Long tail vs short tail keywords: which should you target first?

Short tail keywords are broad searches of one to three words with big volume and brutal competition: "running shoes", "crm software". Long tail keywords are longer, more specific phrasings with small individual volume and much clearer intent: "best crm software for a two person real estate team". The practical difference is not length, it is specificity, and the practical answer for most sites is that the long tail comes first. It converts better, ranks faster, and there is far more of it: almost 95% of all keywords get 10 or fewer searches a month by Ahrefs' count of their US database, and Backlinko's study of 306 million keywords found that phrasings of four or more words make up the overwhelming majority of what people actually type.

What short tail keywords really are

A short tail keyword, also called a head term, is the plainest name for a topic. "Keyword research." "Standing desk." "Payroll software." The volume numbers are the biggest you will see in any tool, which is exactly why every site in the niche wants them and why the first page for each is a fortress of incumbents. But the volume is not even the real problem. The real problem is ambiguity: someone searching "standing desk" might want to buy one, compare brands, read about back pain, or find a DIY plan. Google resolves that ambiguity by ranking a mixed bag, usually led by giant marketplaces and publishers who can satisfy several intents at once.

That ambiguity has a second cost people rarely price in: even when a mid-size site does crack a head term, the traffic behaves like a crowd wandering through a lobby. Lots of visits, vague intentions, thin conversion. Head terms are branding assets. They are rarely revenue assets on their own.

What long tail keywords really are

The long tail is everything specific: questions, qualifiers, comparisons, use cases. "Payroll software for restaurants with tipped employees" is a long tail keyword not because it is six words, but because it names one situation precisely. The person typing it has already done their thinking. They do not need to be persuaded that they have a problem; they are checking whether you are the answer to it.

Individually these terms look like rounding errors, ten searches here, forty there. In aggregate they are most of search. That distribution is the entire strategic point: the demand you can win this quarter is spread across hundreds of small phrasings, not concentrated in the one big term your dashboard keeps showing you. If you want the full method for surfacing them, we wrote a separate guide on how to find long tail keywords, and the long tail keyword tool automates the mining and the clustering in one pass.

Long tail vs short tail keywords: the actual trade-offs

Put side by side, the trade balances like this:

• Volume per keyword: short tail wins by an order of magnitude or two. This is the only column it wins.
• Total available volume: long tail wins. The many-small-numbers add up to more than the few big ones, and you can plausibly win the small ones.
• Difficulty: long tail wins decisively. Fewer sites target each phrasing, and the ranking pages are usually weaker.
• Conversion: long tail wins, and it is not close. Specific queries come from people later in their decision, so the same hundred visitors are worth several times more.
• Time to rank: long tail terms can move in weeks on a modest site. Head terms take years, if they move at all.
• Content cost: a wash, with a catch. Long tail needs more pages, but each page is smaller and better defined. Head terms need fewer, heavier, endlessly-updated pages.

There is one more asymmetry worth naming. When a long tail page misses, you have lost one small page's effort and learned something about the niche. When a head term campaign misses, you have usually lost a year of content, links and attention. The long tail is not just higher expected value; it is cheaper variance.

The order of operations: long tail first, head terms as a consequence

The clean way to think about it: you do not choose between long tail and short tail, you choose the order, and the order is almost always tail first. Every long tail page you rank builds topical coverage around the head term, earns internal links pointing at your eventual pillar page, and generates the engagement signals that make the head term reachable later. Sites that rank for "crm software" today mostly got there by first ranking for two hundred phrasings underneath it. The head term was the consequence of the campaign, not the target of it.

Head terms also lean on authority you may not have yet. The first page for a serious short tail keyword is a list of domains with deep link profiles built over years. You can close that gap two ways: slowly, by publishing pages other sites choose to cite, and deliberately, by placing genuinely relevant editorial backlinks on real publications that point at the pages you need to strengthen. Either way, the tail-first plan gives those links something worth landing on.

When does short tail come first? Roughly three cases: you already have the authority (an established domain entering an adjacent topic), the niche is so new that the head term is still soft, or the term is your own category name and you need to own it for credibility even at break-even economics. If none of those describe you, they are not your fight yet.

How to split your keyword list in practice

Take your keyword universe and sort it into three buckets, ignoring word count entirely and reading specificity instead:

• Winnable and specific: clear single intent, modest difficulty. These become pages now, one page per cluster of phrasings that mean the same thing.
• Valuable but heavy: the head terms and near-head terms. These become pillar pages you build once and improve quarterly, fed by internal links from bucket one.
• Big and vague: high volume, mixed intent, no plausible page format. These are dashboard decoration. Let them go.

The step people skip is the clustering in bucket one. Twenty long tail phrasings are usually five pages, not twenty; "payroll for restaurants", "restaurant payroll software" and "payroll software for restaurant owners" are one page wearing three hats. Split them and the pages compete with each other, which is how a long tail strategy quietly turns into a keyword cannibalization problem. Group first, then write. The full SEO keyword research pass in Keywordpilot does the grouping automatically and hands back the buckets in priority order.

What the numbers say when you get it right

A tail-first plan produces a distinctive traffic curve: nothing for a few weeks, then dozens of small streams that individually look like noise and collectively become the majority of organic revenue. It rarely produces a single hero keyword you can screenshot, which is why it is underused; the strategy photographs badly and banks well. The head-first plan produces the opposite: a beautiful target, years of nothing, and a postmortem. Choose the boring curve.