Last week, we saw Google updating the as yet not fully rolled out Scroll Pack with stronger and more obvious calls to action for user engagement.
This week I am seeing yet another, subtler, update; Google has added more and more colorful icons for each section of the Scroll Pack result. This is the second UI update since the Local Scroll Pack was first seen in late May.
While that sounds inconsequential as I write it, on the mobile display it dramatically increases the visibility of the pack and is quite eye catching. Clearly Google is tweaking the Local Scroll Pack to drive engagement the Local Finder.
On restaurants, the full range of attribute icons are visible on various searches.
On non-restaurant searches, the suggested alternative categories show the bright, red Maps pin instead.
Note the use of bright red pins in place of attribute icons.
When the suggested alternative search categories are expanded, the Local Scroll Pack becomes a rainbow of seductive color. With the first rollout, I worried that users would not interact with the local listings as frequently. But, while we await full rollout, this weeks and last’s upgrades have made the new Scroll Pack highly visible and likely to increase user interaction.
The new(ish) Local Scroll Pack, which remains with very limited visibility, has been updated with a bolder call to action to view more locations. In addition it now includes additional personalization gleaned from Maps and greater detail if a user is looking for more categories related to the search.
The arrows indicate areas that have changed over the past month.
In the initial preview, it was a concern whether users would be willing to both scroll AND click to see more locations. The default view only showed 1.5 locations and a scroll only surfaced 3. If the user wanted more they needed to click either the main selection view more button or choose one of the drop downs. One commenter on Twitter, only half kidding, referred to it at the “Local 1.5 Pack”.
With as many as 5 ads seen above the pack, Google seems to have realized that users were not, in fact, digging deeper. You can see images of the previous displays here, here and here.
Note also the inclusion of the Google Maps set, favorites choice in the display. A minor update has been to delineate which additional categories would be visible on the last drop down.
I am still only seeing this new display when logged in on my iPhone on either Safari or Chrome. It appears to showing based on my Google user as if I emulate a different phone on my Chrome desktop browser, I also continue to see this new display.
I have been seeing this new Scroll Pack almost continuously since late May.
I am curious how many of you have also seen it or currently see it and if you have seen it on Android?
Google Antitrust: Is It Enough for Yelp? – My bi-weekly conversation with David Mihm. We discuss the changing user behaviors and other headwinds that might keep Yelp from returning to its former glory, even if there is an anti-trust settlement.
Found Bytes – with David Mihm I had the pleasure of being interviewed by David Mihm on his new ThriveHive Podcast, Found Bytes on the topic of Google as the New Home Page.
I have been seeing a test of a new Local Scroll Pack result when logged in on my iPhone and using Safari. There is a different style presented for retail/restaurants than for service business. Brand packs have a slight variation as well.
In the brand results, there is the card style scroll pack below the ability to follow or learn more about the brand. However when one of the query suggestions is selected then the photo style result is presented. The net result of this proximity is that the user is immediately presented with competing alternatives.
It is interesting the obvious and highly visible call to action to “follow” the brand.
This is the view of this new look when a multi-location brand is queried:
It is interesting that this new Brand Scroll Pack is being displayed and obviously tested simultaneously with a segmented display that splits the results of a typical, single location brand search into multiple segments in the results. See Joy Hawkin’s write up on the Local Search Forum.
The Local Scroll Pack, first spotted last week, has now shown up in a greater range of categories including retail, restaurants, ERs and lawyers. Last week the new pack design was spotted in searches for the moving companies, HVAC, locksmiths and garage door companies.
In retail and restaurants the Local Scroll Pack includes large image thumbnails and included 5 results in the right to left scroll. The restaurant Scroll Pack included NO calls to action in the first view or second views. It required a click to the profile in the Local Finder and then another into the Business Profile to call or get driving directions.
The lawyer Local Scroll Pack, like those for service area businesses, did not include images. In these however, the only visible call to action without visiting the Local Finder, was a click to call option.
The retail/shopping oriented Local Scroll Pack, like the restaurant result required the user to click twice to get through the Local Finder and into the Business Profile (Knowledge Graph) before being presented with any call to actions.
Obviously, this is being tested and retested across multiple categories. The frequency of the test across these many categories implies that the feature is being prepped for release. How soon and with what changes, we can’t know.
Google appears to be testing (rolling out?) a new Local Scroll Pak display that provides a horizontal scrolling view of the top local search results.
The result, first pointed out to me by Todd Moss, is currently showing on my iPhone for the search Moving Companies San Francisco. It is also showing for similar searches in most other markets as well as searches for Garage Door companies, HVAC and Locksmiths. It is not showing for plumbers or handy man.
The Scroll Pak is interesting for several reasons.
The first being that while the organic local display takes up less space, Google provides two additional alternative Scroll Pak searches immediately below the primary Pak that can be access via a drop down.
The Scroll Pak is literally placed three full swipes down the page. On my iPhone 8 that means that the user has to scroll about 2700 pixels before it comes into view.
Above it? Two LSA ads, three AdWords ads and an organic Yelp results with its own scrolling list of four site-links of the ten best local movers.
When seeing this, one has to wonder if this is Google’s potential anti-trust response to Yelp’s criticism of favoring their own local results?
Below the Scroll Pak are two more spammy Yelp doorway pages, a local website and then “Interesting find” stories before delivery five more organic results for the likes of ThumbTack and Angies’ List and few more local websites. Obviously, other than for Yelp, organic search results do not appear to be a solid client acquisition strategy for the other directories.
The whole of the mobile display is nine iPhone screens long, with Ads taking up three and a half of them.
Obviously Google is maximizing display space for pay to play and while Yelp with their massive number of doorway pages did manage to get some visibility, it is limited.
While this might be a way to deal with Yelp’s antitrust complaints it is just as likely a way to minimize the visibility of local listings in categories that are filled with spam. And as long as they have minimized the potential visibility of the bad actors they might as well maximize their ad inventory and income.
Apparently Google has turned on their Reserve relationship with Trip Advisor allowing consumers to purchase tickets directly from the Business Profile (aka Knowledge Panel).
Only problem? Once activated within Google, a business can’t shut this off from within Google.
I noted last week seeing these Reserve with Google “buy tickets” button in Business Profiles for a museum in Atlanta that was driven by Trip Advisor.
The Buy Tickets buttons, apparently now visible on many more venues, are very visible on both mobile and the desktop and are an obviously clear call to action like the Book button for restaurants and spas.
Interestingly, Universal Studios which has locations in both Hollywood and Orlando, only has the button on their Hollywood location. Both locations use the TA booking platform. So either this is an early test or someone in the org turned it on without telling others.
When the Buy Tickets button is clicked you are brought to an e-commerce interface allowing you to select the date, the type and number of tickets and then asked to pay with Google Pay. If you don’t have a Google Pay credit card set up yet, you are then directed to do so.
Unlike, the Reserve scheduling tools for spas, this completes the whole transaction from scheduling to the purchase for TripAdvisor and the venue directly on and within Google.
The mobile and desktop experiences are very similar with the exception that the buy button is white with an outline on my iPhone.
Reserve with Google, at least with spas and gyms cost nothing to participate in beyond the costs of the scheduling software that the business was already using. Once turned on by the business, the business was provided with some analytics about the Google based interactions but in the case of Open Table, those analytics were removed.
Whether that is the case with the Buy Tickets functionality through Trip Advisor is not clear. In December Trip Advisor was noted on the Reserve Partners page as “coming soon”. Now however they are noted as a full partner on that page.
Where is the (effing) Off Button?
However what is clear for all Reserve buttons is that once they are turned on, whether automatically by Google or by the business, is that there is no way to turn it off via the GMB dashboard.
Obviously this whole thing is confusing to the businesses involved as indicated by this post in the forum:
The poster, justifiably, thinks that “this buy tickets option isn’t for our own tickets, it’s for some 3rd party site selling hollywood tours”.
This is so “Googly” to create a feature that makes sense for them, mostly makes sense for the consumer but is totally screwed up for the business.
In the case noted in the forum by Universal Studios, it appears that the Buy Tickets feature was inserted without the businesses knowledge or consent and there appears to be no way for them to easily remove it.
Google has started the rollout of “Short Names”; the ability of a business to create a custom, short URL that then can be used for easy access to their business profile on Google.
Customers can enter the short name URL in the browser’s address bar, like “g.page/[yourcustomname]”, to go directly to your Business Profile.
The short name is created in the Info area of the GMB. The name has to be at least 5 characters in length and not previously taken.
So in the case of Barbara Oliver Jewelry the short name would be g.page/BarbaraOliverJewelry which then redirects to her search profile page on mobile and to her Maps profile page on the desktop. The name must be at least 5 characters in length.
Currently the feature is not yet surfaced in the API and for bulk users, short names have to be claimed individually on an individual listing basis.
In discussing this feature with Google, they noted that this not so much a feature for a special or custom name but to be able to provide a long term, easy way for consumers to interact with a businesses listing.
Google via the GMB interface will encourage a business to claim and share this short URL where it makes sense and where the business could gain value.
The current use cases are limited but the short name will allow for more transactional (in the very broad sense) capabilities down the road. The example shared by Google was as a review link.
Businesses will be able to change their short name although there will be limits on the frequency of the changes.
There is a newly released help page: Create a short name & URL for your business where Google details how to create a short name as well as how users and business owners can “flag a name for impersonation, offensive, fake, spam or inappropriate content issues”.
This appears to be very early days for a feature that can be expanded into many users. So while it is not useful today, it will become more so over time and it makes sense to choose a name as soon as you are able.
How well Google will police and moderate the inevitable conflicts and abuses remains to be seen.
TL;DR On the shoulders of others, a technique that can help you uncover Google Local Spam at greater scale.
The Inspiration
Over the years, Google Maps has lost many advanced search features* making spam identification harder and harder. A few weeks back Dave DiGregorio, a local SEO in NJ, posted a tip on Twitter that helped at ferreting out some of the spam:
Using the allintitle command in the Local Finder and then zooming out allows you to find up to 200 instances of spam. Better than a fish!
While Dave’s technique might surface some of the spam, it still left me feeling frustrated due to the fact that Google Maps was restricted to showing only 20 results and providing no option to see more. Thus to develop anything approaching a significant list, one had to craft the search for each and every geography that held that spam.
And I knew that for some of these terms, particularly those generated by lead gen networks, there were hundreds if not thousands of spammy listings in play in virtually every town and city in the US.
But I took Dave’s idea and refined it using the Local Finder on the desktop* (as opposed to Google Maps) to allow me to surface hundreds of related spam listings instead of the 20 that Maps were serving up.
The Technique
Start with a local search on Google.com that returns the 3 Pack
Select the option to view More Places
In the Local Finder’s search field, replace what you have with the command: allintitle:spammy business name or allintitle:”spammy business name” (using the quotes to be more precise). For example: allintitle:”cheap car insurance”, or allintitle:”dog bite layer“.
Using the zoom out Map button, bring the whole country into view.
The search will continue to change as you zoom out to view more of the US.
If you zoom too far, the listings will disappear so zoom back in one step for maximum coverage
Scroll to the bottom and you will that you now have as many as 10 pages, each with 20 listings.
Here is a video detailing the technique:
Why do you need this?
Google typically responds much better to reports of massive abuse at scale than tales of a single mom and pop down the road adding some title spam.
If you really want to clean up the space for your local client in say, Philadelphia, it is more likely to be taken down if you provide details for hundreds of listings.
The sad reality is that you should be able to report 1 or 2 listings and have their magical, world changing AI figure out the other 2000 similar listings.**
Big surprise, it can’t or won’t. It must be working on more important problems.
What’s Next
The problem with this technique is that while you can find hundreds of listings, Google needs you to report the URL for the listing for them to do anything. That still adds up to a ton of work.
I am working with Darren Shaw and the good folks at Whitespark and to capture these listings with the URL at bulk to make my job of reporting them to Google easier. I will keep you posted as to our progress.
A favor to ask
If you use this technique to uncover spam at scale I would love to see some of the searches!
Footnotes
*In days gone by you could do name, phone number or URL searches in Maps and it would return hundreds and hundreds of listings. This evened the playing field somewhat and gave me the tools to find and report massive networks. But as Google scaled back these advanced features Maps became less and less of a tool that could be used to uncover the very spam that was populating it.
*This command and technique ONLY works on the desktop and won’t work on mobile. And likewise for any sample Urls in this post .
* This still means that you have to put in a fair bit of volunteer work to get Google to take notice and that sucks. But if it helps the client and you are getting paid, there could be worse things to do.
*While 200 listings is better than nothing, we still can not easily determine how many there might be beyond those 200.
In 2009 I met with the then head of Google Local, Carter Maslan, and asked why they allowed so much spam to continue to exist in Maps after they were alerted to it. He assured me that it wasn’t very visible and it was being used to “inoculate the system” against future abuses.
The Disease
He spoke as if it was just a simple case of the chicken pox and the vaccine was already in hand.
Well, it turns out, more than 10 years later that spam in Local was more like the Spanish Flu pandemic of 1918* and the vaccine is still in the labs. And people are dying.
A screen grab from 2009 showing the density of “Locksmiths” in NYC.
OK people are not dying but Google Local is. The scale and scope of lead gen spam in Google Local could be fatal to consumer trust.
Who knows when Rank Brain will come out of its coma to solve the issue and truly vaccinate the system.
The Evidence
A case in point is a lead gen network that I have written about AND reported to Google both via the normal reporting mechanism as well as filing a detailed report via Twitter*.
Problem is, either they didn’t believe me or else they can’t or won’t do anything about it, because it still exists at quite large scale.
All of the spam listings have a similar pattern and include an common insurance search phrase, often note that they are open 24 hours and frequently include a city name before or after the keyword laden name.
When you dial any of the listings you are put through to a sophisticated automated call tree that asks the same qualifying questions each time:
if you already have insurance or no
if you have been continuously insured for 12 months or no
what is your zip code
which insurance carrier you currently have
Based on the answers the system sends you along to an insurance agent that can sell you a competing product. The idea is that if you already have Allstate and are calling for “cheap car insurance” you must want a different brand.
I have dialed dozens and recorded several and the call tree is exactly the same for all of the listings.
The only thing that changes is that, depending on the zip you are in and which insurance you have, you will be forwarded to a different “local”* agent as a “live lead”.
You can hear a sample of that call tree that occurs on every one of the spam listings in Google Local here:
The Scope of the problem
I have identified hundreds listings associated with this particular vertical lead gen network at Google. There are multiple keyword based listings in many markets and coverage is country wide.
Given the Google Local predilection to return “branded” search results, front page visibility is quite high in many markets.
When visiting my son in Santa Fe last month, I performed this search and the lead gen spam ranked #1.
And clearly, Google is not the only service being spammed as these listings are common in Bing as well.
Just because Bing has the same problem, Google by virtue of their bigger reach and their monopoly position has a much greater responsibility to solve this issue.
How it works
Essentially, it is a live lead generation service. Unlike lead gen that involves a form and follow up, this puts the searcher directly in touch with a business that has an interest in servicing the immediate inquiry.
In speaking with several of the agents on the other end, they are only charged if they accept the call. If a call is less than 2 minutes in length, they are charged for it but can request credit. If the call is not answered it goes on to the next agency in the lead generator’s queue.
Generally these agencies say that the lead quality is fair. Given the search term “cheap car insurance” it is no wonder.
According to the company that is selling these leads, close rates run from 30-50% depending on whether the agency is a national brand or a local agent and whether they are “transactional’ in their approach and properly qualify the caller.
The cost of the lead is $42 for an insured auto driver and $35 for an uninsured one. If an agent in fact does close 50% then it could cost as little as $84 to sell a single policy. In speaking with an agent at the end of the line, it would appear that the close rate is not that high. If the close rate is in the more likely 30% range* it would put the cost of selling a single policy closer to $125 under the program.
From the point of view of the lead gen company its a good deal as there is obviously zero acquisition cost on Google Local spam.
According to their sales rep, they have “from 30-50 different campaigns from publishers through various mediums to drive consumers to call a phone number.” And we know that at least two of those “publishers” are spam networks on Bing and Google.
I assume that worse case for the company behind this is that the Google Local spam dramatically drives down their costs even if they do actually have additional paid campaigns. And given that Bing Local spam is also being used, it is not clear how large a part of their profit spammy listings play.
Given the reach of Google Local, I assume that they make significant money by leveraging these fake listing calls.
Who is behind it
Clearly lead gen technologies can be outsourced to get all of the technologies needed to provide the service. VOIP phone numbers and call centers can all be assembled from a range of outsourcing companies for a solution. That being said in most other cases of Google Local lead gen that I have examined you find human call centers answering for the spammy listing.
But in this insurance case, the call tree technology is somewhat distinctive and seems to point to a single company. The company was identified by one of the insurance agents I spoke with and the specific call tree technology I experienced was confirmed by the company’s sales rep in a subsequent conversation.
When I asked the lead gen company’s sales rep about the use of Google Local for lead gen she assured me that it was a paid ad. I took the time to show her the spam, which temporarily slowed her down and lead her to pause but she quickly recovered and moved on with her sales effort. While she recognized it as spam, it is not clear how much she knows about her company’s “efforts” in this arena.
On this company’s website they note that they also provide lead generation services for medicare supplement, home security, rehab and the mortgage industries. I have yet to find any examples of these industries that have similarly spammy listings using this same call technology but it bears investigation. The rehab industry is not known for fair play.
Who is really behind it?
The one question that isn’t yet answered is whether this particular lead gen company or any of the ones that we have been seeing create their own local spam or whether that is outsourced to companies that specialize in generating these listings at scale.
It seems to me that someone, somehow has figured out how to insert these listings into Google at scale. Whether it is through a rogue employee working for one of the companies that is white listed to add listings or a rogue contractor for Google or a criminal element this is currently unclear.
What is the harm?
Someone on Twitter asked me if there was in fact any harm to this technique since the customer got an insurance quote and a local agency somewhere got the business.
I would contend that the harm is in fact significant.
First and foremost the consumer is being deceived. They are looking for an agency in their town, city or suburb and are then being pre-qualified and shunted off to the highest bidder. It is deceptive advertising at best. In reality it is more like a two sided marketplace where the buyer and seller are clueless to the actual con that is taking place .
Secondly, local businesses that deserve and need the exposure of the Google Local pack are being pushed out of view and losing valuable opportunities that Google Local promises.
These harms are palpable and don’t speak to the larger potential harm that is being done to erode trust in the whole of the local ecosystem.
All the while the lead gen company is taking its receipts to the bank* and Google sits doing little about this problem. Essentially both companies have externalized the costs of this problem to the unsuspecting consumer and other businesses.
Footnotes:
* It is estimated that about 500 million people or one-third of the world’s population became infected with this virus. The number of deaths was estimated to be at least 50 million worldwide with about 675,000 occurring in the United States.
* Google may have removed a few of the listings that I passed their way but removing a few and ignoring the obvious pattern is definitely akin to the boy with his thumb in the dike. As a volunteer, life is not long enough to send them hundreds and hundreds of links when they could easily user their tools to surface these.
*Local is a stretch. Many of the listings are situated in rural communities but the agencies picking up the calls are typically larger agencies hundred of miles away
*30% is still high for a lead gen service and the real time search and answer nature of the process can explain that
*I estimate that there are at least 500 insurance listings from this particular network. Although there could be more. Any one call is a marginal incremental cost to the company. If you estimate that each of these 500 locations generates 10 calls a month at $42.00 that comes out to $252,000 year. Not a bad return. If you multiply that by the number of industries they might be in, things really start looking up.
