When a PPC account produces more keyword ideas than you can test, a consistent opportunity score helps you decide what to add, pause, or promote. This guide shows how to rank terms using intent, expected conversion performance, cost, impression share, competition, and profitability, with a repeatable spreadsheet method you can revisit as your data changes.
Overview
PPC keyword prioritization is not a contest to find the largest search volume or the lowest cost per click. The best term is the one that connects a relevant searcher with an economically viable conversion opportunity. A keyword with modest demand may deserve attention if it signals strong buying intent. A high-volume term may belong lower on the list if it attracts research traffic, has expensive clicks, or produces low-value conversions.
Use an opportunity score to make this judgment visible. The score does not replace campaign data, search term analysis, or testing. It creates a common framework for comparing candidates that have different levels of evidence.
A practical scoring model combines five dimensions:
- Intent: How closely does the query indicate a need your offer can satisfy?
- Conversion potential: How likely is a qualified click to complete the target action?
- Economics: Can the expected revenue or profit support the estimated cost?
- Reach: Is there enough eligible demand or impression opportunity to justify a test?
- Manageability: Can you give the term a relevant landing page, ad message, budget, and measurement setup?
Before scoring, confirm that conversions are recorded consistently. The PPC conversion tracking and UTM guide can help you check attribution inputs before using them in a profitability analysis.
How to estimate
Start with a candidate list from keyword research, search term analysis, competitor or product research, and existing account data. Keep close variants together only when they share the same intent and can use the same landing page. Otherwise, separate them before scoring. A keyword clustering workflow is useful at this stage.
Then score each candidate from 1 to 5 in the following fields:
| Field | 1 means | 5 means |
|---|---|---|
| Intent | Broad, informational, or weakly related | Specific need with clear commercial relevance |
| Conversion potential | Little supporting evidence | Strong historical or closely comparable performance |
| Profitability | Likely below your target economics | Comfortably within your target economics |
| Reach | Very limited eligible impressions | Enough demand for a meaningful test |
| Execution fit | No suitable ad or landing page | Strong message and destination already available |
Apply weights that reflect your business. For example:
Opportunity score = (Intent × 30) + (Conversion potential × 30) + (Profitability × 25) + (Reach × 10) + (Execution fit × 5)
Because each rating is from 1 to 5, the result ranges from 100 to 500. The weights are assumptions, not universal benchmarks. A lead-generation business may weight lead quality and profitability more heavily, while a new account with little conversion history may give reach and intent more emphasis for initial testing.
For candidates with enough data, estimate profitability separately:
Expected conversion rate = conversions ÷ clicks
Expected cost per conversion = average CPC ÷ expected conversion rate
Expected contribution per click = expected conversion rate × contribution per conversion
Here, contribution per conversion should reflect the value you are willing to use for acquisition decisions, not necessarily total revenue. Account for margin, refunds, sales costs, or lead close rates where those inputs are available. If the result is uncertain, use a conservative range rather than a single precise number.
Inputs and assumptions
Build a spreadsheet with one row per keyword or tightly related keyword group. Use these columns:
- Keyword or search-term cluster
- Platform and campaign
- Intent rating, 1–5
- Historical or estimated conversion rate
- Conversion-potential rating, 1–5
- Average CPC or CPC range
- Conversion value or contribution per conversion
- Profitability rating, 1–5
- Impressions, impression share, or demand estimate
- Reach rating, 1–5
- Execution-fit rating, 1–5
- Weighted opportunity score
- Recommended action
- Test date and review date
Do not treat platform estimates as guaranteed outcomes. Forecasts can change with bids, budget, geography, match type, audience settings, seasonality, and competition. Label every input as historical, estimated, or assumed. If a keyword has no reliable data, lower the confidence rating rather than presenting an estimate as fact.
Also separate keyword opportunity from keyword eligibility. A potentially valuable term may not serve because of low volume, targeting settings, budget limits, or account structure. For serving issues, use the keyword status troubleshooting guide before concluding that the opportunity is poor.
Match types should be recorded in the sheet because the same wording can behave differently under different targeting controls. Review the actual queries that triggered ads, not just the keyword you entered. This is where search term analysis reveals irrelevant themes and new high-intent language. Add confirmed exclusions to a maintained negative keyword list rather than repeatedly judging the same waste.
Worked examples
Assume the following three illustrative candidates are for the same product and are being considered for a controlled search campaign. The figures are examples for demonstrating the method, not benchmarks.
| Candidate | Intent | Conversion | Profitability | Reach | Fit | Score |
|---|---|---|---|---|---|---|
| Specific product solution | 5 | 4 | 4 | 3 | 5 | 425 |
| Category comparison | 4 | 3 | 3 | 5 | 4 | 360 |
| Broad problem topic | 2 | 2 | 2 | 5 | 3 | 245 |
The specific product solution ranks first because its strong intent and execution fit outweigh its smaller reach. Launch it as the primary test, with an ad and landing page that directly address the stated need. The comparison term is a reasonable second test, but it may require broader messaging and closer monitoring of conversion quality. The broad topic should not automatically be rejected; it can be tested later with a capped budget, or used for content and audience discovery if direct-response economics are weak.
Suppose the first candidate has an illustrative average CPC of 4 currency units and an expected conversion rate of 5%. Its estimated cost per conversion is 4 ÷ 0.05, or 80 currency units. If the acceptable contribution per conversion is 140 units, the term may have room for testing. If the expected conversion rate falls to 2%, the estimated cost per conversion rises to 200 units, changing the recommendation. That sensitivity is why a range is more useful than a single forecast.
Use the score to choose the next action, not to make a permanent judgment:
- High score: Add, promote, or expand carefully with a defined test budget.
- Middle score: Test only after improving the landing page, ad relevance, targeting, or measurement.
- Low score: Pause, exclude, or hold for a different objective unless new evidence changes the assumptions.
When to recalculate
Recalculate scores during weekly optimization sessions when meaningful new search terms, costs, or conversions appear. A weekly review should usually focus on actions: promote proven terms, investigate deteriorating economics, and add negatives for irrelevant traffic. Use a monthly review for broader ranking changes across campaigns and products.
Rescore candidates after a landing-page change, offer change, tracking correction, budget shift, geographic expansion, or major change in conversion quality. Revisit assumptions before a product launch and during predictable seasonal periods; demand and competition can alter both reach and cost. If pricing, margin, or sales close rates change, update profitability even if advertising metrics look stable.
Keep the previous score, new score, date, and reason for the change. This simple history prevents unexplained decisions and shows whether an improvement came from better demand, better execution, or a changed assumption. For a repeatable monthly process, pair the sheet with a PPC keyword reporting template.
Finish each review by selecting a short action list: the top terms to expand, the terms to test next, the terms to pause, and the negative themes to add. Opportunity scoring works best when it narrows the next decision without pretending to predict the entire campaign. Update the inputs, record the evidence, and let the next controlled test improve the score.