How to Run SEO Audits: In Search [Episode 70]
Don't forget, you can keep up with the In Search SEO Podcast by subscribing on iTunes or by following the podcast on SoundCloud!
May 12, 2020 |
The In Search SEO Podcast
How to Get the Most Out of Your SEO Audits: Summary of Episode #70
Matthew Tenney of GlassesUSA.com joins us to talk about how to run an SEO audit
- When and how to run the most efficient SEO audit possible
- From among the piles of data, how to know what to focus on when doing an SEO audit
- How to best qualify the data you get back from an SEO audit
Plus, we go into the May 2020 Core Update... its roll-out, its impact, its scope and size, etc!
With over a decade in online marketing and SEO, Matthew specialises in affiliate and e-commerce sites. He has developed innovative and data-driven audits focussed on on-page, off-page and technical issues. He currently works at GlassesUSA.com as Director of SEO.
GlassesUSA.com is the fastest growing, leading online eyewear retailer. Offering a variety of high-quality glasses in a wide-range of styles. Perfect for single vision glasses or sunglasses. GlassesUSA.com was built on the belief that purchasing eyewear shouldn’t break the bank and is on a mission to change the way consumers purchase eyewear.
Rank Risk Index
The January 2020 Core Update Data Results
Why Ranking Above The Fold Matters More Than Ever
Google is Suggesting Searches From Search History
LinkedIn Was Temporarily Off Google’s Index
Search Console Now Shows SpecialAnnouncement Schema Reports
Google Showing More Local News for COVID-19 Results
Google Released New Popular Products Report
Follow the podcast on Twitter
What Made the May 2020 Core Update Unique? [00:04:36 - 00:20:18]
No introduction is needed as we discuss the Google May 2020 Core Update. Let’s kick it with the basic facts, just in case you don’t know what just hit the SERP.
On May 4th, Google announced that they were releasing what they named the May 2020 Core Update. By the next day, our Rank Risk Index picked up elevated levels of rank fluctuations recording a risk level of 87. By the 6th fluctuations, levels hit a whopping 116 and the rest is history.
First off, this update caused a ton of uproar. How could Google release an update during COVID-19?! People were claiming that sites may not have their full teams available to deal with the drop-off in rankings and it’s not fair that Google hit them now. While Mordy understands their point, Google still needs to offer the best results, pandemic or not, which puts them in a tight spot.
Back to the update, it was slow to start. The levels of rank volatility were not core update worthy. They were high, but not legendary. At first it seemed that Google was taking it easy, with the pandemic going on, but they were not as this update turned into a beast. When Day 2 of the update came around, the fluctuation levels jumped like crazy, way up there even for a core update.
So how big was it in comparison to Google's last core update? Well, the January 2020 Core Update was super big, way up there in the annals of core update greatness. The May 2020 update edges it out at the top and bottom of the SERP. If you look at the top 3, 5, and 10 results, the May update slightly edges out the January 2020 update in terms of volatility increases.
There was, however, something very odd about this update. What’s odd is its treatment of YMYL sites at the top of the SERP when compared to what we usually see with core updates. Usually, niches like Health and Finance see heavier levels of volatility compared to other niches. That was not so much the case here as the YMYL and non-YMYL niches all showed relatively uniform volatility increases. What’s odd as it only happened once so far with these core updates and it wasn’t nearly to this extent of uniformity.
Mordy could speculate that with COVID-19 and how it’s changed the web and impacted content creation, what it means for user intent and user behavior, and what we’ll expect to see for entire niches that Google upped the ante for everything. Perhaps they’re really profiling all types of niches with extreme scrutiny because of it.
Of course, you would have to imagine that COVID-19 is playing a role here. The COVID-19 SERPs have been fluctuating like crazy the entire time as there were some unconfirmed updates during April that threw the COVID-19 SERPs into a tizzy.
Mordy did see Google playing with the local government sites on the COVID-19 SERPs where different state sites with different information on COVID-19 were going up and down. For example, Ohio went up in some cases and Florida and New York went down. Why? We’re not sure. Maybe it’s related to the extent COVID-19 was impacting that state at the moment.
Another interesting occurrence was when a lot of sites in the health niche that you would not expect to lose rankings did lose rankings. Sites like the CDC and Harvard's health site saw some rank reduction. Mordy did not look too deeply into this but he suspects that as the pandemic came in Google turned up the volume on some of these sites which makes sense, but as the pandemic wanes in certain places or as it becomes the new normal (as in the breaking nature of this all tones itself down) Google is turning that volume back down.
It’s all crazy stuff. Expect to hear more from Mordy in another post coming soon.
How to Properly Run SEO Audits: A Conversation with Matthew Tenney [00:20:18 - 00:52:54]
Welcome to another In Search SEO podcast interview session. Joining us today is a personal favorite of mine, a most gentle soul, one of the hidden gems of SEO. And today we are bringing him to the forefront. Please welcome the director of SEO at GlassesUSA.com, Matthew Tenny.
Thanks so much for the intro. Very nice.
Nice to have you. So I have to admit, I'm still feeling guilty because I have a really bad prescription and before we were chatting I ordered glasses online but I ordered from Zenni. Are you okay with that? We're still okay with doing the interview, right?
Yeah, I don’t know. I may have to cut it short.
So we're talking about site audits today and there's a ton to talk about. I want to first make sure everybody's on the same page. In case you're a newbie in the SEO industry, when we say site audit, what are we talking about?
You know, there are a lot of different site audits and I think it's good to kind of focus on one. Obviously, we can talk about a few. There's on-page, off-page, and technical. I bring a lot of experience with the on-page side. Obviously, this is critical for having good content and getting rewarded both from users and Google. In general, with a site audit, we're talking about looking at your entire site, looking at the health of it through different factors and measures, and checking out to see what needs to be fixed.
It’s a funny thing because people are so hyped up on whether it's a core algorithm update, ranking factors, links, or whatever it is. Look, Google has a certain budget so it could look very holistically, it could use machine learning to analyze your site, or it could just look at your headers.
Absolutely. Once you know the foundation of SEO then it’s up to figuring out how to optimize it which is the key.
With that, when you decide to do a site audit, what's the deciding factor telling you that you need to sit down right now and spend the next 18 hours doing a site audit?
It's funny that you say 18 hours. I used to work at a company that had over 60 SEOs in-house so I have a lot of experience to estimate how long these things take and 18 hours is fairly accurate. Even just for an analysis. Ultimately, it depends on the size of the website. The reason that people are going to do audits varies. Obviously, if you got it or had a loss of traffic you need to go in and figure out what you need to fix. This is a great way to analyze where you're at in terms of your benchmarks, in terms of what's not working, and what needs to be fixed.
Another opportunity is when somebody comes in, let's say an agency or somebody who is not familiar with a website that they worked on in the past, it's a great way to learn the website. When you go into the site from A to Z, you basically want to know what Google can find and how good the quality of what Google can find is.
Those are the two big ones that give you a starting point. The last one is potentially going through this experience and then building it into your routine. As an SEO, you're going to gain experience from doing the site audit and potentially, in the future, break it up into batches instead of doing a full site.
One reason why I'm a huge proponent of full site audits is that I've had experience and I've even read where you really get evaluated on your entire website. At some point, if enough pages are low quality, it's like poison. You're poisoning your entire website and you're potentially putting your entire website at risk. We're talking about either some sort of manual action for low-quality thin content, or you get an algorithmic hit and you get to the third page or fourth page. So the routine we're talking about, we're talking about an opportunity to learn the website and also for when you get hit.
Yeah, it's really interesting. Using a site audit is a great way to become more intimate with your website and to understand what's really happening. There are many times when I've looked at a site audit for our site and I see all these issues and you freak out, but then you start to realize that it’s not really an issue. There’s a reason why this page is like this and why that page is like that. When you run that first site audit can you really gain a new pervasive understanding of the site or do you have to do multiple audits, see how things are going, see why pages were created, what's it there for, how is it working, etc. so that you can really take that nuance look at what a site audit can offer you?
So you're asking if you can really understand the site in-depth and really understand what needs to be changed?
I’m talking about off-the-bat in the first site audit.
So there needs to be an understanding of the business, of the vertical that you're in, and of the content. Otherwise, the action items are going to be fairly on the surface and they're not going to get deep. I do believe once you get that, you establish an understanding of the business. Like anything, you need to know what the messages and the purpose is and what the goals are. Besides that, establish the benchmarks. Once you establish the benchmarks, then you can really have actionable items for that first site audit.
One tip that I do want to give is after one person goes through the audit, have somebody else review the audit. Sometimes somebody that's very familiar with the website and looks at it every day will get blinders. If they do an audit, great, they can come up with actions. But it's really important to share that audit and get additional feedback from somebody that hasn't seen this website for a long time or they're not as familiar with it and you get that unbiased view. That way you can really do some good checks and balances.
Yeah, that's a good point because there's an advantage to having somebody who's very familiar with the site look at the audit versus having some who was no clue what they're looking at.
You mentioned before about the idea of batches. I was wondering, at what point do you start to say, "You know what, okay, fine. I don't need to run the full audit on the site anymore. I can look at my links or I'll look at some content aspects or some content metrics...” How do you get to that point? How do you know when it’s batch time?
It's a good question. Most of the audits that I do are a full site audit and then it becomes more of a routine. After that first site audit, I call it more like a quick and dirty, the batches that you can do are top-performing pages and lowest performing pages which you check out every three months. Break it into numbers. The whole idea here is that in that first audit, you're likely going to identify useless pages, pages that don't bring value to users, pages that waste your budget (per se for Google), and that you need to get rid of. The goal is to identify what's causing those pages and fix those bad practices or unintended circumstances. Fix those for the future so you won't have them repeating and then you'll understand what batches you should do.
If you got hit manually or algorithmically then do a whole site audit. Don't do it in batches. Yes, it takes time but like I said, there are enough pages that can hold down the weight of the site where your entire website can get hit. At least you're going to identify patterns on those very low-quality pages that you need to start fixing in the future and prevent when you're ordering content or uploading or working on a new layout. So there are very big insights that you'll get and moving forward that you build into your routines.
You mentioned algorithm updates. I'm wondering, have one of your site's been hit by an update, you’ve done an audit, and then you found the problem?
Yeah. One of the best ways to learn is sometimes you go through mistakes. I’m somebody who has worked in companies that had purchased previous sites or purchased networks of sites where we sometimes inherited unhealthy assets. When that's the case, doing the audit was one of the best ways to identify the quality not being good. We're talking about metrics like time on page, bounce rate, and conversion rate. These are solid metrics that helped us evaluate the health of the site. Obviously, there are some metrics that we can get into a little bit more in this conversation but those are the highest quality ones that we were looking at that helped identify what to fix.
I do want to get into a deeper dive into the metrics in a second, but something you said kind of sparked my curiosity. You mentioned before you work with sites where you were acquiring assets and now you're working at Glasses USA where I'm assuming you're not acquiring any assets. What does the site audit look like now?
I think people who work in the e-commerce industry can probably identify with this. There are different industries and there are a ton of products. It's an opportunity to work with the product team, the development team, and operations and fine-tune what we want to communicate to Google and what we want from a user perspective. The audit is going to be complex. I haven't done it yet. I do have a plan and it's going to go in and clean up a lot of pages or remove pages.
Obviously, my rule of thumb is if it doesn't have traffic and if it doesn't have backlinks then goodbye or evaluate if it needs to be changed, but most likely not. Then pages that maybe don't have organic potential, no search volume, users, etc. It's not a high intent, search volume, or search worthy page. But from a user perspective, when they get to the website and they go through their funnel of purchasing glasses, it may be very valuable for them. So in that case, part of this audit is going to be a bit untraditional where I'm going to go through a lot of pages and either noindex them or maybe adjust the canonical to a similar page. It's a different type of audit because there are so many pages, legacy pages, and products. So I'm excited to get into it. There's a lot of cleanup to do. Look, until now, the company's done great organically and this extra optimization is just going to take it a step further.
Right and that could really be the difference between ranking five versus ranking three and that can be the difference between ranking above or below the fold.
Fingers crossed. Good luck to you. I don’t envy you. That'll probably take more than 18 hours. I can feel your dread coming through the electric wires that are connecting our two computers together.
It's all good and I always look at it as there's so much more potential. It’s exciting and I'm really glad that I got this ahead of me.
Pro tip. If you're out there and you're dreading the site audit, everybody is. It's not just you.
Yeah, ultimately think about the experience that you get and the result being a big boost in money. That's what I'm here for. That's what keeps me motivated.
Nice. So let's talk about metrics. You've run your site audit and now you have four bazillion different metrics to look at. How do you figure out where to start? Does it depend on the site? Where do you go? Because there are literally a million places you can start.
It's a good question. Ultimately know what your goals are and what makes the business tick. That should drive what the key metrics are. It’s going to be different from business to business. Ultimately, we're talking about a combination of analytics, rankings, and let's say some technical tool, whatever you like, to get some additional info. I like to get a whole story or almost the whole story of each page. What I mean by that is putting all these metrics together and first and foremost establishing a benchmark. Let's say as an example, bounce rate and conversion. What you can do is build a formula that will flag problematic pages. So you establish your benchmarks and say, "Let's say my bounce rate is X and the conversion rate I want is Y and I want to find any page that has a higher bounce rate and a lower conversion rate.” The formula exists out there. You can do this through a BI team or you can do this manually in Excel and boom, you got a flagged page.
Conversion rate can also be different types of conversions. Conversion rate could be registration, subscription, whatever it is, it doesn't have to be a purchase. First of all, make sure that you have that metric. Make sure that it's accurate. A tip that not everybody may be familiar with is that bounce rate can be affected significantly by events or something that causes an interaction. So technically you want to communicate to a developer or maybe your analyst to check out if events are being triggered as a non-interaction. If it's a non-interaction, then your bounce rate does not get affected. Basically, I want to recommend having a bounce rate that reflects somebody going to the page and leaving. A bounce rate can mean something different for other companies and businesses. From my experience, somebody got to the page and then they left right away without going to any other page. And in order to do that, you need to have clean analytics.
So that's one problematic page. Beyond that, we're talking about what's the rank of let’s say your main keyword for that page. Are you ranking at the top? Let's say you're ranking at the top and you're getting no traffic. What does that tell you? I'll ask you, Mordy. If you were in that situation.
If I'm ranking really well and there's no traffic? It's on my target audience.
Exactly. So you need to probably optimize either by going after a different audience/keyword or you need to scratch that idea as it’s a waste of your budget.
That's a definite vanity metric. I’m number one for this keyword that garners me no traffic.
This brings me to the topic of long-tail keywords. There's no room half the time for those unless they’re high intent. In other words, let's say you do a query for Parkinson's. You have like five or six sites that are going to be like WebMD, the Mayo Clinic, the Cleveland Clinic, etc. Then you're going to also have Harvard and Hopkins in there. Then you're going to have something like the Cancer Society of America. And maybe cancer.net, some non-organization website as well. There's no room for anything else unless you're going after some particular nuance longtail keyword for Parkinson's.
Okay, I'm done ranting.
So back to other metrics that you should make room for. Internal linking. I think that's nice. You want to know if that page has internal links and the reason for that and the last metric is backlinks or referring domains. I think this tells almost the whole story and it really helps with action items. Maybe the content is good, or the bounce rate and conversion rate with time on page are decent. It just needs a little boost with some PR or getting some more exposure on the site and giving it some nice internal links.
What are you looking for with internal links?
A few things. One is quantity and the other is going in a little bit deeper of the actual words that are being used. From my experience, you obviously want to do what's good for the users. More sites are using what I call "money words” linking to pages, rather than natural words. That's a little way to give a little boost and optimization by adjusting the words that are being used. Obviously, you don't want to go in and just adjust the specific words. You want to adjust the actual page and make the page valuable. Go in there and really give some value besides just having two or three words or three words that are linking.
What's your approach here? Let's say you're a car dealership and you're linking to a mechanic. You can link to the word ‘auto parts’ or you can link something a little more specific or a little more often called longtail. There have been these two approaches in SEO forever. Do you just go for that very exact keyword or do you give Google a bit more context with a little bit of a longer anchor text?
So we know now that Google understands the context and it can be long and exact. Ultimately, trying to manipulate things in order to gain rankings is a no-no. You want to do what works for the users within the context of the site. To group these and understand where you're at is to group the long tails and the exact into money. I’ll say this is money and the natural and domain is the other half of it. It's important as an SEO to know where you're at. Ultimately, you can try different things on your own website and instead of using the longtail you can try the exact and maybe it will be good from the user and from a search engine perspective.
How often do you change the text around just to get that anchor text right?
Not as often. It's more in these audits or if I'm pushing a specific page and I realized that the page has been buried. There's a lot of potential and value to that page so let's make sure that we give that page more exposure. We're talking about a header menu, sidebar, or within other articles that get some nice traffic or PR. That's a good method to make sure that that page will get an additional boost.
I have to ask a very generic question at this point and I'm sorry for doing it but I have to. What are the general mistakes that you see? I think you have to ask this question to everybody who you speak to, only specific to the topic, of course. What are the general mistakes you see people making when they're doing a site audit?
First of all, you want accurate stats. That's one. Two is general content where you check you’re being consistent with your site and your page. Maybe you have somebody in your company or you can get an expert that's good with UX to really help you see what you're doing right and wrong. In a general rule of thumb, there’s a UX perspective and then there’s the actual quality of the content. A lot of times SEOs take over sites, I don't know the percentage of SEOs, but I notice a lot. They're very text-heavy and it's really not necessary. Users expect and react to something much better when you're breaking up the text with graphical elements whether it be images, video, listicles, bullet points, whatever it is. I notice a lot of times that that's missing.
Headers are huge. It's great for breaking up the text. It's nice for users. Let’s say you have a video that you're hosting yourself that doesn't necessarily have a title within it if you're hosting from a third party. So putting a header above a video explaining what the user is about to see helps. The same idea if you have a table like a comparison table and adding a header above the table.
You want to be consistent. A lot of times I noticed inconsistencies from page to page and that's a lot of times a big insight on pages that are succeeding as opposed to the pages that are not succeeding. A lot of times they're missing those other things. Maybe that's because they’re legacy pages and the current SEO didn't work on those old pages, or they didn't have the right strategy in place, or you were doing A/B tests to gain insights and you only did those changes to your top pages and those neglected pages didn’t have that same treatment. I'm sure I'm missing some patterns. The other big thing is breaking up long text, good UX, and making sure that pages have internal links. That's SEO for you.
It’s true as I always think about text creation from an educational perspective. One of the accommodations in teaching which I used for kids with special needs is called chunking where you do a little bit with them, do something else, and then you come back to it. Images is a way to do that and one of the things that I do see in some of the bigger SEO sites is that the only use of images is for teaching purposes so it’ll have a graph or a chart in there. But it's okay to throw an image in there just as an eye break. Obviously, don’t throw any random image up in there as it has to relate obviously to the content. Don't just throw an image of your kids.
Think about the emotions that you get as well from the image. Images give off an emotion and tell a story so it's important.
Yeah, I spend way too much time picking out images. I try to find something a little bit offbeat and a little bit more creative.
Optimize It Or Disavow It
Your options are this, either a technical audit or a content-based audit. Which one do you do?
In general, I would go with content. I mean, we know even Google publicized the three most important SEO things: RankBrain, content, and links. In general, most sites and Google and search engines can really handle a lot of tech issues. But at least have somebody with some technical knowledge just go through and make sure that your robots txt, your disavow, your canonicals, and whatever are up to par.
Make sure it’s intact because without that, spending any time on on-page doesn't matter for organic value. Maybe for other marketing channels, but I would say having a really quick, short tech, just to make sure things are in place and then go with the on-page.
Okay. So that's definitive. That's not "It depends.” Make sure you're being crawled and then go to content.
Well, thank you for that. I appreciate you coming on the show.
Thank you. Anytime. This was great by the way.
SEO News [00:53:49 - 00:57:08]
Google is Suggesting Searches From Search History:
Google is running a new search suggestion that offers a search suggestion based on your recent search history!
LinkedIn Was Temporarily Off Google’s Index:
The www version of LinkedIn dropped out of the Google’s index for a bit. This had nothing to do with Google, but with something on LinkedIn’s side.
Search Console Now Shows SpecialAnnouncement Schema Reports:
Search Console now shows errors and performance metrics for the SpecialAnnouncement markup.
Google Showing More Local News for COVID-19 Results:
Google says it has gotten more local with the news around COVID-19 saying it has moved to show more local news sources. Google also said news results for its Top Stories carousel for COVID-19 queries need not be AMP!
Google Released New Popular Products Report:
Google has a new toy for product marketers. The Rising Retail Categories tool tells you which products are getting hotter and more popular.
Tune in next Tuesday for a new episode of The In Search SEO Podcast