Six months ago, one of our clients was publishing consistently and going nowhere. Two posts a week, decent grammar, on-topic keywords, and organic traffic that refused to move past a few hundred visits a month. Nine months later, that same blog was pulling in over 400% more organic traffic, with the growth compounding month over month rather than plateauing. Nothing about the publishing cadence changed. Almost everything about the process behind each post did.
This is the exact process we used, broken down into the four things that actually moved the number: where we found topics worth writing about, how we validated demand before writing a single word, what "high quality" meant in practice rather than in theory, and why differentiation - not word count - was the real lever.
Start Where the Questions Already Are
Most content teams start with a keyword tool. We start with Reddit. Keyword tools tell you what people type into Google. Reddit threads tell you what people actually mean when they type it - the confusion behind the query, the objections they haven't voiced yet, the exact words a real buyer uses instead of the sanitized phrasing that shows up in search-volume reports.
The process is simple and repeatable. Search Reddit directly for your topic area, then read every thread with more than a handful of comments. Not the post itself - the comments. That's where the real questions live: the follow-up someone asks because the original answer didn't fully satisfy them, the "but what about..." replies, the disagreements between commenters that reveal a topic has more nuance than any existing article captures. Every one of those threads is a content brief someone else wrote for you, for free.
We also run site:reddit.com "[topic]" directly in Google rather than relying on Reddit's native search, since Google indexes and surfaces relevant threads more reliably. Pair that with site:quora.com and the "People also ask" boxes on the SERP itself, and you have a list of real questions in real language - not the flattened, generic phrasing that keyword tools tend to output.
Check Who's Already Winning - and Why They're Vulnerable
Before writing anything, we run a competitive gap pass using site: search directly against competitor domains. Searching site:competitor.com [topic] shows you exactly how a competing site has structured their coverage of a subject: what they've published, what angle they took, and - just as importantly - what they left out.
This step matters because most content teams skip straight to writing without checking what already exists. The result is a fifth article saying the same thing as the four before it, none of which rank particularly well because none of them differ enough to be worth Google - or an AI engine - preferring over the others. Reading the top five ranking pages for a target query before writing isn't optional. It's the only way to know whether you're about to add something new to the conversation or just add to the noise.
Keyword Research Is the Floor, Not the Strategy
We still use keyword tools. Search volume, keyword difficulty, and related-term clustering are useful inputs. But they're the floor of the process, not the strategy. A keyword tells you that people are searching for something. It doesn't tell you what they actually want when they get there, and it definitely doesn't tell you how to write something worth ranking.
The keyword research step exists to confirm a topic has demand and to surface the related terms and subtopics worth covering in the same piece, so the article can build topical depth rather than just hitting one narrow phrase. It's a checklist item, not a creative direction.
What "High Quality" Actually Means
Every content brief says "high quality." Almost none define it in a way that changes what gets written. Here's what it meant in practice for this client, and what actually moved the traffic number:
- User-first structure. The answer to the reader's actual question appears in the first two paragraphs, not after 400 words of preamble. If someone came from a Reddit thread asking a specific question, the article answers that specific question immediately, then goes deeper.
- Specificity over generality. "Improve your site speed" is generic advice that a hundred other articles already say. "Compress hero images to WebP and lazy-load anything below the fold - this alone cut our test site's LCP from 4.1s to 1.8s" is specific, verifiable, and not interchangeable with any other article on the topic.
- Original angles, not aggregated ones. A huge share of underperforming content is aggregation - restating what five other articles already said, in a slightly different order. We required every post to include at least one angle, framework, or piece of original analysis that wasn't already sitting on page one.
- No filler. No "in today's fast-paced digital landscape." No throat-clearing intros. No padding a 600-word answer out to 2,000 words because a content template said so. Length followed substance, not the other way around.
Differentiation Is the Real Growth Lever
If there's one finding from this project we'd want another marketer to walk away with, it's this: the biggest driver of the traffic increase wasn't publishing frequency, word count, or even backlinks. It was differentiation. Every single post that meaningfully outperformed its category average had a clear, identifiable reason to exist that wasn't already covered by the top-ranking pages for that query.
Sometimes that meant a genuinely contrarian take, backed by evidence, on a topic where the consensus advice was outdated. Sometimes it meant original data from the client's own product or customer base that no competitor could replicate. Sometimes it was as simple as answering the follow-up question that every other article left unaddressed - the one buried in the Reddit thread comments. The common thread was that every post had to pass a simple test before it got written: if this exact article already existed somewhere on page one, would we still publish it? If the honest answer was no, the topic got reworked until the answer was yes.
The Repeatable Version of This Process
Distilled down, the process that produced the 400% increase looks like this: mine Reddit and Quora threads for real questions in real language, run site: searches against ranking competitors to see what's already covered and what's missing, use keyword tools to confirm demand and surface related subtopics, write with a user-first structure that answers the actual question immediately, and refuse to publish anything that doesn't clear the differentiation bar. None of these steps are exotic. What made the difference was doing all of them, consistently, on every single post - not just the ones that felt important.