardize your on-page template across every new piece

Decide once, then never re-decide: how your title tags are written, how your H1 relates to your title tag, where your primary keyword appears, how internal links are formatted and what your URL structure looks like.

I keep a one-page template that goes on every content brief we send writers. It saves hours of editing per article and produces consistent results across writers who have never spoken to each other. Standardization is what lets you scale; ad-hoc decisions are what burn content teams out.

7. Build content in clusters, not in isolation

For every commercial keyword you target, plan a cluster: one pillar piece and three to five supporting pieces that link inward to it. Search engines reward sites that demonstrate topical depth, and clusters are the cleanest way to demonstrate it.

A single well-built cluster of six pages around one commercial topic will outperform 30 disconnected blog posts every time. Test that against your own analytics if you doubt it. The math is one-sided.

8. Earn links through original data, not cold outreach

Cold outreach link building has been the lowest-yield activity in SEO for at least three years. The replacement is original research: publish one piece per quarter that contains data nobody else has — even if your sample size is small. Journalists and bloggers cite primary sources because primary sources make their work easier.

Last year, one of our small-sample data pieces earned more high-authority backlinks in two months than a previous client’s six-month outreach campaign. The ratio wasn’t close.

9. Track three metrics monthly — and ignore the rest

The SEO industry has trained founders to obsess over dashboards. The truth is that three numbers tell you almost everything: how many of your targeted commercial keywords are ranking in positions 1 to 10, how much qualified organic traffic those rankings produce and how many of those visits assist a conversion.

Ranked positions tell you if your work is paying off. Traffic tells you if the rankings are valuable. Assisted conversions tell you if the traffic is worth the next month of investment. Everything else is noise until you’re operating at meaningful scale.

10. Audit, dedupe and prune every quarter

Most sites lose more SEO performance to keyword cannibalization, duplicate intent and stale content than they gain from new publishing. Every 90 days, audit your existing content: which pages are competing against each other for the same query, which are pulling impressions but no clicks and which are pulling neither?

Merge the cannibalizing pages. Refresh the impression-rich but click-poor pages with better titles and meta descriptions. Redirect the truly dead ones to their nearest healthy cousin. Pruning is unglamorous work that often produces the single biggest one-quarter SEO lift any site will ever see.

The system above isn’t proprietary. Every step is something a careful agency would do, in roughly the same order. What separates the founders who win at SEO from the ones who plateau isn’t access to a secret framework. It’s the discipline to do all 10 steps in sequence, on a quarterly cadence, for two to three years before judging the results.

By the time you do hire an agency, you’ll know exactly what to ask. The ones that can’t answer those questions will filter themselves out before they bill you.

When founders ask me how to start with SEO, they usually expect a tool recommendation. The honest answer is that the tool isn’t the problem. The sequence is.

For the past few years, my agency has been refining the same end-to-end SEO process — the one I now use with every new client and break down across an eight-module curriculum we run for our team. The steps work because they’re ordered. Most founders fail at SEO not because they skip steps but because they do them in the wrong order: writing content before researching keywords, building links before fixing crawl errors and chasing traffic before defining what kind of traffic moves their business.

Here’s the 10-step sequence I follow, in the order I follow it. You can run all of it yourself for the cost of two free tools and a few weekends.