See exactly what Google rewards, for any keyword.

Most keyword lists die at the briefing stage, because the keyword looked fine in a spreadsheet and nobody looked at the actual search results before assigning it. SERP analysis is the ten-minute check that prevents that. It answers two questions: what does Google think this query means, and can we realistically compete?

Read intent from the result mix

The results Google already shows are the clearest statement you will get about what the query means. Count the formats in the top ten:

• Mostly listicles and “best X” roundups: commercial investigation. The searcher is comparing options.
• Mostly product or category pages: transactional. A blog post will not rank here regardless of quality.
• Mostly guides, definitions, and explainers: informational.
• A visible mix: the intent is genuinely split, and you can often win by covering both angles on one page.

The practical rule is that you do not get to argue with this. If nine of the ten results are comparison listicles and you brief a how-to guide, you have chosen to rank for something Google is not asking for.

Check who ranks, not just what ranks

The second pass is about authority. Note the domains. If the top ten is Wikipedia, a couple of major publishers, and two entrenched category leaders, you are looking at a fortress SERP. Better content does not beat that on its own, and a site with little accumulated authority will not break in.

This is where a lot of content budget gets burned: the page gets written well, it is genuinely useful, and it lands on page seven because the query was never winnable at that authority level. Recognising a fortress early is worth more than writing a better article.

When you hit one, the move is to go more specific rather than to try harder. A narrower long-tail variant with a weaker top ten is a page you can actually win.

Record the SERP features

Note what sits above and around the organic results:

Feature	What it changes
AI Overview	Pushes organic down, absorbs clicks, opens a citation opportunity
Featured snippet	Winnable with a direct 40 to 55 word answer near the top
People Also Ask	Gives you exact-match FAQ questions to answer
Video pack	Signals the intent may be better served by video than text
Shopping results	Strong transactional signal, deprioritise editorial content

People Also Ask is the most immediately useful of these. The questions in it are real queries phrased the way people actually type them, which makes them the right wording for an FAQ section.

The template

Record one row per keyword before anything gets briefed:

Keyword | Intent | Dominant format | Top 3 domains | Fortress? | SERP features | Decision | Reason

The Reason column is the one people skip and later regret. A keyword skipped with a written reason is a decision you can revisit when your authority changes. A keyword skipped silently gets rediscovered and re-argued every quarter.

Doing this across a client roster

One SERP check is ten minutes. Doing it for every keyword across a dozen client brands, then keeping the results current as SERPs shift, is where it stops being feasible by hand. That is the part AltoRank automates: SERP checks per client, briefs written against what the SERP actually rewards, and a human approving before anything publishes. See how it compares on the alternatives comparisons, or start with keyword clustering if you have not grouped your list yet.