SKAGs vs Thematic Ad Groups: A Decision Matrix and Monthly Workflow for Google & Microsoft Ads

Overview

A SKAG places one target keyword, or one tightly controlled keyword concept, in an ad group. A thematic ad group places several closely related keywords together when they share search intent, ad messaging, landing page, bid logic, and negative-keyword requirements.

SKAGs can provide detailed control over bids, ad relevance, search-term analysis, and landing-page alignment. Their cost is management overhead: a large account can become difficult to navigate, maintain, and test if every small variation receives its own ad group.

Thematic grouping is usually easier to scale. It reduces duplication and lets an advertiser write one coherent set of ads for a clearly defined intent. The risk is over-grouping. If keywords require different messages, landing pages, or commercial treatments, combining them can hide useful performance differences.

A practical decision matrix

Score each proposed group from 0 to 2 for every factor: 0 means low separation, 1 means moderate separation, and 2 means strong separation. Add the scores before building or restructuring.

Use the total as a guide rather than a platform rule. A low score generally supports a thematic group. A high score supports a split, provided the expected control is worth the additional maintenance. A middle score calls for a testable structure: begin thematically, label the keyword concepts clearly, and split only when performance or operational needs justify it.

How to estimate the right structure

Estimate structure with three questions: what must be controlled separately, what can be reported together, and what can be maintained consistently?

1. Cluster by search intent. Use a keyword research tool, search-term report, or keyword clustering tool to separate queries by the problem the user is trying to solve. Do not group terms merely because they share a word.
2. Map each cluster to a destination. If two clusters need different product, service, category, or comparison pages, treat that as a strong reason to separate them. A useful cluster should have one clear landing-page promise.
3. Draft one ad concept per cluster. If the headline, benefit, qualification, or CTA must change substantially, the keywords may need separate groups. An ad copy optimizer or headline analyzer can help review clarity, but it cannot replace intent mapping.
4. Assign a control requirement. Record whether the group needs its own bid priority, budget treatment, audience setting, location logic, or conversion goal. Separate control is valuable only when someone will act on it.
5. Check scale before building. Count the number of groups, ads, destinations, and recurring review tasks. A theoretically precise account structure is not useful if it prevents regular search-term analysis and negative-keyword maintenance.

A simple planning estimate is:

Monthly maintenance load = ad groups × review minutes per group + shared campaign review time.

Use your own observed review time rather than a universal benchmark. If a proposed SKAG structure doubles the workload without creating a clear optimization action, a thematic structure is usually the more durable starting point.

Inputs and assumptions

Before restructuring Google Ads or Microsoft Ads, capture the inputs below in a keyword management spreadsheet. Keep platform performance separate from assumptions so the structure can be recalculated when conversion tracking, bids, or business priorities change.

• Keyword and match-type data: keyword, match type, search terms, impressions, clicks, cost, conversions, conversion value, and status.
• Intent label: informational, comparison, category, product, service, brand, or another label appropriate to the account.
• Message and destination: primary promise, CTA, landing-page URL, and any required qualification.
• Control needs: desired bid priority, budget priority, audience treatment, location treatment, and conversion action.
• Quality diagnostics: ad relevance, landing-page experience, expected click-through rate, and other available quality indicators. Treat these as diagnostics, not as the only reason to split.
• Business thresholds: target CPA, return requirement, minimum conversion volume for a decision, and acceptable spend without a conversion. Set these from your economics.

Match types should be selected deliberately and reviewed through actual search terms. Build a negative keyword list around clear incompatibilities, such as different products, audiences, locations, or intent stages. Avoid adding negatives simply because a term has not converted yet; first consider volume, conversion delay, tracking quality, and the value of the query.

Use consistent names. For example:

• Google Ads: CA_Search_NonBrand_Product_Exact_StandingDesk
• Microsoft Ads: US_Search_Brand_Service_Theme_Installation

Keep the naming convention short enough to scan. Include only fields that support filtering or decisions: market, network, brand status, intent, and theme. Store longer notes in the spreadsheet rather than forcing every detail into the ad-group name.

Worked examples

Example 1: Keep a thematic group together

Suppose an account contains “adjustable standing desk,” “height adjustable desk,” and “sit stand desk.” The terms target a similar product category, use the same product page, and can share a message about desk height range and delivery. Their matrix score is low because intent, destination, message, and bid priority are aligned. Begin with one thematic group, then review search terms for product-type mismatches and add carefully chosen negatives.

Example 2: Split a high-value concept

Now compare “standing desk for home office” with “standing desk for corporate office.” The product may overlap, but the audience, proof points, order size, and landing-page content could differ. If the business has separate offers or wants distinct CPA targets, split the concepts. A SKAG is not mandatory; two focused thematic groups may provide enough control without creating one group for every close variant.

CSV-ready planning example

campaign,ad_group,keyword,match_type,intent,landing_page,structure_action,negative_note
US_Search_NonBrand,Standing_Desk_Theme,adjustable standing desk,Exact,category,/standing-desks,keep thematic,exclude unrelated furniture terms
US_Search_NonBrand,Standing_Desk_Theme,height adjustable desk,Phrase,category,/standing-desks,keep thematic,review search terms
US_Search_NonBrand,Corporate_Standing_Desk,standing desk for corporate office,Exact,commercial,/business-solutions,split theme,separate commercial qualifiers
US_Search_NonBrand,Home_Office_Standing_Desk,standing desk for home office,Exact,consumer,/home-office-desks,split theme,separate consumer qualifiers

For each row, connect performance to an action. A high CTR with weak conversion rate may indicate message-to-page misalignment rather than a reason to create more groups. A low CTR with strong conversion rate may call for ad testing before restructuring. A low Quality Score should prompt a relevance and landing-page review; see the Google Ads Quality Score guide for a diagnostic workflow.

When to recalculate

Review the structure monthly, but do not force a change every month. The purpose of the review is to identify evidence for a split, merge, bid adjustment, negative, or copy test.

1. Verify measurement first. Check conversion actions, revenue values, attribution settings, and URL parameters before judging a group. Use the conversion tracking QA checklist when data appears inconsistent.
2. Review search terms. Identify new intent clusters, irrelevant queries, and terms that deserve their own landing page. Add negatives at the correct campaign or ad-group level, then document the reason.
3. Find merge candidates. Flag groups with the same message, destination, bid priority, and negative logic. Merge only after confirming that separate reporting is not needed.
4. Find split candidates. Flag themes with distinct conversion rates, CPA patterns, search intent, landing pages, or business value. Split the smallest useful unit, usually a concept or intent theme rather than every spelling variation.
5. Adjust bids and budgets after structure decisions. Use actual cost, conversion, and value data. Do not interpret a small difference as meaningful without considering volume and conversion delay.
6. Schedule copy tests. Change one major variable at a time where practical, such as benefit, qualifier, or CTA. Record the start date, audience, landing page, and success metric. An A/B test duration calculator can help plan the review window, but the final decision should account for traffic quality and business seasonality.

Recalculate sooner after a major landing-page change, product or service launch, tracking implementation, budget shift, market expansion, or platform targeting change. Refresh the decision matrix and negative-keyword logic whenever match behavior or campaign settings change. For broader account maintenance, combine this process with a PPC budget and bid management workflow, and keep naming and UTM conventions consistent using a UTM builder and campaign tracking guide.

The durable rule is simple: use SKAGs or tightly controlled groups when separate control creates a specific optimization action; use thematic groups when shared intent, messaging, destination, and economics make combined management more useful. Revisit that decision with fresh search-term and conversion data rather than treating either structure as permanent.