What I Learned from 50 Examples of the New Local SERP
First impressions can be deceiving. When users began sharing examples of the new “mega map” desktop layout for Google’s local pack – confirmed as an official rollout by Google on December 9 – they all shared a similar look and feel, with a wide layout that took up the entire main content area above the fold. Especially notable was the lack of ad placements above the new local pack, creating a particularly immersive search experience where essentially the first half of the main SERP page was all organic and all local. Here’s an example:
One can imagine that this more prominent placement might lead to greater click-through rates for local packs, making it more critical than ever to rank in the top three results for local searches. The increased real estate for the local UX might also lend itself to the introduction of new features we haven’t seen before.
After conducting more than 50 “local intent” searches, however, I’ve found that not all of them return the new “mega map,” nor is the new layout as consistent as it at first appeared it would be. The range of searches I tried includes generic keyword searches for brick-and-mortar stores, such as the example above, as well as searches for local service providers, chain stores, products, and more. I tried covering a broad base of searches covering a range of categories. I made sure all of my searches would be interpreted as local by appending “san francisco” to each query.
Here’s a summary of my overall findings:
- Of the 50+ searches I conducted, 80% returned the new layout; 16% returned the old local pack; and 4% (two examples) returned something else
- Of searches returning the new layout, nearly half (48%) had from one to four text ads at the top of the page, causing the local pack display to go beyond the top of the fold
- Searches for local service providers returned Local Services Ads, usually in addition to top-of-page text ads
- Searches for products tended to trigger the old version of the local pack, usually accompanied by product ads
- Searches for chains tended to trigger the new layout
- Only two of the new local pack results included any sponsored listings – and these were anomalous examples where the local pack was halfway down the page – whereas 44% of the old local packs had a sponsored listing in the first position
This last finding is notable, suggesting that Google may be moving away from ad placements in the local pack, at least when it chooses to display the full-width layout prominently. (I did find sponsored listings when clicking through from the new layout to the local finder.)
The Above-the-Fold Mega Map
Searches for brick-and-mortar businesses like department stores, fabric stores, donut shops, hair salons, comic book stores, and pharmacies returned the above-the-fold edition of the new layout, with no ads intruding at the top, as did searches for chains like Starbucks and Target, searches for leisure activities like laser tag and horseback riding, and miscellaneous other examples like charter schools, emergency rooms, and natural stone exporters. The common thread in these searches seemed to be the primary intent of finding a brick-and-mortar location rather than a product or a service.
The Mega Map with Ads
Several searches returned the new layout with one to four top-placed ads above it, including searches for accountants, allergists, bicycle stores, and car rental. I didn’t see too much of a pattern in these categories aside from, perhaps, the availability of ad inventory, or the possibility that the searcher’s needs might be as easily filled by an online product or service as by visiting a store or office. For instance, a search for gift shops in San Francisco returned ads for the online gift stores on the websites for SF MOMA and SF Parks.
Even More Ads for Local Service Categories
For local service categories with Local Services Ads inventory, Google served these ads up as well, generally along with top-of-page ads which appeared beneath the LSA carousel.
Product Searches Show the Old Local Pack
A significant minority of searches, some 16%, returned the old local pack layout, which we might now call the “mini map.” These searches did seem to fall into a distinct pattern: they were searches for branded products or specific product categories, such as Xbox, Dyson vacuum, Super Mario Brothers Playset, high def TVs, and eyeglasses. Again, I always included “san francisco” in order to convey local intent, but Google made sure to display product ads in these SERPs wherever possible, presenting the user with the alternative to purchase the same products online. Is Google retaining the old local pack as part of a redefined product search experience?
The Ambiguity Factor
What happens when local intent is less clear? I tested this by leaving the geo-modifier out of the search query and found that this produced a different result in some cases. In the example below, Google isn’t sure whether I want a local business or general information about sushi, and so I see an old-style local pack alongside a knowledge panel about sushi as a topic.
In fact, old-style local packs plus a knowledge panel were the most common variation when the geo-modifier was removed, appearing in 32% of searches that previously had returned the new layout. Those searches seemed to fall into a clear pattern where the user might be looking for general information rather than local results, and included terms such as laser tag, Tae Kwon Do, insurance agent, photographer, and web hosting. In three cases – lighting store, rare books, and fabric store – Google showcased product ads rather than a knowledge panel, also defaulting back to the old local pack layout. The appearance of the old layout would seem to indicate, now, that Google is less confident in the user’s local intent.
Anomalies in Local SERPs
There were a few odd results that didn’t fit any of the above patterns, such as “off track betting san francisco,” which returned an old-style local pack (I tried other “sensitive” categories like cannabis store and casino, but these returned the new layout). A search for “golf course san francisco” returned what might be termed a “destination carousel” and no local pack at all, the only such example in the group. Note this carousel is another example of a full-width display.
I mentioned in my intro that only two examples of the new layout contained a sponsored listing. These were actually variations without the geo-modifier “san francisco,” as shown below. It’s hard to say why more ambiguity about local intent would make Google more likely to show a sponsored listing. As I’ve noted, the two new local packs with sponsored listings also appeared further down the page. One example is below.
Finally, “boutique hotel san francisco” returned the regular hotel pack, which has been around for a while as a differentiated search experience for hotels. As with product searches, Google seems to want to maintain a separate search experience for hotels and accommodations, unless of course we are in a transitional phase with more changes forthcoming.
Conclusion
Of course, we are always in a transitional phase with Google SERPs. However, given the patterns I’ve observed here, I don’t think I’m seeing a change that is still rolling out, but rather phase one of a new layout that includes a few intentional variations and exception cases. Google has been moving toward a more verticalized search experience in local for some time now, with hotel searches looking different from those for retail stores or service providers, and these layout variations extend Google’s ability to make some vertical search results behave differently from others based on a more refined view of user intent than before.
So too, the new local layout belongs to a trend toward full-width design in search results that may see more aspects of the SERP becoming full-width over time.
Note: I had a colleague perform all of the same searches I did, with some additional variations in query language such as “near me” as an indicator of local intent. Though his results did not match mine exactly, the degree of variation was modest, and the trends I’ve described were largely consistent.