Can You Use Amazon’s Mechanical Turk To Change Google Autocomplete?

For those of us working to influence Google Autocomplete, Mechanical Turk is a great tool to use. You can easily create an account, upload your .csv file and have random people from all over the United States (or the world) perform your Google searches for you.

If you’ve done any research on how to change Google Autocomplete, you know that diversity is key. The fastest path to success is to get your searches done from as many different computers as possible.

One would hope that Mechanical Turk would be a great help in this effort, but based on a couple of tests I’ve run recently, it’s not ideal. Let’s look at the data.

Test Batch #1

I uploaded a list of 18 unique keywords I wanted searched. Each unique keyword search is called a HIT and by default, MTurk only allows 1 person to work on any one HIT. So, it’s possible that 1 person could perform all 18 searches, since each one is a unique HIT.

In my test batch, I required that each keyword be searched 10 times. So, worst case, I’d get 10 different people performing all 18 searches. Best case, I’d get 180 different people searching for a keyword 1 time.

My result was that I got 21 different people to perform the searches. I was hoping for a little more diversity.

Test Batch #2

On my 2nd test batch, I uploaded the exact same .csv a day later – 18 unique keywords X 10 searches per keyword. Again, the worst case scenario would be I’d get 10 people performing the searches; best case would be 180.

In this batch, I only had 14 different people performing the searches. Not very good.

When I merged the two result files together, I counted 33 different people performed the searches. Thus, only two people worked on both batches.

Test Batch #3

On a 3rd batch, I uploaded 11 keywords as unique HITs and required 15 searches for each keyword. Best case, I’d get 165 unique searchers – worst case, I’d get 15.

I fared much better in this instance, getting 36 different people to perform the searches.

Additional Thoughts On Mechanical Turk & Google Autocomplete

There are some other limitations to Mechanical Turk that make it not an ideal solution for trying to change Google’s suggestions.

  1. Batches are completed all at once. I’m not a developer, so I upload batches under the standard interface. In this interface, I have not found a way to “drip” the HITs over time. So, it’s likely that all of your searches are going to occur within a day or two, as opposed to being spread out more naturally. Not ideal.
  2. Not location specific. More and more, Google is localizing autofill results based on where you search from. For instance, if you Google something in Portland, Oregon, you might see entirely different suggestions if you Google it from New York. Mechanical Turk only allows you to specify workers by country, not by city.
  3. Can’t collect geographic information. I track every single search our team performs, so I know when it was done, where it was done and what was searched. Unfortunately, Turk’s TOS do not allow you to collect ANY identifying information on workers. I tried adding a question, “What is your city and state?” and my HIT was rejected. Without city/state information, my reporting data is incomplete.


The upside to Mechanical Turk is that it works. Once you’ve designed your HIT template, it’s easy to upload a task. Of course, you’ll have to spend time doing setup for each client, in order to make sure the search is performed in a way that Google recognizes it.

The big downside is you can’t identify where your searches are coming from. This is a big problem, since I believe that Google is going to continue to integrate localized suggestions in the future, even more so than it already has.

We’ve seen how local search results have become much more local in the last 12 months, I suspect that highly location-specific autocomplete is not far behind.

We use Mechanical Turk on a limited basis to help get search volume and a little diversity, but to rely on it as the only method for changing Google’s autocomplete would be risky.

I’d be interested in hearing feedback from others who’ve tested it to change Google suggestions. What kind of diversity did you get? Can anyone say whether the developer API allows you to geo-target or drip searches over time?

If you have negative suggestions in autocomplete and need help, please call 503-890-6663 for a quote to get it fixed.



Related Searches Appearing At The TOP of Search Results

In a few rare cases, I’ve seen Related Searches appear ABOVE the organic search results. Generally, we expect to see them at the bottom, but in doing some research for a client today, I noticed 3 related searches to “foreclosure attorney” appearing above the SERPs.

related searches appearing at the top of SERPs


I’ve also seen this happen one time with an autocomplete client I consulted with. In their case, two related searches appeared right after they started their campaign and they are still there – 6 months later.

Why do Related Searches Appear Above Organic Search Results?

Great question. In the only case where I witnessed this happen, I believe it was because the client did too many searches with the client’s name all at once. This caused Google to pull a couple of the related searches from the pool at the bottom and place them at the top.

They were trying to influence autocomplete and they just hit it too hard. But that’s just a guess.

You might also enjoy reading our case study on how we fixed Searches Related To for a recent client. Call 503-890-6663 if you would like to learn more.

How Does Google Autocomplete Work?

Great question.

And for now, it’s one that Google appears to give a somewhat clear answer on. The suggestions we see in autocomplete come from two things:

  1. Content
  2. Search activity

how google autocomplete works

The screenshot above is from

What Google is telling us is that if there is no content relevant to your search activity, then the algorithm is less likely to update it’s suggestions for the searches you’re performing.

For example, if we were searching for the term “Joe Smith Lollipops,” it would help if Google had indexed pages with “Joe,” “Smith,” and “Lollipops” on them. In my opinion, it would help even more if all three of those words were in the title of at least one web page, as the title of a page is a strong indicator to Google what that page is about. Without relevant content, it’s less likely that Autocomplete will change.

So, looking at this from a reputation managment perspective, if you want to push out a negative suggestion, you’re going to need both search activity and content that is both about the same topic.

Creating Relevant Content For Autocomplete

For clients who don’t have a strong presence online, it can be difficult to come up with enough unique search terms to make your campaign successful. Here’s how you can create the content yourself (or we can do it for you for an additional fee).

From our example above, if Joe Smith wanted to append the word “lollipops” as a suggestion in autocomplete, I would recommend creating at least one great web page all about “Joe Smith Lollipops.”

The easiest way to do this is to setup a blog using free software at WordPress, Blogger (need to be logged into your gmail account), or Weebly. I’d claim and then I’d write 500 words of unique content about Joe’s love for lollipops or how Joe always gave his grandchildren lollipops when they visited, or whatever. It’s not rocket science, it just has to be unique.

This is sort of a silly idea of how you could go about creating content that’s going to be relevant to your searches, but I use it to demonstrate my point. This isn’t to say that you can change Google autocomplete WITHOUT content; maybe you can. I’ve seen other people do it. I just haven’t invested the time to do it myself.

How Did The Negative Suggestions Get There In The First Place?

Since Google is telling us CONTENT + SEARCHES = SUGGESTIONS, it begs the question, “If no one did any searches, how did the negative term get into autocomplete in the first place?”

I’ve seen many cases where clients have had ZERO search volume for their negative term and yet, it appears as a suggestion. In my case study that compares what we see in autocomplete and how it compares to the data in Google’s Keyword Planner, I make the conclusion that “either the keyword planner data that reports search activity is wrong,” or “content alone can cause a new suggestion to first appear in autocomplete.”

My personal feeling is if you have low search volume for your name or business name, it’s much easier for new content alone to influence the suggestions about you. Names and businesses with higher volume are going to adhere more closely to Google’s indication that content plus search activity equals autocomplete suggestions.

We could theorize this all day long, but in the end, all you want is for the negative stuff to go away, regardless of how it got there. And that’s what my team and I help you do.


I believe search activity is more important than content, when it comes to Google autocomplete. I’ve seen mulitple cases like the goofy example above using “Joe Smith Lollipops,” where one, well-written, unique, optimized piece of content combined with adequate search activity has caused autocomplete to include the term(s) we were searching for.

In my case study about a client who’s negative suggestion and home city started with the same two letters, we had to create a series of blogs for terms beginning with “Du—“. No matter how silly those blog titles were, once we performed the searches and accumulated enough volume, Autocomplete updated to include them as suggestions next to the client’s name.

Google is constantly updating it’s algorithm to modify what we see in it’s organic search results and I suspect they may one day modify how their autocomplete algorithm displays suggestions. Let’s hope they tell us when/if they do. For now, we’ve got a decent understanding of how it works, and that makes it fairly straightforward how we can change it.

If you have questions about the suggestions that appear next to your name or business name, give me a call at 503-890-6663. Your consultation will be confidential and I’ll give you an idea of what it will cost and how long it will take to change the suggestions next to your name.




Does Location Matter To Google Autocomplete?

You bet it does! As Google becomes more localized in it’s search results, the suggestions we see from one city to the next can also be different.

We make it a practice of checking a few different cities – including the one where you do business – to see if Google’s autofill is the same.

If the suggestions are the same in a variety of different cities, our work is easier. All we have to do is have our team perform the necessary searches, no matter where they are located.

If, however, the autocomplete results are different in your city, than they are in another city, than it’s probably going to take a number of localized searches to ensure your campaign is a success within your home city.

How To Check Google Autocomplete Suggestions In Different Cities

Thankfully, Google makes it as easy as 1-2-3 for us to see the suggestions from one city to the next. Follow the steps below the image.


1) Go to Google and under the search bar, click “Search tools” (#1 in the image).

Note – You won’t see this option if Google has just been launched and is in the center of your screen. You have to perform a dummy search and hit enter. Then the Google search box will move to the top of your screen and you’ll see the “Search tools” option.

2) Next, click the little downward arrow (#2 in the image).

3) Then enter your location (#3 in the image) and hit “Set.”

Now, you can see how your Google suggestions appear to users in that city. It’s becoming more common to see different suggestions between the cities where you do business and cities outside of your market.

For example, a real estate investor who flips houses in Los Angeles and San Diego may see that “scammer” is the 1st Google suggestion in those cities, but in Chicago, Boston, and Seattle, “scammer” is the 5th term. This is a clear indication that search volume may be different for the different cities, online content may be geo-targeted, and it’s going to require a specific campaign made up of search activity in Los Angeles and San Diego in order to change those local results.

I know it’s a bit confusing, but we work with you to make sure you get the results you want to see in the cities that are important to you and your business.

How This Effects Your Autocomplete Campaign

Our search team is spread out all over the world. We have pretty good coverage and we’re always growing; however, if your Google auto-suggestions are different in different cities, then we have to account for that.

If you wind up working with someone else besides me to help with your problem, make sure they are accounting for localized suggestions; otherwise, it’s possible your autocomplete problem could be fixed in every city EXCEPT the one where you live and do business.

Certainly, it’s more time consuming to target searches for you in a local market. It requires careful planning to ensure proper search volume AND proper location, in order to get the results you want to see in your home city. If needed, we ramp up with additional team members who live in your target city, to make sure you get the search activity where you need it and achieve the results you desire.

You can also join the effort by following our Autocomplete Fix – DIY instructions. Your family, friends, colleagues, and volunteers can aid our effort. Every little bit helps.

Can Autocomplete “Go Local” At Any Time?

Yes. I’ve seen at least two cases where when we began a campaign for a client, there autofill suggestions were the same in multiple cities. However, a couple months into the campaign, we noticed different suggestions in different cities.

Usually what happens is if a client lives or does a lot of business in a certain city (or cities), they are more prone to seeing different suggestions in those cities then they are in the rest of country.

We’re working on a case like this right now where we’ve fixed the client’s negative results in every city EXCEPT the two major cities where he does business. So, we’ve concentrated our efforts with people who live in those cities to ensure that that we build enough search activity to knock out the negative suggestion.

Certainly, “going local” creates much more work, particularly for high volume clients – as you need to have a diverse base of people within a city performing your positive searches on a regular basis. We’ve got good coverage in most major cities and we’re building our network each month, so that we can have greater control.

Have an autocomplete problem you need fixed? Give Mike a call at 503-890-6663 for a private consultation.

Google Autocomplete and Keyword Planner

In the case study below, I’ll examine the relationship between Google Autocomplete and how it correlates to it’s Keyword Planner data.


In the case below, we successfully helped a client remove the word “scam” as one of the suggestions when you type her name into Google.

Now that the case is in the rearview mirror, let’s see how the data we collected during our work matches up with what’s being reported by Google’s free Keyword Planner tool.

Please note that all names and keywords in the example below are changed to protect the confidentiality of the client; however, the numbers, dates, and all other facts are actual.

Chronology Of Our Work With The Client

April 3, 2013

Client first contacted us because her name was being autocompleted with “scam” when she typed “Jane Doe” into Google.

Jane doe scam

We did some keyword research and found that Google reported ZERO search volume for the keyword term “Jane Doe scam,” which makes us wonder why Google suggests it in the first place, but that’s another story.

With the client’s help, we assembled the following list of 17 positive search terms she had relevant existing content online for.

  1. Jane Doe chevy
  2. Jane Doe bookworm
  3. Jane Doe scrapbooking
  4. Jane Doe hathaway
  5. Jane Doe st mark’s school
  6. Jane Doe oxford university
  7. Jane Doe pigeons
  8. Jane Doe ABC Electric
  9. Jan Doe admin
  10. Jane Doe socializing
  11. Jane Doe time stamp
  12. Jane Doe UK
  13. Jane Doe elementary
  14. Jane Doe spades
  15. Jane Doe twitter
  16. Jane Doe blog
  17. Jane Doe worcester

April 4, 2013

We began performing searches for all of the positive terms listed above, aiming for about 20 searches per month.

April 30, 2013

After about 4 weeks, the term “Jane Doe scrapbooking” is added to the list of suggestions. “Jane Doe scam” is still one of the suggestions.

Jane doe scrapbooking

June 8, 2013

“Scam” is pushed out of the search results after about 2 months of performing positive searches. Of course the client was very happy with the result and after tapering down the searches to about 50% of our previous volume, we stopped searching altogether on July 31, 2013.

Jane doe google autocomplete fixed

October 24, 2013

After about 3 months of performing no searches at all, we decided to check Google to see what the Jane Doe autocomplete suggestions were. We see there are NO SUGGESTIONS now.

It appears as though Google’s autcomplete algorithm has recognized that there has been no search volume (at least by our team) over the last 3 months and the suggestions have been updated to reflect that.

Jane doe no autocomplete suggestions 3-4 months later

Matching Autocomplete Suggestions What Google Keyword Planner Data

October 27, 2013

Our next step was to take a look at the data provided by Google’s Keyword Planner. We ran a report with our original 17 positive search terms to see what Google would report for the search volume for each one. We also added “Jane Doe” and “Jane Doe scam” to the list.

Notice how Google reports search volume for:

  • Jane Doe chevy – 30 searches in July
  • Jane Doe admin – 20 searches in July
  • Jane Doe blog – 10 searches in July

There are ZERO reported searches for “Jane Doe scrapbooking,” even though we searched for that term as much as the other terms.

Keyword Planner Data, all countriesOct '12Nov '12Dec '12Jan '13Feb '13Mar '13Apr '13May '13Jun '13Jul '13Aug '13Sep '13
Jane Doe90203020903202607030202010
Jane Doe chevy0000000003000
Jane Doe bookworm000000000000
Jane Doe scrapbooking000000000000
Jane Doe hathaway000000000000
Jane Doe st marks school000000000000
Jane Doe oxford university000000000000
Jane Doe pigeons000000000000
Jane Doe ABC Electric000000000000
Jane Doe admin0000000002000
Jane Doe socializing000000000000
Jane Doe time stamp000000000000
Jane Doe UK000000000000
Jane Doe elementary000000000000
Jane Doe spades000000000000
Jane Doe twitter000000000000
Jane Doe blog0000000001000
Jane Doe worcester000000000000
Jane Doe scam000000000000

We would have expected to see each term in the list above have 10, 20, or 30 searches per month in the months we actually performed the searches. But we didn’t. Instead, we see that only 3 of the 17 terms we searched show any volume at all. And even then, the volume does not match up with the months in which the actual work was completed.


From a small case study like the one described above, we can glean lots of insights into Google Autocomplete and Google Keyword Planner

    1. As you can see from the chart above, the search volume being reported by Google’s Keyword Planner is far different from what we would have expected to see. Not only is the volume incorrect, but so is the timing. Thus, it appears as though the data in the keyword planner should only be used as a guide, and not as fact.
    2. There is no search volume for the term “Jane Doe scam,” going back through October 2012, which begs the question, “How did ‘scam’ get their in the first place?” This either confirms our assumption in point #1 above or it could indicate that Google uses content signals to form autocomplete suggestions.
    3. After two months of performing positive searches in Google, Jane Doe’s autocomplete problem was fixed. After about 3 months of performing ZERO searches, all suggestions were gone. For keywords with low search volume, it appears to take between 2-3 months for the suggestions to change. 
    4. As overall search volume disappears, so do the suggestions. After 3 months of performing no searches, we noticed that when we type the client’s name into Google now, no suggestions appear. It’s reasonable to expect “Jane Doe scam” could appear again as a suggestion, just like it did in the first place. We’ll be sure to check on it in the future, but for the time being, our assumption is that once autocomplete is fixed, some ongoing maintenance might be a good idea as insurance against the negative term(s) re-appearing.

Have an autocomplete problem you need fixed? Give Mike a call at 503-890-6663 for a confidential consultation.