Google Ads N-Gram Analysis: Find Wasted Spend in Search Terms

What is N-gram analysis?

An N-gram is simply a sequence of words extracted from a search query.

Take this Google Ads search term:

A one-word analysis might identify:

• cheap
• running
• shoes
• men

A two-word analysis could identify:

• cheap running
• running shoes
• shoes men

Instead of evaluating that query in isolation, N-gram analysis combines performance data from every search term containing the same word or phrase.

That lets you answer questions such as:

• Which words repeatedly consume budget without generating conversions?
• Which phrases appear frequently in high-value searches?
• Are certain themes producing unusually strong CPA or ROAS?
• Which query patterns account for a significant share of total search-term spend?

Why use N-gram analysis in Google Ads?

A normal Search Terms report is organized around complete queries. The problem is that poor or strong performance often comes from a pattern shared by many different queries.

Imagine these search terms:

Individually, none may look disastrous. But aggregating the word“cheap” reveals:

That pattern is much easier to recognize when the queries are aggregated.

1-gram vs 2-gram vs 3-gram analysis

1-gram

One-word N-grams are best for discovering broad themes such as locations, product types, modifiers or intent signals.

2-gram

Two-word combinations add context and are often more actionable.

3-gram

Three-word phrases are more specific. They can reveal valuable long-tail themes, but naturally appear less often.

There is no universally best N-gram length. In most account reviews, analyzing one through three words together provides a useful balance between broad patterns and specific intent.

How to export your Google Ads Search Terms report

1. Open the relevant Google Ads account.
2. Navigate to Search terms.
3. Select the date range you want to analyze.
4. Include the performance columns relevant to your analysis.
5. Download the report as CSV or TSV.

Useful columns normally include:

• Search term
• Impressions
• Clicks
• Cost
• Conversions
• Conversion value

Which N-gram metrics actually matter?

Example: identifying wasted spend

Imagine an ecommerce account where the following N-grams appear:

The word free immediately deserves investigation because it has consumed $390 without a conversion.

But wholesale appears much more frequently while generating conversions at a substantially lower CPA.

That could influence:

• negative keyword decisions,
• new exact or phrase-match keywords,
• campaign segmentation,
• landing page messaging,
• ad copy themes,
• budget allocation.

Don't automatically turn poor N-grams into negative keywords

N-gram analysis is evidence — not an automatic negative keyword generator.

A phrase showing poor performance may still need context.

Before excluding a pattern, consider:

• the amount of data available,
• conversion lag,
• attribution differences,
• match-type behavior,
• brand versus non-brand intent,
• whether the phrase is commercially relevant,
• and whether one unusual query is distorting the aggregate.

A practical N-gram analysis workflow

1. Export your Search Terms report.Use enough history to represent meaningful account behavior.
2. Analyze 1–3 word N-grams.Start broad, then investigate more specific phrases.
3. Sort by spend.Find where budget is concentrated.
4. Review zero-conversion patterns.These can expose wasted spend quickly.
5. Review conversion and ROAS leaders.Look for themes worth expanding.
6. Check sample size.Don't make decisions from a tiny number of occurrences.
7. Return to the original search terms.Understand the context before changing targeting.