Most people hear “competitor analysis” and reach for a $129/month Ahrefs subscription, then drown in Domain Rating charts they cannot act on. For a content site you have not launched yet, that is the wrong tool. You need three focused hours, a spreadsheet, and an honest read of who Google actually shows for your target queries — because in 2026 the question is no longer just “can I outrank them,” it is “will anyone click my result at all.”
TL;DR
Before you write article one, spend three hours doing this: pull the live SERP for 30 representative queries in an incognito window, tag every result by site type, and count how often an AI Overview sits at the top. If indie sites, forums, and blogs hold more than ~60% of the organic spots and AI Overviews are not on every query, the niche is winnable. If big publishers dominate or AI Overviews blanket the SERP, change the niche or the angle. Cost: three hours and zero paid tools. The alternative is finding out twelve months and 200 articles later.
Why the SERP, not the search volume, decides this
The strongest predictor of indie content-site success is not how much search volume a niche has. It is who already occupies page one. A query with 40,000 monthly searches that is wall-to-wall Forbes, Healthline, and Wikipedia is worse for you than a 2,000-search query where the top 10 is indie blogs and Reddit threads.
Two things changed this calculus in 2026, and both raise the bar:
- AI Overviews eat the clicks even when you rank. Ahrefs analyzed 300,000 keywords (150,000 with AI Overviews, 150,000 without) comparing December 2023 to December 2025 Search Console data, and found the presence of an AI Overview correlates with a 58% lower clickthrough rate for the position-one page. The drop is 50.8% at position 2 and tapers to 19.4% at position 10. AI Overviews sit at the very top of the page roughly 85% of the time, so they intercept the click before your result is even seen.
- Zero-click is now the default. Depending on the dataset, 58-65% of all Google searches end without any click to an external site, and for searches that trigger an AI Overview that figure rises to roughly 83%. You are not just competing with the other ten results; you are competing with the answer box.
The practical takeaway: a niche where Google rarely shows an AI Overview is worth far more to a solo builder than one where it shows on every query, even if the second niche has higher volume.
Are you ready to do this? Quick self-check
- You have a niche idea but cannot name the top five sites that rank in it.
- You think “competition” means total search volume rather than who occupies the SERP.
- You picked the niche because it pays well, not because the competitors are beatable.
- You have never read a future competitor’s best article end to end.
- You are about to buy a domain before doing any of this.
If two or more of those are true, the next three hours are the highest-leverage work you will do all year.
The free toolkit (skip the $129/month subscription)
You do not need a paid SEO suite for a pre-launch read. Everything below is free as of June 2026:
| Job | Free tool | What it tells you |
|---|---|---|
| See the live SERP | Incognito Chrome/Firefox window | The actual top 10 plus any AI Overview, without your search history skewing results |
| Quick authority glance | MozBar (free Chrome extension) | Domain Authority and Page Authority overlaid on each result — a directional, not exact, signal |
| Topic and seasonality | Google Trends | Whether interest is rising, flat, or seasonal |
| Reverse-engineer intent | The SERP itself | Forum results, video carousels, and “People also ask” reveal what Google thinks the searcher wants |
Use a clean incognito window per session so personalization and prior clicks do not reorder the results. If you want one paid spot check later, Semrush’s free tier lets you pull a competitor’s top organic pages a handful of times a day — but it is optional, and not needed for the go/no-go decision.
Step 1: Build the query list
Take your top 30 candidate queries — the long-tail questions from your niche research. Search each one in incognito and record the top 10 organic results. A spreadsheet column like this is all you need:
query,position,domain,site_type,result_format,aio_present
"firebase rewrites for spa",1,firebase.google.com,vendor-docs,html,yes
"firebase rewrites for spa",2,stackoverflow.com,forum,qa,yes
"firebase rewrites for spa",3,medium.com/@someone,personal,blog,yes
...Tag site_type as one of: vendor-docs, big-publisher, wiki, forum, indie-blog, youtube, news. Add an aio_present column (yes/no) for whether an AI Overview rendered. Those two columns decide the verdict; the rest is context.
Step 2: Classify the SERP
After 30 queries, tally the site_type distribution and the aio_present count. Three patterns emerge:
- Indie-dominated (more than 60% indie-blog, forum, or personal): green light. Consistent, specific writing can rank here.
- Mixed (30-60% indie): yellow light. Pick the sub-angles where indie sites still rank and avoid head-on collisions with the big publishers.
- Big-site dominated (under 30% indie): red light. Change the niche or pick a much narrower angle.
Then overlay AI Overviews. If an AI Overview appears on more than half your 30 queries, downgrade the verdict one level — even a green-light SERP loses most of its clicks when the answer box is always on top. One nuance worth knowing: being cited inside the AI Overview is not nothing. Sites quoted in the overview earn roughly 120% more clicks per impression than uncited sites on the same query, so a content site that consistently gets cited can claw some of that back. But you cannot count on that as a new site.
Step 3: Read the top three competitors’ best work
For your two or three strongest competitors, read five of their highest-ranked articles end to end — not skim, read. Note:
- Article length and depth (word count, number of H2 sections, presence of original screenshots or data).
- Section structure: is there a repeatable pattern you can match or beat?
- Which questions they answer well, and which they leave open.
- Whether the writing reflects first-hand experience or reads like generic AI output.
- Their internal linking: how many on-topic articles do they link, and how deep does the topic cluster go?
The page sitting at #1 today is the floor, not the target. If you ship something thinner, you will not win, and Google has gotten sharper at telling thin from substantive.
Step 4: Pick your angle
After classifying and reading, commit to one of these:
- Same topic, deeper execution. Works when competitors are clearly thin or AI-generated — add original screenshots, exact numbers, and a real opinion.
- Same topic, narrower audience. Write the same material for a more specific reader (for example “for solo founders” instead of “for engineers”).
- Adjacent topic, under-served. Find the sub-niche your competitors skipped or covered badly.
- Walk away. The honest call when you found 30 strong competitors, a constant AI Overview, and no gap.
Write the decision into a one-page positioning doc. You will reread it every time you draft, and it stops you from quietly drifting back to a topic you already proved is unwinnable.
Common mistakes
- Skimming instead of reading. Titles do not show you what “good” looks like; the body does.
- Trusting Domain Rating to judge competitors. DR and DA are directional and gameable. The live SERP position is the ground truth.
- Counting off-intent big-publisher pages. If Forbes ran a generic listicle that does not match the query intent, a specific indie page can still win that slot.
- Ignoring AI Overviews. Tracking only the ten blue links and skipping the answer box is the most expensive mistake of 2026.
- Picking the angle before the analysis. The angle should fall out of what the SERP shows, not your prior beliefs.
FAQ
Do I really need to read 15 competitor articles? Yes. Two hours of reading is the cheapest insurance against twelve months of writing the wrong thing at the wrong depth.
What if the SERP is all YouTube? That is a strong signal the intent is visual. Either compete in video or pick queries where text intent clearly dominates (how-to comparisons, error fixes, and definitions still skew textual).
How do I tell AI-generated competitor content from human writing? Look for generic intros, no specific numbers, no original screenshots, and no opinion. Machine-written content usually fails all four. Beating it is mostly a matter of specificity — exact figures, real steps, and a point of view.
Should I include Reddit and forum results? Yes. A Reddit thread ranking on a query means Google sees forum-style intent, and a structured, well-organized article can usually outrank a messy thread. Treat Quora and Medium the same way — reachable.
A free tool says Domain Rating 70. Does that mean I cannot win? Not on its own. DR measures the whole domain, not the page that ranks. Check whether the specific ranking URL is strong and on-intent; a weak page on a strong domain is beatable with a better page.
How often should I redo this? Re-run a 10-query spot check before each new content cluster, and re-check AI Overview coverage every few months — Google’s rollout and the CTR numbers move quarter to quarter.
Related
- How to Pick a Niche That Has Real Search Demand
- Content Site Monetization Paths
- How to judge if a topic has real search demand
- When to Buy a Domain
- Bilingual or Single Language: Which Should Your First Site Be
- When It Makes Sense to Add a Second Language
Tags: #Indie dev #Website planning #competitor-analysis #SEO