Google is testing another way to compress Search ads: show only part of an advertiser's sitelink set by default, then make the user manually expand the ad to reveal the remaining links. The experiment adds a numeric indicator telling searchers how many additional sitelinks are hidden behind the collapsed control.
The variation was documented on September 4 by Barry Schwartz at Search Engine Roundtable. Arpan Banerjee first shared the format, and Schwartz said he was able to reproduce it independently. In the initial state, a limited set of sitelinks is visible; clicking the expansion control opens the rest.
Expandable sitelinks are not new. Google has tested multiple versions over the years, including mobile formats, arrows, expandable descriptions and other ways of hiding secondary links until the user interacts. What is notable here is the updated collapsed-by-default treatment and the explicit count showing that more links are available.
The ad is asking users to make a second choice
Sitelinks traditionally work by exposing several alternative destinations directly inside an ad. A searcher might ignore the advertiser's main landing page and jump straight to pricing, a product category, store information or another high-intent page.
The collapsed format changes that interaction slightly. The user first sees the primary ad and a subset of its sitelinks. To inspect the complete set, the searcher has to decide that the ad is interesting enough to expand.
That extra interaction is small, but it creates a meaningful distinction between links that are immediately visible and links that are technically attached to the ad but hidden until requested. For advertisers, “eligible to show” and “actually seen without interaction” become different states.
Google may be balancing richer ads against limited SERP space
Google encourages advertisers to provide multiple assets because they give the system more options when assembling an ad. Its current sitelink documentation recommends creating at least four sitelinks and notes that shorter link text can make it possible to display more links.
The company also says advertisers that increase the number of sitelinks per campaign to six can, on average, see up to 3.5% more conversions at a similar cost per conversion, based on Google's internal May 2025 data. More available links can therefore be useful to both the advertiser and Google's optimization system.
But more links also make an ad physically larger. On a search results page where sponsored results, organic listings, shopping modules, AI features and other elements compete for vertical space, showing every available asset at once can create very tall ad units.
A collapsed format offers a compromise: Google can preserve access to a richer sitelink set without dedicating the full amount of screen space to it before the user signals interest.
The counter makes hidden inventory visible
The numeric indicator is an important detail. Simply collapsing links risks making users unaware that additional destinations exist. Showing a count communicates that the advertiser has more options available even though Google is not displaying them immediately.
That can function almost like a disclosure affordance: the ad remains compact, but the interface tells the user there is more to explore. A person looking for a specific page may choose to expand rather than clicking the main headline.
The effectiveness of that design will depend on whether searchers understand the control and consider the hidden links worth another interaction. A count may increase curiosity, but it may also be ignored by users who scan search results quickly.
Above-the-fold sitelinks could become more valuable
If Google were to expand this format, advertisers might need to think more carefully about which sitelinks are most likely to appear before expansion. Today, advertisers create assets and Google determines which eligible combinations to show based on expected performance and relevance.
Google's asset documentation explains that the system selects additional assets for individual searches when it expects them to improve performance. Sitelinks can be supplied at account, campaign, asset-group or ad-group level, and assets from different levels can sometimes serve together.
Advertisers therefore do not have a simple fixed ordering that guarantees a particular sitelink will always occupy the first visible position. If only a subset appears before the expansion control, Google's selection logic becomes even more consequential.
A high-value link such as Pricing or Book a Demo could receive very different exposure depending on whether it appears immediately or behind the collapsed section.
This could affect sitelink click distribution without changing advertiser setup
One of the interesting aspects of the test is that advertisers may not need to change anything for the user experience to change. The sitelink assets can remain configured exactly as before while Google alters how many are initially exposed.
That means performance shifts could appear even when campaign settings are stable. A sitelink that historically received meaningful direct traffic might see fewer clicks if it is increasingly hidden behind an expansion control. Conversely, the links chosen for the default state could capture a larger share of sitelink interactions.
Advertisers monitoring asset performance should therefore avoid assuming that every change in click volume comes from bids, copy or demand. Search-result presentation itself can redistribute attention among assets.
Google already treats ad format as dynamic
The experiment fits Google's broader approach to Search ads. Advertisers provide headlines, descriptions and assets, but Google's systems decide how to combine and present those components for a particular auction and user.
Official guidance says assets such as sitelinks and images are shown when Google predicts they can improve results. The company can also use automated assets and flexible formatting to assemble a final ad that differs from a static creative preview.
Collapsed sitelinks extend that logic from selection into interaction design. Google is not only deciding which assets deserve to appear; it is testing whether some of them should remain one click away until the searcher explicitly asks to see more.
A smaller default ad can change competitive spacing
Search advertisers care about ad assets partly because larger ads occupy more visual territory. A sitelink-rich ad can push competing results farther down the page and give one advertiser several clickable destinations in a prominent block.
If Google compresses that block by default, the surrounding results may become visible sooner. On mobile in particular, where vertical space is scarce, the difference between a fully expanded sitelink set and a collapsed control can materially change how much of the next result appears on screen.
That does not necessarily make the format worse for the advertiser. A more compact ad may be easier to scan and less overwhelming. The relevant question is whether reduced initial exposure is compensated by better engagement from users who intentionally expand it.
The test could produce a useful intent signal for Google
Manual expansion also creates a new behavioral event. A user who clicks to reveal additional sitelinks is signaling interest in the advertiser while not yet committing to a destination.
Google has not said that this interaction is used as a ranking, bidding or optimization signal, and there is no basis for assuming that it is. But from a product-design perspective, expansion gives Google measurable information about whether users find the additional choices useful.
If the format is being evaluated experimentally, Google can compare how often users expand, which hidden links they choose afterward and whether the compressed presentation changes overall ad engagement. Those outcomes would help determine whether the space-saving design is worth a broader rollout.
Advertisers should not reduce their sitelink inventory because of the test
The existence of collapsed sitelinks is not a reason to stop supplying multiple assets. Google's current guidance still encourages advertisers to provide a healthy set of relevant sitelinks so its systems have enough options to match different searches.
Google recommends at least four sitelinks at account or campaign level and at least two for mobile. It also advises using unique link text, keeping labels concise and adding descriptions because richer information can unlock additional ad formats.
The sensible response is therefore not to provide fewer links in an attempt to avoid the collapsed interface. It is to make each link useful enough that, whether it appears immediately or after expansion, it offers a meaningful destination for the query.
The distinction between configuration and presentation keeps growing
Paid search once gave advertisers a relatively direct relationship between the ad they built and the ad the user saw. Responsive Search Ads and automated assets have already weakened that one-to-one relationship. Advertisers now provide a pool of components and Google's systems assemble the final experience.
This sitelink test reinforces the same direction. Even after Google decides which links are relevant enough to attach, the interface can determine which are visible, which are hidden and how the user reveals them.
For PPC teams, that makes SERP observation increasingly important. The Google Ads interface can tell an advertiser which assets exist and how they perform in aggregate, but it cannot always communicate every presentation experiment happening in live search results.
This is a test, not a confirmed redesign
The most important qualification is that Google has not announced collapsed sitelinks with a numeric expansion counter as the new standard Search ad format. Schwartz describes it as a test, and expandable sitelink experiments have appeared in several forms before without becoming the only presentation users see.
Advertisers should therefore avoid redesigning campaign structures around one observed interface variation. The useful action is to monitor whether the format appears more frequently, whether Google documents it, and whether sitelink performance changes in accounts where the test is visible.
If the format expands, it would represent another step toward a Search results page that reveals information progressively rather than displaying every available ad asset at once. For users, that can mean a cleaner initial result. For advertisers, it means some of the links they worked to create may increasingly depend on one extra click before they can compete for attention.
The sitelinks are still there. The experiment is testing whether Google should make searchers ask to see all of them.