Kyle Roof


Source: SEOSignalsLab

Pick His Brain!

I’d like to introduce one of our members, Kyle Roof, for our next ‘Pick His Brain’ session and I want to thank him for the participation.

Kyle knows a thing or two about SEO testing as he regularly publishes his findings in a private community called SIA.

To prove that Google’s algorithm is a mathematical formula, he has ranked #1 spot on GMB and has beaten everyone in our 30-day Rank or Go Home challenge after almost 6 weeks.

He is also the co-founder of High Voltage SEO with over 50 clients and the co-creator and math behind Page Optimizer Pro.

If you have any questions related to SEO testing or running an agency, you can pick his brain.

Here are the rules.

1) I’ll let the thread go on until he asks me to stop. Theoretically, this thread can continue until the FaceBook stock value goes to zero.

2) Please, no snarky remarks. I will not tolerate any intentional negativity. We are here to learn from each other’s success and strategies.

3) Please do not PM him and bother him. If you have a private question, ask for his permission on this thread when appropriate.


Table of Contents

What are your biggest ‘must do’ tasks when trying to rank 1st in map packs?

Verified and optimized GMB, connect your Google properties with your site as best as possible, niche specific citations.

Where does one go about finding the best niche relevant citations?

One of the first things we’ll do is compare the backlink profiles of top competitors.

You’ll see the links that they share – more often than not, those links are citations.

If I’m not mistaken both Brightlocal and Whitespark will help you find these types of links.

You can also do it in majestic or seo profiler.

Not sure, I haven’t used to compare multiple sites at once.

When you say Google properties which specific ones are you referring to?

GMB, google+, youtube

What are the 3 things you look at when trying to rank?

1. The keyword cluster (the secondary terms you can win bu properly optimizing for a top level term)

2. Word count, keyword and variations frequecy and placement of my primary keyword

3. The number of linking domains I need.

Did you hire a Latin writer for the page you’ve ranked for the challenge?

My team at hvseo has the best Latin writers on the planet!

Seriously though, as mentioned previously.

We decided to do use lorem ipsum to prove the point that on page seo is math and placement of exact keywords and variations.

The rest of the words don’t matter.

What are some techniques you recommend for Lead Gen for a new SEO Agency?

Lead gen is super tough. Your best lead gen is your current clients.

You may need to take on some early clients at lower than you’d like at first to establish that referral base.

Can you share the case study about the GMB?

Just responded below on a general outline.

Verified GMB, optimize it, build niche citations.

What did you do to rank the GMB in the contest?

I got a verified GMB and then optimized it.

When I say I, I mean my team.

The optimization was done according to what a lot of the strong GMB guys in here like Jordan Pearce recommend.

We then did a competitor citation audit and built out niche specific citations.

What do you recommend someone to do to try and rank merchant pages on Goggle in affiliate marketing who just cant compete with existing brands who already rank for key terms without any backlinks to their merchant page specifically?

I would guess that their backlinks are their internal linking structure (if you aren’t seeing any backlinks to those pages).

Nevertheless, on page is winnable. Google is showing you the sites that it likes.

Pick out the ones that appear to be winning with their on page and tune your pages to theirs.

What are the formulas for link building?

On my team at hvseo, we have a full time link builder.

We start with topic often looking at topical trust flow to get a feel.

We’ll then try for minimum 3rd party metrics, e.g. DA 20.

However, we’ll still evaluate sites lower than that if they appear to be growing.

What tool do you use to determine where to place the keywords on a page?

Page optimizer pro. I’m obviously biased, but it will show you where to place your keywords and how many times to use them.

I then look to Cora for the rest of the tuning I need.

What is your technical SEO formula for dynamic sites?

Javascript sites are a real pain.

I would try to talk the client into static versions of their most important pages.

How would you go about ranking in local without an actual address?

I obtained a local mailing address to get the verified GMB postcard for the competition.

How u crack algo?

I conduct and publish 7-10 seo tests a month. Most of which are single variable.

I then use a tool like Cora from Ted Kubaitis to track algo and serp shifts.

You get Cora reports from Ted, or he advises you on the changes to make based on the Cora report. I looked at Cora and found it confusing.

I run Cora myself and my team and I tackle it.

For a service area business where the address is hidden on the GMB do you still need to add the full address that Google verified on your website and in any citations as part of your NAP or will just the abbreviated address shown in the public GMB be ok. I know having the exact address on your website and citations is vital when showing your full address on GMB but what about hidden service area businesses?

Not really sure.

It would seem that you’d want NAP consistency.

Which I think is the point of doing citations and putting the address on your site.

That said, proximity is the number 1 factor for maps.

You see business in the maps all the time that don’t even have a website.

So, maybe not helping yourself but it’s not a death sentence.

PWA or AMP which on you like most?

I don’t really play with either.

Probably good for news sites but I’m not sure if they benefit regular sites just yet.

How much time delay are you noticing for Google to process on page changes?

Definitely can happen quickly.

I wait for two cache dates before making new changes if I didn’t see any movement check out your site by doing

That will give you the last time google has taken a screenshot of your site

What would you advise to help push a site that ranks very well locally but just can’t get into the map pack. What kind of approach would you suggest?

I’m sure there are others more qualified to troubleshoot the maps.

But the first thing I’d check is that your GMB category is the same as those that are currently in the map pack.

After that I’d do a citation audit.

See what the competitors have that you don’t.

Do the anchor texts on internal links carry the same weight as the external links anchors? If we hit the anchors hard on internal links, could it cause an overoptimization?

I’ve found that 2-3 internal links are equal to 1 external link of the same value.

As such, I monitor my internal anchor text ratio with the same calculation.

What’s your SEO community, as I am interested in joining other resources?

SIA. I’m just a part of the community though I am the lead tester.

It’s a paid community so I’ll leave off the url.

Does google detect certain links belong to the navigation links like menu, footer, side bar just because of their duplication all over the site? or via html tag?

Google “sees” the page after the rendered DOM

Do u publish each test inside your SIA membership?


Are you like to use multiple schema in one site?

I’ll use business/organization schema on only one page.

On other target pages I’ll use schema for a product or service.

Are you use Organization schema for new site. if how much time Google takes to show?

For this competition site, the schema stars popped out in about a week, maybe a little less.

What service do you use to generate your schema for your clients sites?

I use a template. copy the schema that I used on this site and adapt as you need it.

RELATED:  Pick His Brain! with Simon Cornelius

Do you like to generate sitemap daily?

I let plugins handle that.

What was the winning factor behind your rhinoplasty website?

On page, all the way.

Literally no backlinks ? Not even web 2.0s?

Did a little with citations, and some T1s.

Nothing major.

The biggest push was on page.

Do you have a website where I can subscribe to learn from what you do?

Not really….there’s SIA which is a paid community.

You can also learn a lot just from the free credits from POP.

How would you go about ranking an ecommerce store with just products and categories as content?

Hopefully you are going up against other sites that are ranking just products and categories.

If so, you can use those competitor pages for your on page roadmap

Based on your ranking there were no backlinks at all beside the citations. On page optimization was your weapon to lend as #1 on Google Maps. Was the keyword density a factor to rank since the content was written in Latin?

It’s important to think of ranking signals independently.

If you like the term density, that is fine, you can also use term frequency.

Either way, they are implemented signal by signal.

I put the exact keyword phrase, and it’s variations a specific number of times in each signal.

Also, just for clarification as mentioned above, we did build out some T1s in addition to the citations.

I see those as citations really.

How do you monitor your internal links anchors?

Right now, by hand

What is T1’s? Titles with H1 headings?

Tier 1. sites that you feel are safe to link to your money site regardless.

The challenge for me is that it takes a while to get them “trusted” or validated – then it’s like rocket fuel for traffic and rankings. Do you have any suggestions to shorten that cycle or a methodology after hitting publish?

I wonder if you what you are seeing as trusted/validated is just the amount of crawls Google needs in order to get through all the content, index it and rank it.

Maybe a thought would be to try rolling out the content in smaller chunks / phases.

Just spitballing an idea.

What were your two biggest tactics to win the contest?

On page for organic. verified gmb + citations for the maps.

Would you say it takes far less content to rank on page one for local keywords like MyCity Roofing or MyCity Plumbers? It seems you would not need a 700 to 1000+ word article to rank on page one for those type of keywords because no one wants to read it; they just want to call for your service to solve their problem. Is that right? I could be wrong?

I target the competitors who are winning based on their on page for that target keyword and then will aim for a word count that is 10% +/- that average.

Do also you also think it would be far harder to rank organically locally with no real business address and no citations?

Organically, no. I don’t think so. Maps might be a different story.

1. Keywords in title, url, h2, and once in body of the article. 2. Anchor text optimization (tie post that is performing well to it) 3. Use some LSIs in the body of the content. 4. Use CogSEO content assistant to integrate words into the body of the article It works for me… Is this what you do too?

As a quick approach where you need to get content out quickly,

I’m sure that will work just fine. I might skip the h2 and use an h1.

For competitive keywords or terms where you have the time, budget, and inclination.

I think you’ll want to go with the scientific approach and figure out exactly.

How many times to your keyword and its variations in those signals that you identified as well as other signal locations.

You’ll get better results when you tune your optimization specifically to what google is rewarding for that particular keyword.

Best place to get citation cheaply?

In our agency we brightlocal and whitespark. Somewhere between $3-$5 per citation built

If a website has 2 languages (German+English) and the main keyword is actually written exactly the same in both languages. How do you rank web pages for both languages? Both languages are important to the client but their English version pops up in German search.

Hreflang might help you here. Worst case, consider moving the german page over to a subdomain (only if you aren’t risking rankings).

On a subdomain, you can claim the property in SC and designate the region.

Is there any relation between having a AdWords campaign and seo?

Just having one, no I don’t think so. I’ve seen plenty of sites drop in rank that have had robost adwords spend.

Do you even use web 2 for T1? Is it spun or readable? if so what is the math(percentage something) for density of anchor text that you want to rank for?

I build branded web 2.0s as part of any project. I pretty much use those like citations.

I will also build niche 2.0s from time to time, but that content will be unique.

Do you use these web 2 for your tier 1 linking directly to money site? Would you even recommend the usual linking for web 2 which is directly to the money site and link to authority sites as well?

Yes, the web 2.0s I mentioned above I use them to link directly to the money site.

These branded web2s. Are they getting content syndicated from the website via IFTTT? Or is it just some static content that sits there and adds no value.
2) What web2s are we talking about here?

I use them like citations so they have the businesses info on them.

I don’t use IFTTT. Tumblr,, blogspot, etc.

Can you show like a sample how you setup these web 2 to be citation-like setup?

It’s just NAP with a business description

Did you point the citations links to the homepage and do you think you would have got a ranking boost if you would have pointed all your on article content pages to the homepage instead of the inner page and built out the homepage instead?

No, in this case, I pointed them to the target page.

What are your tests with 301s and the best way to implement them?

I’ve proved that they work….that they will pass juice.

I don’t have a test proven best practice, but I do like setting up hosting and 301ing domains from there rather than the registrar.

Jose Ponce Jr. is a stud when it comes to this sort of thing.

Have you done any testing 301s from cloudflare without hosting?

That’s sweet. I haven’t played with that yet.

What is your favorite test result from the last 6 month’s?

Great question and tough to answer.

I’ve found two new ranking factors this year, but perhaps my favorite is proving (again) that LSI is a ranking factor.

LSI causes such flame wars in SEO. You never hear an argument that says “I have a test, let’s talk about it. ”

You only here LSI is “old” or someone can’t find a Google patent on it.

Those are nice, but not argument ending statements. It’s also important to properly define LSI (which is probably where most things go off the rails).

LSI are other (latent) words that come up when talking about something that gives the conversation meaning (semantics).

LSI are not variations of your keyword or synonyms or long tail related searches.

If your keyword was “Car” LSI might be engine, tires, Toyota, mph.

Those are words that come up in a conversation about cars to give meaning/context to the conversation.

Those words are a ranking factor.

Did/are you doing a case study on this? Especially interested in how you manage your internal link structure.

I’ve run several single variable tests.

These get posted in SIA.

So i tried to use page optimizer pro and for some reason it was unable to get information from my website at all. Any clue why this might be?

You mean it didn’t run? did you use http/https for all urls?

Did you fetch variations?

Does the domain of your target url match your page url?

If you did the above and it and it still won’t work, send me a screenshot of your settings.

Does page optimiser pro colate all your knowlage from Sia programmatically.

The knowledge that I have from tests is the foundation of the tool’s algorithm

Does the tool taje into consideration the length of the competitors articles? (For me not every long article beats short ones)

You’ll get the competitor average to consider

How do you go about testing? A lot of people report on tests but I have never seen someone teach how to test.

Happy to teach it. The concepts are fairly straightforward.

Where people get off the rails is trying to test multiple things at one time without realizing it.

If you search Kyle Roof seo, you’ll see my agency’s media page.

There’s a video on there from a couple of years ago that explains the basics.

That was the first time I presented seo testing concepts.

The only thing to mention is that the video talks about doing two pages, that’s a simplified explanation.

You’ll want to set up at least 5 pages in your test environments.

Do you have any examples or case studies where you’ve significantly reduced content based on the suggestions in the report and then seen a ranking increase? I’m mainly asking about word count, not deoptimizing specific keyword placement. I ran a report and have a page with 2500 words and it is giving me a target of around 650. I know that just like with any tool, you have to use your judgement on what to actually change, but I’d really love to hear any stories of drastically reducing word count and ending up with a better ranking.

Usually if your word count is off by a lot, the rest of your signals are off as well.

RELATED:  Pick His Brain! with Gareth Daine

That said, many users have had success reducing word count, but you’ll want to tread very lightly….especially if you are ranking well already.

If you are ranking well, make tweaks one at a time and with word count, do it incrementally.

What are the main reasons you end up needing Cora?

Cora does correlational analysis on over 500 potential ranking factors.

POP doesn’t do that type of analysis and only provides recommendations on about 40 signals.

What is your 3 step strategy for speed optimization?

Keyword in: 1. meta title, 2. h1, 3. body copy.

Do you use guest posts to rank sites? And if so, what is necessary for the guest posts to really work (some guests posts don’t have any effect on some site’s ranking)?

I do use guest posts. We have a full time link builder on our team.

Look for cache dates on the page where you will get the link or a similar page to what you will get.

Make sure that those pages are getting crawled regularly.

When comparing content on the competitors site, what if it is an eccommerce site with 400 -500 words of content below the product images. Would I have to consider the number of times the keyword is found within the products on the page or just the content?

You’ll want to consider keyword placement and frequency in the places the impact rank regardless of competitor word count.

What do you like to put in your seo reports based on your typical offer for clients to convey maximum value?

In my agency we have worked a ton on reporting over the last several months.

We have packages based on the number of pages that we will optimize.

We then wanted reports that show the performance of those pages.

I think it’s imperative to get clients off of keyword reports and onto performance reports.

In a performance report you will monitor stronger KPIs like traffic and conversions.

We use datastudio and supermetrics to create the reports.

Do you no follow your outbound authority links on your main page?

Usually not. I would consider doing that if you have a bunch of outbound links.

I think it’s rare that I would have more than one or two on a page.

1. Do you use a standard wordpress theme or a modded one for your testing? your testing sites, do you use a dedicated ip or just a standard shared hosting?

Right now I have 20 or so active testing sites.

Most are wordpress, though a few are html.

Five of them are a stripped down version of wordpress.

In terms of hosting that’s also mix.

Most are on standard shared hosting across a few providers (godaddy, hostgator, greengeeks, a couple others) but I do have a few with dedicated ips.

Generally speaking the sites are set up based on something that is trying to be emulated for a particular test.

Do you have any tutorials or training on how you used your tool and cora to determine on-page?

I’m working on a tutorial now.

I could also put together a live session for Q&A.

Have you tested embedding YouTube videos, if it hurts/ helps rankings?

In 2015 I ran a test and found no impact for just having a video on a page. Obviously, that was a few years ago.

1. In the webinar you stressed the importance of choosing competitors with good onpage (very reasonable). How do you determine which sites are doing their OnPage right? A domain with strong backlinks can also have spot on OnPage, so it shouldn’t be discarded. Do you just discard those with thin content? 2. Do you use POP as a base for covering the most important factors and then fine tune with Cora? 3. What’s your stand on TF-IDF?

1. thin pages, sites that are probably winning due to other factors (e.g. youtube, yelp), sites that are pages you can’t build (e.g. directory pages).

A page can have both good on page and off so a strong backlink profile doesn’t disqualify a site.

As mentioned in the webby I also like to check the search in quotes to see which sites stay in the top 10 and if any moved up.

Those sites are probably strong in on page.

2. yes, that’s how I do it

3. Both POP and Cora are fans of signal specific TF.

What does POP output that Cora doesn’t? Not judging, just curious to know since I haven’t used Cora apart from having a look at two reports once and haven’t started using POP.

POP and Cora take different approaches to how the recommendations are come up with.

POP bases those recommendations on the specific competitors that you want to run against.

Cora is looking at all competitors.

Additionally, Cora is doing a complete correlational analysis of over 500 potential factors and pulls out the ones that appear to be moving the needle.

POP’s algorithm is looking for the critical min/max from 40 factors derived from specific analysis.

When you say “run them for difficulty,” does that mean competition?

I use the difficulty score from MOZ as a quick benchmark.

1. How does POP deal with plural/singular variations? I mean, probably (maybe you’ve tested this) Google treats “best strollers” and “best stroller” as the same term. But if I’m not wrong, POP will count those as different terms. Am I wrong? 2. What happens with “best strollers” and “best of strollers”? 3. When I fetched variations for a Spanish site in Spain, it just got one, which is weird as it’s a popular term. I tried with ‘EU’ (although there are many languages spoken in the EU) and ‘South America’ (thinking they mostly speak Spanish too). This isn’t a problem for me as I usually gather variations in a different way, but just curious.

For exact keyword counts pop will only consider exactly what you put into the tool.

Keep in mind it’s going to be showing you what your competitors are doing with that keyword.

In terms of “are these the same keyword” Google will show you that.

Do the search both ways and see if the serps change.

If they don’t, or pretty much don’t, then they are the same keyword.

I would guess that the opposite of what you go with as your main will show up as a variation.

If it doesn’t, you can add it in.

2. Those might be related.

Do the search for best strollers, scroll to the bottom of the serps and see if best of shows up.

If so, you can probably win it by properly optimizing for the primary keyword.

3. We’re working on a pick a language option now which might help pull more accurate variations.

That said, the variations come from Google.

They are identified in the SERPs so it’s possible there aren’t many variations.

Limited variations are especially common with one word keywords.

What would you “ must do” for local GMB optimisation. And what resources would you recommend for doing so.. coaching, guide or service provider.

GMB optimization and niche citations. .

For citations I use brightlocal and whitespark.

I haven’t viewed his new course but I go to Jordan Pearce with my GMB questions.

Where do you think they collect these documents? Can they be any of the web pages that got indexed or more specific and trustable sources?

I would think that the source is pages that are currently indexed for a specific phrase and ranking well according to strong on page signals.

When using pop I assume you leave out directory sites, so if there’s only a few actual competitors on page one do you include off page 2 or just use a small sample set?

As a part 2, if you’re already in position 2 or 3, are you trying to mimic 1 exactly or testing slightly more or less optimization to beat it?

Yes, if you don’t have a directory site, I would leave off directory pages.

Do your search in quotes. Take a long hard look at any sites that move up in ranking to page 1.

Those might be good to include into your competitor list as there’s a good chance their on page is strong.

Where can we get this awesome tool?

Whats the difference in terms of SEO, is there any bigger advantage of using pages??? And can I still get good result with post??…….. Also what is your opinion on SILO..

No difference between a page and a post.

I always set up silos for my target pages.

When using POP what is the best practice for choosing the right EXACT keyword? Do you just pick the one with the most searches?

You’ll want to pick out the primary keyword.

Do those searches above, do the SERP results change? They probably don’t change too much.

If they don’t, you might be able eye ball which version sites are optimizing for in their meta title and h1.

That’s the one I would go for.

Sometimes things are a bit scattered so plot them in a sheet and weight towards sites that are similar to yours.

If i’m using POP for ranking ecommerce, should my competitor links list for that term only be e-commerce sites that are ranking (as opposed to combining ecommerce and information sites together in the list)?

I would go with the ecommerce sites.


Leave a Reply

Your email address will not be published.