Google May 2020 update explained, 400+ sites checked, 270+ human manual hours invested

Aim100IM

Jr. VIP
Jr. VIP
Joined
Nov 6, 2010
Messages
3,174
Reaction score
961
Website
www.35DollarSeo.com
Google did a major broad update on 4th May and many of you have certainly seen huge rank shuffle in you rankings.

Since 7th may - I (and 3 of my other team members) have put a lot of hours into analyzing and understanding this google update and we have many takeaways.


WHAT EXACTLY DID WE DO? -
We checked more than 400 sites. We checked a lot of things about these 400 sites. (We actually checked 1000+ websites but approx. 400 websites were checked in huge detail. This article is a work of 270+ human manual hours.

We sorted all sites in different 3 categories.
Category A: Sites those saw rank loss ~after the update. Let's call these LOOSER SITES.
Category B: Sites those had almost same rank ~after the update. (Hardly 5-10 rank PlusMinus OR sites those saw rank loss for some keywords but not for all.)
Category C: Sites that were benefited from the update. Let's call these WINNER SITES.

We cross-checked all these sites again on 22nd May (bcoz Google takes some time to roll out its updates completely) and then we picked sites accordingly and did cross-check things again so that we can reach to a clearer conclusion.


How does Google rank a website? -( understanding the BALANCE) -
This paragraph has nothing to do with the work we did, but reading this paragraph will make it easier to understand the complete article.
Google never ranks a website on the basis of a SINGLE THING. Google always checks many things while deciding if they should rank your website.


For example:
For your niche- You can see a site ranking 1- which has ZERO social media votes.
OR
For your niche - You can see a website ranking 1- which has only 5 pages on it.
OR
For your niche- you can see a website- which has only 10 links.

but that does not mean that you should also make ZERO social votes and do only 5 pages and have only 10 links to rank.


If you get some social media votes then your chances of ranking will increase.
If you have a lot of articles on your website- then your chances of ranking will increase.
If you have a lot of good links- then your chances of ranking will increase.....


SEO is about giving your website POSITIVE POINTS from many aspects. Be it social media. Be it links or any other thing.
There is no single thing- which did cost you your rank drop. There are many factors involved and I will try to cover many of those.



Let's BEGIN!

1. Links to homepage VS links to inner pages:
Most of the winner sites have average 30-50% backlinks to homepage And 50-70% backlinks to inner pages.
Most of the looser sites have a higher percentage of backlinks to HOMEPAGE and a lesser percentage of links to inner pages.

POINT TO NOTE: Check your website links percentage. Make sure your inner pages backlink count is MORE THAN your homepage backlinks count.

2. Exact brand anchors percentage:
Most of the WINNER SITES have close to 30% exact brand anchors.
Most of the looser sites have close to 13% exact brand anchors.

EXACT Brand anchor example:
If your site is Football. com then brand anchors are football, foot ball.
We did not calculate BRAND ++ keywords IE we did not count "post title + brand anchors" (or other variation of brand anchors) as brand anchors (otherwise brand anchors percentage would be way too higher).

POINT TO NOTE: Make sure to have a lot of EXACT broad anchors.


3. Generic anchors percentage:
Most of the WINNER SITES had close to 10% generic anchors.
Most of the looser sites had close to 5% generic anchors.

POINT TO NOTE: Make sure that you have close to 10% generic anchors.


4. Social media votes:

Most of the winner sites had close to 2000 social votes.
Half of the LOOSER sites had close to ZERO social media votes.
Other half LOOSER SITES had way tooo high social media votes (10k+ average).

We checked LOOSER SITES social media votes manually and it was a clear sign of SOCIAL MEDIA SPAMMING.
Their social media votes were on very spammy profiles and looked 100% spam.

We also checked WINNER SITES social media votes manually and it was clear- that their social media votes were coming from half-decent profiles ~ definitely not as spammy as LOOSER WEBSITES social media votes.

Most of the winner sites- social media votes- were spread on many different pages/posts of their website.
Most of the looser sites- social media votes- were targeted on just a few pages/posts of their website and it looked spammy too.

Google did a great job understanding social media presence of a website and you should not take this lightly now.


POINT TO NOTE: Get social media votes from half-decent profiles atleast. Do not get social media votes from very spammy social profiles.
Get social media votes slowly and never buy thousands of social votes in a single shot.
Get social media votes for different many pages of your website and never get those for just a few urls.
Most important- your social media presence should not look too spammy. It should make sense. Try to get votes for pages/posts of your urls- those you think- are good enough to get shared on social websites.




5. Which social sites are preferred by google:
Most of the winner sites had social media votes from -
A. Fb (likes, comments, share) (close to 20%-50% of total votes)
B. Pinterest (Close to 20-50% of total votes)
C. GetPocket, Folkd, Bitly, VK, Buffer, Reddit

Point to note:
Get social media votes from facebook and pinterest mainly.
After these 2- try to get votes from GetPocket, Folkd, Bitly, VK, Buffer and Reddit.
If I be you- I will add some twitter votes too.



6. Backlink niche relevancy:
Most of the winner sites had many NICHE RELATED backlinks. Many unrelated niche backlinks were posted on the right "category" of a backlinking website.

Most of the looser sites had fewer NICHE RELATED backlinks. Their backlinks were also posted on WRONG "category" of the website.

POINT TO NOTE: Get links from niche relevant websites. If you are getting links from unrelated sites- make sure to choose RIGHT CATEGORY/TAG as much as possible.



7. Total Article count on a website (on page):
We checked 400 websites in detail but we also checked 600 other websites- just for finding this next line (and 1 more line you will read later)-

In a single/same niche- we checked top 10 competitors of a winner site and we found that - a winner site had almost 3 times more articles than a looser website.

POINT TO NOTE: Make sure to have more articles than most of your top 10 competitors (avoid authority competitor sites obviously as this is close to impossible to have more articles than them).


8. Word count for your articles? (On page):
Most of the winner websites had an average 1000 words long articles on their website.
Most of the looser websites had average 700 words long articles on their website.


POINT TO NOTE: Make sure that you have average 1000 words long articles on your website.


9. Anchors used?:
For most of the winner sites-> when we looked at different pages of a website- we found that each page has atleast 5 different anchors used.

For Most of the looser sites-> when we looked at different pages of a website- we found that each page has only 2-3 different anchors pointing to the page only.


POINT TO NOTE: While making backlinks- make sure that most of your website pages- have atleast 5 different anchors pointing to them.

10. User time spent on a website:
Most of the winner site's users spent average 53 seconds on the website.
Most of the looser site's users spent average 1.49 minutes on the website.

POINT TO NOTE: Nothing.

11. Video backlinks:
GOogle gives a lot of weightage to video backlinks now.
Most of the winner sites have a lot of video backlinks pointing to them.
Most of the looser sites have either ZERO OR very low count of video backlinks pointing to them.

POINT TO NOTE: Go make some youtube videos and get link to your website from there.
ALSO get video backlinks from other video sites such as dailymotion and more.



12: Relevancy of a website (ON PAGE):
Most of the winner websites had LESSER count of BROAD ARTICLES on them.
Most of the looser websites had HIGHER count of BROAD articles on them.
Broad anchors example: If your website is football .com then a possible broad anchor is: FUN ACTIVITIES OR Sports.

It looks like- google now cares a lot about the relevancy of a website ~on page wise.

POINT TO NOTE: You need to stick to "ONLY AND ONLY YOUR NICHE" more now..... Don't do a lot of broad articles on your website now.


13. LSI keywords used for backlinks:
We checked 400 websites in detail but for this step- we checked another 600.
Winner sites had "more backlinks using LSI keywords" then looser websites (we found this after checking top 10 competitors of a winner website).

POINT TO NOTE: LSI keywords. Go get more of those.



Those are all the findings- we were able to make.



How to re-gain your ranking:
In most of the cases- A single factor is never responsible for giving you a rank drop or rank gain.
It is always decided by many factors. If you have lost your ranking then you should cross-check all the points written above and get the ones- you don't have currently.


PS: We used "MOST OF THE" sites word many times in the article. By "MOST OF THE" we mean 40-80% websites (bcoz as written above- a single factor can not decide a website rank drop or gain).


PPS: We checked most of the mid-size OR low sized websites only. What we have written above- is definitely not about huge authority websites that have millions of articles on them.



I hope you find this article useful.
Thank you for reading!
 

Michael44Z

Registered Member
Joined
Jan 11, 2018
Messages
84
Reaction score
28
Interesting info, thanks for putting this together.
 

safeSEOboost

Jr. VIP
Jr. VIP
Joined
Mar 22, 2010
Messages
2,343
Reaction score
1,731
Website
bit.ly
@Aim100IM as you said it sounds like you put a lot of work into this, so thanks for the effort.
Interesting findings, appreciated.
 

rafark

Jr. VIP
Jr. VIP
Joined
Jan 15, 2013
Messages
2,406
Reaction score
1,464
Website
bit.ly
Nothing new under the roof, it seems. Thanks for sharing your findings with us. Did you manage to check the keyword density of the winners and losers?
 

Akash Mondal

Senior Member
Joined
Apr 29, 2018
Messages
993
Reaction score
213
Very nice read. Packed with info.
Thanks for all the hard works.
 

splishsplash

Jr. Executive VIP
Jr. VIP
Joined
Oct 9, 2013
Messages
1,868
Reaction score
5,794
Website
wolfofblogstreet.com
Hey buddy,

The data you've collected here is really good, but analysis like this isn't usually statistically valid. It's impossible to do this by eye-balling. You need to do some math to see if the correlation or difference is statistically significant. It really might seem like your conclusions are true, but for something to be statistically significant, then it has to have a probability of randomly happening of less than 5%. So if you have "most" sites in the winners group having more natural social shares, then it could be that in your sample set, the probability of the number of winners having more natural social media shares occurring in a random sample is 8%. This might seem low to you, as a human.. Which it is. Only 8% of the time we'd expect to see these extreme results, so it's not very common, therefore it must be the cause, right? But sadly no. It has to be < 5%. Not even 5.1%.

Let me give you an example using heads or tails coin flips.

A flip of a coin has a 50% chance of being heads. We would expect out of 40 tosses to get 20 heads and 20 tails. But what if I told you I had a biased coin. A magic coin. And I flipped it 40 times, and 28 of the times it was heads, and 12 it was tails. Amazing. But the probability of that happening is 6.8%. Put another way. If this scammer went into 14 pubs trying to sell this coin, then one in 1 of them you could expect a toss of 28 heads and 12 tails. Not that impressive is it? But to the human mind 28/12 is a "wow! we are onto something here. a pattern"

This is why we set p=0.05. 5%. 1 in 20.
Some studies will set p=0.01, 1% or 1 in 100.

So with that in mind, you really can't conclude any sort of meaningful patterns based on eye-balling data. If you did this test on 20 more sets you would get different results. There's probabilities each time you do it that you get a certain result, because you're taking a sample, not the total population. And that sample is tiny compared with the population.

Let's look at how we could maybe turn this into something more useful and find some real conclusions from all that hard work..

Have you classified all this data in spreadsheets? If so, I could work with you to do some proper analysis on the data. I'm doing a study where I'm asking people to submit their own sites so I can ask more detailed questions, but if you've got data for all this below we can definitely do some very useful analyses.

So, we have multiple data points here. But the interactions between them are potentially complex.

First, you've eye-balled differences in groups(These are not actually correlations, you've done an eyeball version of a difference in groups test rather than a correlation test)

Your dependent variable here is a categorical variable, winner, neutral and loser. This will be referred to as the DV.

Now, you aren't looking for a correlation between, say brand anchors, and something else. A correlation is only appropriate for measuring associations between ordinal/continuous data. Ordinal is ranks, like google rank, and continuous would be things like distance, speed.

You're looking for a differences in your groups, so is there a difference between sites in 3 types of social share groups, none, natural or fake.

First, we would want to test these all mathematically, individually, one at a time and see if there's a statistically significant difference in the groups.

That's done with a chi squared test for homogeneity.

Then you would want to test to see if the independent variables are really independent or correlated with each other. Ie, maybe more brand anchors is correlated positively with more natural social shares, so is it really social shares that are responsible for the different group, or is it the increase in brand anchors.

pm me if you want to have a chat about it and share the data for further analysis.




Google did a major broad update on 4th May and many of you have certainly seen huge rank shuffle in you rankings.

Since 7th may - I (and 3 of my other team members) have put a lot of hours into analyzing and understanding this google update and we have many takeaways.


WHAT EXACTLY DID WE DO? -
We checked more than 400 sites. We checked a lot of things about these 400 sites. (We actually checked 1000+ websites but approx. 400 websites were checked in huge detail. This article is a work of 270+ human manual hours.

We sorted all sites in different 3 categories.
Category A: Sites those saw rank loss ~after the update. Let's call these LOOSER SITES.
Category B: Sites those had almost same rank ~after the update. (Hardly 5-10 rank PlusMinus OR sites those saw rank loss for some keywords but not for all.)
Category C: Sites that were benefited from the update. Let's call these WINNER SITES.

We cross-checked all these sites again on 22nd May (bcoz Google takes some time to roll out its updates completely) and then we picked sites accordingly and did cross-check things again so that we can reach to a clearer conclusion.


How does Google rank a website? -( understanding the BALANCE) -
This paragraph has nothing to do with the work we did, but reading this paragraph will make it easier to understand the complete article.
Google never ranks a website on the basis of a SINGLE THING. Google always checks many things while deciding if they should rank your website.


For example:
For your niche- You can see a site ranking 1- which has ZERO social media votes.
OR
For your niche - You can see a website ranking 1- which has only 5 pages on it.
OR
For your niche- you can see a website- which has only 10 links.

but that does not mean that you should also make ZERO social votes and do only 5 pages and have only 10 links to rank.


If you get some social media votes then your chances of ranking will increase.
If you have a lot of articles on your website- then your chances of ranking will increase.
If you have a lot of good links- then your chances of ranking will increase.....


SEO is about giving your website POSITIVE POINTS from many aspects. Be it social media. Be it links or any other thing.
There is no single thing- which did cost you your rank drop. There are many factors involved and I will try to cover many of those.



Let's BEGIN!

1. Links to homepage VS links to inner pages:
Most of the winner sites have average 30-50% backlinks to homepage And 50-70% backlinks to inner pages.
Most of the looser sites have a higher percentage of backlinks to HOMEPAGE and a lesser percentage of links to inner pages.

POINT TO NOTE: Check your website links percentage. Make sure your inner pages backlink count is MORE THAN your homepage backlinks count.

2. Exact brand anchors percentage:
Most of the WINNER SITES have close to 30% exact brand anchors.
Most of the looser sites have close to 13% exact brand anchors.

EXACT Brand anchor example:
If your site is Football. com then brand anchors are football, foot ball.
We did not calculate BRAND ++ keywords IE we did not count "post title + brand anchors" (or other variation of brand anchors) as brand anchors (otherwise brand anchors percentage would be way too higher).

POINT TO NOTE: Make sure to have a lot of EXACT broad anchors.


3. Generic anchors percentage:
Most of the WINNER SITES had close to 10% generic anchors.
Most of the looser sites had close to 5% generic anchors.

POINT TO NOTE: Make sure that you have close to 10% generic anchors.


4. Social media votes:

Most of the winner sites had close to 2000 social votes.
Half of the LOOSER sites had close to ZERO social media votes.
Other half LOOSER SITES had way tooo high social media votes (10k+ average).

We checked LOOSER SITES social media votes manually and it was a clear sign of SOCIAL MEDIA SPAMMING.
Their social media votes were on very spammy profiles and looked 100% spam.

We also checked WINNER SITES social media votes manually and it was clear- that their social media votes were coming from half-decent profiles ~ definitely not as spammy as LOOSER WEBSITES social media votes.

Most of the winner sites- social media votes- were spread on many different pages/posts of their website.
Most of the looser sites- social media votes- were targeted on just a few pages/posts of their website and it looked spammy too.

Google did a great job understanding social media presence of a website and you should not take this lightly now.


POINT TO NOTE: Get social media votes from half-decent profiles atleast. Do not get social media votes from very spammy social profiles.
Get social media votes slowly and never buy thousands of social votes in a single shot.
Get social media votes for different many pages of your website and never get those for just a few urls.
Most important- your social media presence should not look too spammy. It should make sense. Try to get votes for pages/posts of your urls- those you think- are good enough to get shared on social websites.




5. Which social sites are preferred by google:
Most of the winner sites had social media votes from -
A. Fb (likes, comments, share) (close to 20%-50% of total votes)
B. Pinterest (Close to 20-50% of total votes)
C. GetPocket, Folkd, Bitly, VK, Buffer, Reddit

Point to note:
Get social media votes from facebook and pinterest mainly.
After these 2- try to get votes from GetPocket, Folkd, Bitly, VK, Buffer and Reddit.
If I be you- I will add some twitter votes too.



6. Backlink niche relevancy:
Most of the winner sites had many NICHE RELATED backlinks. Many unrelated niche backlinks were posted on the right "category" of a backlinking website.

Most of the looser sites had fewer NICHE RELATED backlinks. Their backlinks were also posted on WRONG "category" of the website.

POINT TO NOTE: Get links from niche relevant websites. If you are getting links from unrelated sites- make sure to choose RIGHT CATEGORY/TAG as much as possible.



7. Total Article count on a website (on page):
We checked 400 websites in detail but we also checked 600 other websites- just for finding this next line (and 1 more line you will read later)-

In a single/same niche- we checked top 10 competitors of a winner site and we found that - a winner site had almost 3 times more articles than a looser website.

POINT TO NOTE: Make sure to have more articles than most of your top 10 competitors (avoid authority competitor sites obviously as this is close to impossible to have more articles than them).


8. Word count for your articles? (On page):
Most of the winner websites had an average 1000 words long articles on their website.
Most of the looser websites had average 700 words long articles on their website.


POINT TO NOTE: Make sure that you have average 1000 words long articles on your website.


9. Anchors used?:
For most of the winner sites-> when we looked at different pages of a website- we found that each page has atleast 5 different anchors used.

For Most of the looser sites-> when we looked at different pages of a website- we found that each page has only 2-3 different anchors pointing to the page only.


POINT TO NOTE: While making backlinks- make sure that most of your website pages- have atleast 5 different anchors pointing to them.

10. User time spent on a website:
Most of the winner site's users spent average 53 seconds on the website.
Most of the looser site's users spent average 1.49 minutes on the website.

POINT TO NOTE: Nothing.

11. Video backlinks:
GOogle gives a lot of weightage to video backlinks now.
Most of the winner sites have a lot of video backlinks pointing to them.
Most of the looser sites have either ZERO OR very low count of video backlinks pointing to them.

POINT TO NOTE: Go make some youtube videos and get link to your website from there.
ALSO get video backlinks from other video sites such as dailymotion and more.



12: Relevancy of a website (ON PAGE):
Most of the winner websites had LESSER count of BROAD ARTICLES on them.
Most of the looser websites had HIGHER count of BROAD articles on them.
Broad anchors example: If your website is football .com then a possible broad anchor is: FUN ACTIVITIES OR Sports.

It looks like- google now cares a lot about the relevancy of a website ~on page wise.

POINT TO NOTE: You need to stick to "ONLY AND ONLY YOUR NICHE" more now..... Don't do a lot of broad articles on your website now.


13. LSI keywords used for backlinks:
We checked 400 websites in detail but for this step- we checked another 600.
Winner sites had "more backlinks using LSI keywords" then looser websites (we found this after checking top 10 competitors of a winner website).

POINT TO NOTE: LSI keywords. Go get more of those.



Those are all the findings- we were able to make.



How to re-gain your ranking:
In most of the cases- A single factor is never responsible for giving you a rank drop or rank gain.
It is always decided by many factors. If you have lost your ranking then you should cross-check all the points written above and get the ones- you don't have currently.


PS: We used "MOST OF THE" sites word many times in the article. By "MOST OF THE" we mean 40-80% websites (bcoz as written above- a single factor can not decide a website rank drop or gain).


PPS: We checked most of the mid-size OR low sized websites only. What we have written above- is definitely not about huge authority websites that have millions of articles on them.



I hope you find this article useful.
Thank you for reading!
 

Starblazer

Jr. VIP
Jr. VIP
Joined
Feb 28, 2019
Messages
2,365
Reaction score
2,984
Totally agree with the number of posts point. I frequently find sites with a ton of posts as Google started considering them authoritative.
 

YujinTan

BANNED
Joined
Jan 7, 2018
Messages
4,975
Reaction score
958
guess we just suck thumbs? Even we change , another Google updates reserve back etc? who knows? Only Google themselves decide etc, they are boss.
 
Top