How Google May Respond to Reverse Engineering of Spam Detection

The ultimate goal of any spam detection system is to penalize “spammy” content.

~ Reverse engineering circumvention of spam detection algorithms (Linked to below)

Four years ago, I wrote a post about a Google patent titled, The Google Rank-Modifying Spammers Patent. It told us that Google might be keeping an eye out for someone attempting to manipulate search results by spaming pages, and Google may delay responding to someone’s manipulative actions to make them think that whatever actions they were taking didn’t have an impact upon search results. That patent focused upon organic search results, and Google’s Head of Web Spam Matt Cutts responded to my post with a video in which he insisted that just because Google produced a patent on something doesn’t mean that they were going to use it. The video is titled, “What’s the latest SEO misconception that you would like to put to rest? ” Matt’s response is as follows:

I’m not sure how effective the process in that patent was, but there is a now a similar patent from Google that focuses upon rankings of local search SEO results. The patent describes this spam problem in this way:

The business listing search results, or data identifying a business, its contact information, web site address, and other associated content, may be displayed to a user such that the most relevant businesses may be easily identified. In an attempt to generate more customers, some businesses may employ methods to include multiple different listings to identify the same business. For example, a business may contribute a large number of listings for nonexistent business locations to a search engine, and each listing is provided with a contact telephone number that is associated with the actual business location. The customer may be defrauded by contacting or visiting an entity believed to be at a particular location only to learn that the business is actually operating from a completely different location. Such fraudulent marketing tactics are commonly referred to as “fake business spam”.

The patent tells us that search engines will sometimes modify how they rank businesses to keep fake businesses from showing, and they want to stop people from spamming local earch results. The patent developed in response to fake spam business listings Is:

Reverse engineering circumvention of spam detection algorithms
Inventors: Douglas Richard Grundman
Assigned to: Google
Patent 9,372,896
Granted June 21, 2016
Filed: November 26, 2013

Abstract

A spam score is assigned to a business listing when the listing is received at a search entity. A noise function is added to the spam score such that the spam score is varied. In the event that the spam score is greater than a first threshold, the listing is identified as fraudulent and the listing is not included in (or is removed from) the group of searchable business listings. In the event that the spam score is greater than a second threshold that is less than the first threshold, the listing may be flagged for inspection. The addition of the noise to the spam scores prevents potential spammers from reverse engineering the spam detecting algorithm such that more listings that are submitted to the search entity may be identified as fraudulent and not included in the group of searchable listings.

A Webmasterworld thread discussed the older patent I mentioned , and provides some interesting commentary on it that is worth reading through: Google’s Rank Modifying Patent for Spam Detection

The patent describes how it might not show any positive results in response to fake business spam to throw off people spamming results and to make it more difficult for people to reverse engineer spam detection patterns. I wasn’t convinced that being aware of this patent would help make it easier for people to spam local search results,

It may sometimes not demote a business after fake business spam has been submitted on behalf of a business, if a spam score added to a score for the listing doesn’t rise beyond a certain amount, as shown in this flowchart from the patent:

spam score

It’s difficult to say whether Google is using the process described in this patent or not (or in the post about the patent I wrote about 4 years ago.)

16 thoughts on “How Google May Respond to Reverse Engineering of Spam Detection”

  1. Hi Bill,
    You have simply done a great job. When I sarted reading your post, I didn’t understand what you meant to say.
    But after giving a read twice i understood the whole thing .
    Thanks for sharing.

  2. Hi Bill,
    It was again an amazing and informative post from you.
    You always give wonderful tips regarding search engine working. These tips are really helpful to improve the SEO of the website.
    Reverse engineering is the approach where a product is analysed to drain the concepts of product development and design. The reverse Engineering of Spam by goggle will allow knowing how spam message has been created.
    In this way it will be much easier to prevent the flood of these spam message. This will help all the website owners to prevent the spam on their Website, Thus it will reduce their time and effort on these headache.

    Thanks for sharing with us.
    With regards,
    Saurav

  3. Hey Bill, great info.

    This really explains why I’ve noticed why some business will be in the listings and all of a sudden disappear without any noticeable changes being done to the site. Im assuming somehow the’re spam score increased which lead them to getting a manual review or demotion.

    Anyways, great info!

  4. Hi Jonathan,

    This does seem like behavior that people have told me that they’ve seen before, and I mentioned with the patent I wrote about a few years ago. It is interesting seeing something like this show up.

  5. Hello bill
    You have simply done a great job. When I started reading your post, I didn’t understand what you meant to say.
    But after giving a read twice i understood the whole thing .

  6. Its a great piece of knowledge , Thanks for sharing the important and valuable information with us , i got some new ideas from it . Thanks a lot for sharing with us .

  7. Thanks for the article. This year I’ve seen 3 smb’s that have had their submissions to google My Business regularly rejected. Each is a new business, new name, new URL, but is the exact same category or use type in a space that was previously occupied by that type of space. The characteristics of the patent suggest that might have created the repeated non acceptance of the G My Business submissions. G “customer service” has been little to no help.

    But these are new businesses…and their characteristics could well trigger this algo spam rejection.

    Two big problems (well more). In the real world. Lots of businesses could well occupy space previously occupied by a similar business. The simple fact is that the new business can use the construction left behind by the old business. It can save hundreds of thousands of dollars. That is why you will see the same use in a space again and again.

    Over the many years Google local/google My Business has simply not handled these new businesses, in spaces previously built out for the same use by a previous occupant, well or accurately. In one form or another Google has been unable to accurately reflect the new business.

    Frankly the danged “stupid algo’s”. There is virtually no reason to keep old web data on an old business that is closed. Especially no reason to do so…when that data gets inputted into an algo that messes up a new business.

    Hey Google. When a business is CLOSED. Kill the data!!!!!! Don’t rely on it anymore. You’ve been seeing this for years. It occurs all the time.

    Fix your algo’s so that they reflect life’s reality not some dumb algo holding onto the past.

  8. @Dave completely agree. The people who lived in my home before me used my address for their local business (which now doesn’t exist). So now I need to either register a new address or do a massive citation clean up (pain).

    It’s the same situation with domain names…

    In my early days of SEO, I’ve built out entire sites that ended up having a penalty on them with no chance of ever ranking.

    I wonder how many legitimate business owners with websites are even aware of situations like that.

  9. Hello Bill Sir, another informative post and this “fake business spam” update is actually required. I have personally (and I believe many like me) have noticed many spam business listings appearing on Google local box ranking which is a real problem for customers and obviously for genuine and authentic businesses, especially after Google reduced the number of businesses appearing on local box.

  10. Is it because of the patent 🙂 . We know Matt cutts clearly mentioned “Patent is just an idea and we may or we may not implement the idea in our algorithms ”

    Need your valuable comments on this 🙂

  11. Hi Shiva,

    I haven’t done enough research into what is being called the possum update to get a good sense of what the harm is that it is supposedly causing. I’m going to have to look at it more clearly.

  12. Hi Shiva,

    It’s had to say whether or not this is the patent involving reverse engineering of spam detection – there are a lot of assumptions and observations being made about what people are calling the “possum” upgrade. It’s hard to say if there really is something there.

  13. Reverse engineering circumvention of spam detection algorithms … by spaming pages, and Google may delay responding to someone’s …

Leave a Reply

Your email address will not be published. Required fields are marked *