The challenge
Furniture is one of the hardest spaces in search. Terms like "cloud couch", "modular sectional" and "sectional sofa" are crowded with big retailers, marketplaces and affiliate review sites with years of authority. A new post aimed at those head terms has almost no chance.
The seed list
17 keywords around modular sofas, cloud couches, deep sectionals, L-shaped sectionals and affordable alternatives.
The volume rule
500 to 1,000 searches a month as a minimum. Anything below was out.
The real goal
High volume with minimal competition, the smallest group in any competitive niche.
The stakes
The list would guide the blog for the next year, so a weak pick costs months of writing.
Why tool numbers alone weren't enough
The brand's founder raised a sharp point early on: Reddit threads rank for a lot of furniture questions, but a Reddit result on its own doesn't mean a keyword is easy. What matters is the thread itself and what sits around it on page one.
That shaped the whole method. If the thread was thin and the rest of page one was small blogs or weak pages, that was a genuine opening. If it was busy and surrounded by major retailers, the keyword went to the bottom of the list or was dropped.
What I did
Mapped what competitors rank for
I pulled the ranking keywords for all three competitor blogs in Semrush to see which topics already earn traffic and which ones they covered badly or not at all.
Expanded beyond the seed list
The seeds set the direction, not the boundary. Many of the final topics came from questions buyers ask before choosing a sofa: comparisons, sizing, materials, care and room fit.
Filtered on demand first
Every topic was checked against the volume floor, then sorted by volume against difficulty.
Checked page one by hand
For each shortlisted keyword I looked at who actually ranks. That check, not the tool score, decided the final order.
Explained every pick
Each topic came with a short "why it's winnable" note. Borderline keywords that were genuinely easy wins were kept and clearly flagged, so the founder could decide.
The results
- 100 long-tail blog topics, each with search volume, difficulty, the page-one picture and a plain-English reason it can rank.
- A prioritized list, so writing can start with the best bets on day one.
- Honest flags on borderline keywords, instead of padding the sheet to hit the number.
- Delivered on the agreed due date and approved by the client.
An honest note: keyword research doesn't rank anything on its own. It decides where the writing effort goes, which is what turns a year of posts into traffic.
Lessons for ecommerce brands
Don't chase the head terms
The questions people ask while choosing a product are where a smaller brand can win.
Read page one yourself
A difficulty score can't tell you whether the ranking pages actually answer the question.
Treat Reddit carefully
A forum result is only an opening when the thread is thin and the pages around it are weak.
Write down the reason
Six months later, it tells your team why each article exists.