Keyword Prioritization Scorecard: A 100-Point Model You Can Copy

What is a keyword prioritization scorecard?

This page gives you one to copy: six criteria, their weights, a rule for what evidence each score needs, four action bands and a worked example. The weights are an editorial judgement, not a finding from data, and the page shows where every input comes from so you can change them.

The criteria and their weights come from the opportunity check that Kanso's research step ran on this topic before the page was written. That check asks whether a topic is worth writing. This page adapts it to ask whether a keyword is worth targeting next.

• Score each keyword on six criteria: relevance (25 points), demand (20), ability to beat the current results (20), no overlap with existing pages (15), a way to add something original (10) and confidence in the evidence (10).
• Attach an evidence label to every score: measured, documented, observed, reported or judgement. The label caps how many points the score can earn.
• Read the total against four bands: 75 to 100 schedule, 55 to 74 backlog, 40 to 54 only with an original element, below 40 park.
• Google Ads uses the words keyword prioritization for an auction rule. It is a different thing, and the next two sections separate the two.
• No ranking data validates these weights. Treat them as a starting point and change them as your own results arrive.

Is keyword prioritization an SEO decision or a Google Ads rule?

It is both: in SEO it is the decision about which keywords to target first, and in Google Ads it is the rule Google uses to pick which of your eligible keywords enters an auction. The two share a name and nothing else.

Search results for the phrase mix the two. Going by their titles, five of the nine organic results the research reviewed on 29 October 2026 are about Google Ads, including Google's own help page in first place.

This page reads three of those results, Google Ads Help, a WordStream article on the Ads rule and a BrightEdge page, plus a fourth Google page on the same rule. Only the BrightEdge page is about choosing keywords for organic search, and it is a vendor's solutions page that describes a process in its own product terms.

If you came here to choose which keywords to write about, skip to the criteria below. If you manage Google Ads campaigns and wondered why a keyword did not serve, the next section is the one for you.

How does Google Ads decide which keyword wins?

Google Ads works down four levels in order: an identical exact match keyword, an identical phrase or broad keyword or search theme, AI-based relevance, and then the highest Ad Rank. Google Ads Help, read on 29 October 2026, says that keywords which are eligible for the same search term do not compete with each other in the auction, and only one of them can trigger an ad for that term.

Google defines identical to include spell-corrected search terms and to exclude plurals and synonyms. The AI step judges relevance from the meaning of the search term, all the keywords in an ad group and the landing pages in that group. The Accelerate with Google India page describes the same system as choosing the right keyword, bid, budget, creative and landing page when several overlapping keywords are eligible.

Matching and prioritization are separate steps. WordStream's Jyll Saskin Gales, in an article last updated on November 22, 2024, describes matching as the system that decides whether you are eligible to show an ad, and prioritization as the system that decides which eligible ad is put forward. She also marks where her article moves from Google's facts to her own opinions: advertisers have less control than they think, Dynamic Search Ads sit low in the order, and Ad Rank still matters.

For a keyword list, the practical advice in Google Ads Help is to structure ad groups by theme, so it is clear which group should match a search. The same page lists reasons an expected keyword may not serve: low search volume status, disapproved creatives or landing pages, targeting differences, budget limits, a different ad format such as Shopping, and advanced search experiences such as Lens, AI Mode and AI Overviews.

What are the six criteria and their weights?

The scorecard has six criteria that add up to 100 points, and each one asks a question you can answer from evidence rather than from a feeling. The table shows the weights, the question, what full marks look like and the evidence label that usually applies.

total = relevance + demand + ability to win + no overlap + originality + confidence, out of 100

Ability to beat the current results should include a look at the results page itself. BrightEdge states that listings above the fold get 69% more clicks than those below it, and that some keywords have no organic listing above the fold because other content types fill that space. That is a vendor statistic with no method on the page, so it carries the Reported label and can lift a score only to the good band, but opening the results page and checking how much room organic results get is a quick check.

Overlap works differently from the other five. If a page on your own site already serves the same intent, the answer is usually to update that page, so the keyword leaves the queue instead of earning a low score.

Why these weights, and how far should you trust them?

The weights are a judgement, not a measurement: nothing in the sources behind this page tests whether 25 points for relevance predicts better outcomes than 20. They encode an order of importance that you can check against your own results.

Relevance is highest because a visit for a keyword unrelated to what you do has no use. BrightEdge frames the same priority as focusing on high-value keywords and prioritizing by organic return on investment, though its page gives no formula. Demand and ability to win share 20 points each because a keyword nobody searches and a keyword you cannot realistically win are equally poor targets.

Overlap gets 15 because it is close to a yes-or-no question. Originality and confidence get 10 each because they adjust a decision more than they make it.

To change the weights, keep the total at 100, change one weight at a time and write down the date and the reason. If you have results from earlier articles, such as clicks and conversions after a few months, let those results move the weights. BrightEdge makes a similar point about iteration: monitor impressions, clicks, time on page, conversions and revenue for new content, then do more with the keywords that worked.

What are evidence labels and why do they cap a score?

An evidence label states where a score's input came from, and the label limits how many points that input can earn, so a guess cannot score as high as a measurement. This is the part of the scorecard most keyword sheets leave out.

Write the label in the sheet next to every score. When one score rests on several inputs, label it by the weakest input that carries it, so a score built on one tool export and one hunch gets the Judgement cap.

The caps do two jobs. They stop a high total from hiding thin evidence, and they show you which cheap piece of data would raise the score the most. A relevance score of 18 on judgement alone is not allowed; either find the product page or customer question that makes it Observed, or accept a score of 12 or less.

How do points, bands and actions work?

Each criterion is scored in four bands of its own weight (weak, partial, good and strong), and the total out of 100 is then read against four action bands. The first table gives the point range for each band.

The action bands decide what happens to a keyword. Only one of the four edges has any support outside our own judgement: the research step described a total of 50 as moderate, and said such a topic is worth writing only with a clear original element.

Expect to move the edges. After a few months, look at which keywords in each band produced pages that earned impressions and clicks, and shift the edges to match what you see.

What does a scored keyword look like? A worked example

Scoring this page's own primary keyword, keyword prioritization, gave 50 out of 100. The research step produced the points and the inputs below on 29 October 2026; the labels are our addition, because the research step does not assign them.

12 + 6 + 12 + 15 + 2 + 3 = 50

The weakest score is originality, 2 of 10, because the research found a single primary source. That is the useful result: the sheet shows which gap to close before writing, and for this topic the gap was an original element, which is why this page publishes open weights and labels.

Demand is a proxy here. The keyword tool returned no volume figure for the primary keyword, so the 6 of 20 rests on four community threads, which is why a Search Console check is the next step once a page is live. The research step states the same limit: it is a decision aid built from free signals, not a Google score or a traffic forecast.

One more caution. We did not see the research step's internal rules for converting inputs to points, so the band guide above is our reading. Read against it, the scores are partial for relevance, demand and confidence, good for ability to win, strong for overlap and weak for originality.

How do you build and run the sheet?

Build it with one row per keyword and one column per criterion, then work through seven steps. The layout below is the whole template, filled in with the example from the last section.

In a spreadsheet, the total for row 2 is =SUM(B2:G2). Sort the sheet by the total column, largest first, and keep the date next to every row.

1. Collect candidates and note where each came from. Search Console queries are the cheapest source, and the guide to striking-distance queries in Search Console shows how to find the ones already close to page one. For a first look at the data, reading Google Search Console as a founder covers the reports.
2. Check your own site first. If a page already serves the intent, update it and drop the keyword from the queue.
3. Gather evidence for each criterion and write down the tool and the date beside every figure.
4. Assign a label to every score using the table above, taking the weakest input.
5. Score each criterion inside the band its label allows, then add the six scores.
6. Read the total against the action bands. For anything that reaches the schedule band, the next document is a brief, and the content brief template with a filled example is built for that step.
7. After publishing, record impressions, clicks and conversions against the row, and re-score and re-weight when you have results.

How does the sheet handle a low-authority site or a small market?

It handles both through the criteria rather than special rules: a new site feels its position in the Ability to beat the current results score, and a small market feels it in the Evidence of demand score. Two questions recur in SEO communities here, and the sources behind this page answer one of them only in part.

Keyword difficulty or topical authority?

Whether a site with little authority should choose by keyword difficulty or by topical authority is an open question in the research, and no source in it settles it. Google's documents in the research do not define either term, and the community threads are asking, not answering.

The sheet does not pick a side. Record the difficulty figure as one labelled input to the Ability to win score, with the tool name and date, because each tool calculates difficulty its own way. Score your existing pages on the topic under Relevance and Overlap.

One option, which is our judgement and untested here: keep the first queue mixed, so that high-difficulty keywords do not take every slot while the site has no pages on the topic.

Small markets with little or no volume data

In a small market the volume column is often empty, so demand has to come from other inputs. The worked example is one case: the keyword tool returned no volume for the primary keyword, and the demand score came from four recent community threads.

Good inputs are Search Console impressions for related queries (Measured), the questions customers send you (Observed) and recent threads where people ask the question (Observed). Label them as proxies, score them within their cap, and replace them with Search Console impressions as soon as a page is live.

What is documented and what is analysis?

The Google Ads rules, the 69% figure and the opinions in WordStream's article are sourced, while the scorecard's weights, caps and bands are our own design. Facts are as of 29 October 2026.

• Documented: Google Ads Help gives the four-level priority order and the list of exceptions.
• Documented: keywords eligible for the same search term do not compete in the auction, and only one triggers an ad (Google Ads Help).
• Reported: listings above the fold get 69% more clicks than those below it (BrightEdge, no method stated).
• Reported: the three implications of the Ads rule are Jyll Saskin Gales's opinion, labelled as such in her WordStream article.
• Documented in the research package: the opportunity check scored this topic 50 out of 100, with the six inputs shown above.

• Analysis: the six weights are a judgement and have not been tested against ranking results.
• Analysis: the evidence labels, their caps and the four band ranges per criterion are our design.
• Analysis: the action bands, apart from the research step's reading of a 50, are our judgement.
• Analysis: the advice on low-authority sites and small markets is a way to use the sheet, not a finding.
• Analysis: the labels in the worked example were added by us.

Where does Kanso fit into keyword prioritization?

Kanso does not score keywords for you: you do the scoring yourself, with this sheet. Two of its agents sit on either side of it.

The SEO Agent audits your site and Search Console data and turns what it finds into recommendations and tracked tasks for your approval, which can feed the demand and overlap columns. The Writer Agent turns Search Console opportunities into long-form drafts in your brand voice, for you to review and publish yourself. Neither promises rankings, and neither publishes to a CMS.

Keyword prioritization: frequently asked questions

Sources

• About ad group and asset group prioritization within a Google Ads account, Google Ads Help
• Broad Match for AI-based keyword prioritisation: Google Ads Tutorials, Accelerate with Google India
• How Keyword Prioritization Works in Google Ads (+What It Means), WordStream, by Jyll Saskin Gales, last updated November 22, 2024
• Keyword Research and Prioritization, BrightEdge