SKAGs vs. STAGs: Which Ad Group Strategy Works Best Now?

The SKAGs vs. STAGs debate comes down to one question: should each ad group hold a single keyword, or a tight cluster of keywords that share the same intent? For most Google Ads accounts in 2026, single theme ad groups (STAGs) win, because close variant matching, responsive search ads and Smart Bidding have quietly dismantled the control that single keyword ad groups (SKAGs) were built to give you. There are still a handful of situations where a SKAG earns its place, and this post covers both sides honestly.

What a SKAG actually is

A single keyword ad group contains exactly one keyword, usually in two or three match types (exact, phrase, and occasionally modified broad back when that existed). The idea was surgical: one keyword, one ad set written to mirror that keyword word for word, one landing page. Structures like this were popular between roughly 2015 and 2019, when an exact match keyword still meant more or less exactly what you typed.

The payoff was message match. If someone searched “emergency plumber Manchester” and your headline read “Emergency Plumber Manchester,” you got a click-through rate lift and, often, a better Quality Score. Better Quality Score meant cheaper clicks, so the extra admin work paid for itself.

What a STAG is

A single theme ad group holds a small group of keywords (typically three to fifteen) that share one search intent and would sensibly be answered by the same ad and the same landing page. Think of “boiler repair,” “boiler repairs near me,” “fix broken boiler” and “boiler not working” living together, while “new boiler installation” sits in its own group entirely.

The grouping rule is intent, not string similarity. If two keywords would make you write a different offer, a different headline or send traffic to a different page, they belong in different ad groups. That test is more useful than any keyword count.

Why SKAGs lost most of their advantage

Three platform changes eroded the case for one keyword per ad group, and none of them are reversible.

• Close variants. Since 2018 and 2019, exact match has covered same-meaning queries, including reordered words, implied words and synonyms. Google documents this behaviour in its close variants help page, and it means your “single” keyword is already matching dozens of queries you never added.
• Responsive search ads. With expanded text ads retired in 2022, you supply up to 15 headlines and 4 descriptions, and Google assembles the combination. Pinning every headline to force exact message match usually restricts the asset mix and can cap performance.
• Smart Bidding needs volume. Automated bid strategies read conversion signals at the campaign and ad group level. Splitting 40 conversions a month across 30 SKAGs leaves every group statistically starved.

The practical result: two SKAGs often compete for the same query, Google picks a winner based on its own ranking of expected performance, and your “control” turns out to be an illusion. Understanding how match types behave today is more useful than trying to fence each keyword off.

SKAGs vs. STAGs: a straight comparison

• Message match: SKAGs still edge it slightly, though dynamic keyword insertion inside a STAG closes most of the gap.
• Data density: STAGs win clearly. Pooled conversion data inside one ad group gives Smart Bidding something to learn from.
• Management time: A 200 keyword account becomes 200 ad groups under SKAGs, or roughly 15 to 25 under STAGs. That is hours per week, every week.
• Asset quality: STAGs let you write fewer, better RSAs instead of 200 thin ones assembled from a template.
• Internal competition: SKAGs create overlap and cannibalised impressions once close variants are in play.
• Reporting clarity: SKAGs give per keyword P&L at a glance, which is genuinely handy for small, high value keyword sets.

When a SKAG is still the right call

SKAGs are not dead, they are just specialised. Keep a single keyword ad group when at least one of these is true:

• The keyword is a brand term. Your own brand, and competitor brands, deserve isolation so bids, ads and budgets never bleed into generic traffic.
• One keyword carries outsized spend. If a single term eats more than roughly 15 to 20 percent of campaign budget, isolating it makes bid control and reporting cleaner.
• The intent is genuinely unique. “Solar panel grants” and “solar panel cost” need different ads and different pages, so one keyword per group happens naturally.
• You are running a strict landing page test. Isolation removes query noise from the result.

Outside those cases, the maintenance cost rarely returns anything. Accounts spending under roughly £2,000 a month almost never generate enough conversions to justify dozens of near empty ad groups.

How to build STAGs that actually perform

Good theme grouping starts in the keyword list, not the ad builder. Work through the process in this order.

1. Cluster by intent during research. Sort your list into buckets like commercial, comparison, price, location and problem-aware before you touch the interface. Our walkthrough on keyword research for Google Ads covers the sorting step in detail.
2. Cap each group at around 15 keywords. Past that, the ad copy starts serving nobody well.
3. Write three to five RSA headlines that fit every keyword in the group. If a headline only suits two of the ten keywords, the theme is too broad.
4. Match the landing page to the theme. One page per ad group is the cleanest rule, and it protects landing page experience, one of the three Quality Score components.
5. Add negatives between groups. Cross-negating stops your “repair” theme from absorbing “installation” queries.
6. Layer extensions per theme. Sitelinks and callouts should reinforce the specific intent, as covered in our guide to using ad extensions effectively.

Once the structure holds, most of your improvement comes from the creative itself rather than the containers around it. The fundamentals in what makes a search ad convert apply either way.

Migrating from SKAGs to STAGs without tanking performance

Rebuilding an account is disruptive, so stage it. A sensible sequence looks like this:

• Export 90 days of search terms and tag each one with its real intent, ignoring which ad group it landed in.
• Build the new themed ad groups in the same campaign, paused, with fresh RSAs.
• Enable the new groups and pause the old SKAGs on the same day, so budget shifts cleanly rather than splitting.
• Expect a learning period of roughly one to two weeks, or until the campaign records about 30 conversions, before judging anything.
• Hold bidding steady through the transition. Changing structure and bid strategy at once makes the results unreadable.

Google’s own ad group best practice guidance recommends grouping similar keywords around a shared theme, which is essentially an endorsement of the STAG model.

Where automated campaign types fit in

Ad group structure matters less in campaign types that do not expose ad groups in the traditional sense. Performance Max uses asset groups and audience signals instead, which shifts your effort from keyword grouping to feed quality and creative variety. If you run both, use search campaigns with STAGs for your high intent terms and let automation handle discovery, as explained in our Performance Max guide. Pairing either approach with remarketing usually lifts blended conversion rates more than any structural tweak.

Frequently Asked Questions

Are SKAGs dead in 2026?

No, but they are a niche tool rather than a default structure, and most accounts should use fewer than five of them. SKAGs still make sense for brand terms, competitor terms and any single keyword taking a large share of budget, while everything else performs better grouped by theme.

How many keywords should a single theme ad group contain?

Between three and fifteen keywords is the practical range, with five to ten being the sweet spot for most search campaigns. The real test is whether one set of RSA headlines can speak convincingly to every keyword in the group.

Do SKAGs still improve Quality Score?

They can lift expected click-through rate slightly, but Quality Score is calculated per keyword against the query, so tight themed grouping with relevant ad copy achieves a very similar result. Landing page experience and ad relevance usually move the number more than ad group size does.

Will switching to STAGs reduce my costs?

Most accounts see management time drop sharply and CPCs stay flat or fall modestly, though the bigger gain is usually a better conversion rate from stronger ad copy. Give any restructure at least 30 conversions or two weeks before drawing conclusions.

Can I mix SKAGs and STAGs in one account?

Yes, and hybrid structures are common in well run accounts. Keep brand and top spending keywords isolated, group the rest by intent, and add cross-group negatives so the two structures do not compete for the same queries.

Want a second opinion on your account structure?

If your ad groups grew organically and nobody has audited them since close variants changed the rules, a structural review often finds wasted spend within the first hour. SEO Quirk works with businesses that want a local agency they can actually speak to, so get in touch and we will look at how your campaigns are put together.