Choosing Keywords for a Site With No Authority

I run a handful of sites, and every single one of them started at zero. No backlinks, no history, no trust. The keyword research I do for a brand new domain looks nothing like the research I do for a site that’s been indexed for three years and has a few hundred referring domains. If you use the same playbook for both, you’ll waste months chasing terms you were never going to win in the first place.

This isn’t a theory post. It’s the filter I actually run new keywords through before I let a writer touch them.

Why authority changes which keywords are winnable

Search engines don’t rank pages in a vacuum. They rank pages against every other page trying to answer the same query, and a big chunk of that comparison comes down to how much the domain and the page have earned trust through links, engagement, and time. A brand new site has none of that. It’s that the competing pages for most decent keywords already have years of accumulated signals a new page can’t match on day one.

That means the keyword difficulty score you see in Ahrefs, Semrush, or whatever tool you use is really telling you two different things mashed into one number: how hard the term is for an established site, and how hard it is for a site with nothing behind it. Those are not the same difficulty. A term that reads as “medium difficulty” for a five-year-old site with real links can be effectively closed off to you for a long time.

What “no authority” actually means in practice

I don’t treat authority as some mystical score. In practice it’s a proxy for a few concrete things: how many other sites have linked to yours, how consistently the site has published without long gaps, and whether the site has any track record of pages actually holding rankings once they get them. A brand new domain fails all three. You have no links, no publishing history, and no pages that have proven they can hold a spot.

That’s the honest starting condition. Once I frame it that way, keyword selection stops being about “what has the best search volume” and becomes “what can a page with none of those three things realistically compete for.”

Start with the difficulty score, then throw half of it out

I still pull keyword difficulty numbers when I’m building a list. They’re useful as a first pass filter to cut obviously unwinnable terms. But I don’t trust the low end of the scale either, because difficulty tools estimate based on the backlink profile of ranking pages, and they miss things like brand recognition, on-site content depth across a whole domain, and how much a topic rewards being an established name versus a good page.

So the tool gets me a shortlist, not a shopping list. From there I open the actual search results by hand for every term I’m considering and look at what’s ranking. That’s the step people skip because it’s slow, and it’s the step that actually matters.

Look at who’s ranking, not just the metric

When I check the results page for a candidate keyword, I’m looking for a specific pattern: are the top results dominated by big established sites, forums, and government or edu pages, or is there a mix that includes smaller independent sites, some of them clearly running on a shoestring? If every result in the top ten is a site I recognize as a major player in the niche, that keyword is not for a new site right now, no matter what the difficulty tool says. If I see a couple of small operators mixed in, especially ones with thin-looking “about” pages and no obvious brand presence, that tells me the term is winnable by a site that’s still building.

I also check how old the ranking pages look. A results page full of content published or meaningfully updated in the last year is a different fight than one where the top pages have clearly sat untouched for a long time. Fresh, thin competition is one of the few honest windows a new site has.

The volume trap

The instinct with a new site is to go after decent volume because you want traffic to show up. I get it, I’ve done it, and I’ve wasted content budget on it. A 2,000 search a month keyword that you have no realistic path to page one for is worth less than a 40 search a month keyword you can actually hold. Traffic on a page you can’t rank is zero traffic. It doesn’t matter how big the number looked in the research tool.

I’d rather build a base of pages ranking for smaller terms first. Those pages start earning the site its first real signals: time on page, return visits, occasional links from people who found something useful and specific. That base is what eventually makes the bigger terms winnable. Skipping straight to the big terms just burns content on pages that sit on page four indefinitely.

Long-tail isn’t a strategy, it’s a starting point

“Go long-tail” gets thrown around as if it’s a complete plan. It’s not, it’s just the mechanism. The reason long-tail works for low authority sites is that longer, more specific queries usually have fewer strong pages competing for them, because most sites don’t bother writing content that specific. That’s the actual mechanic, not some special long-tail magic.

What I look for isn’t length, it’s specificity that matches something my site can genuinely answer better than a generic page. A three-word term isn’t automatically better than a two-word term. A five-word term that nobody has written a real answer for is better than either, if there’s any search volume at all and the intent is clear.

Informational versus transactional at zero authority

Transactional and commercial terms almost always sit closer to money, which means more competitors with more resources are fighting for them. At zero authority, those are usually the last terms I go after, not the first. Informational terms, especially ones tied to a specific problem or a specific edge case, tend to have less commercial pressure behind them and more room for a new page to slot in.

I still map out the commercial terms early, because I want to know where the site is eventually headed. I just don’t send a brand new domain’s first content budget at them. The commercial terms become the target once the site has a run of ranking informational pages backing it up.

Building a keyword ladder

The way I actually structure this is a ladder, not a list. The bottom rungs are narrow, specific, lower volume terms with visibly weak or thin competition. The middle rungs are moderate volume terms where I’ve checked the results page and seen at least a few smaller sites holding spots. The top rungs are the terms I actually want long term, the ones with real volume and real competition, that I’m not touching until the site has a track record.

Every keyword I add to the plan gets slotted onto a rung based on what I saw when I manually checked the results, not just the number a tool spat out. It’s more work per keyword. It’s also the difference between a content plan that produces pages that rank within a reasonable window and one that produces pages that sit invisible for a year.

What I stopped doing

I stopped trusting difficulty scores as a final answer. I stopped picking keywords off volume alone. And I stopped assuming that because a term is technically “low competition” in a tool, it’s actually low competition for a site with zero history. None of those habits cost me nothing to break, they cost me content and time before I noticed the pattern.

What I do now is slower and less exciting, but it produces pages that actually show up. Small, specific, honestly-assessed terms first. Bigger terms once the site has earned something to compete with.

If you want more of this kind of breakdown, on keyword research, link building, and the technical side of SEO, from someone running the sites and eating the mistakes, check out the rest of what we’ve put together here.