Bill Slawski is President of SEO by the Sea and Director of SEO Research for Go Fish Digital
Bill has a Bachelor of Arts Degree in English from the University of Delaware, and a Juris Bill has a Bachelor of Arts Degree in English from the University of Delaware, and a Juris Doctor Degree from Widener University School of Law. Bill has unpicked the techniques, the state-of-the-art and the evolving technology emerging from search engines, especially Google, the last fifteen years.
Bill has helped many of us to understand better how search engines really work; and the impact of search on our daily lives. Bill fuses his passion for facts, science and art with his innate inquisitiveness.
Outside of search research, Bill is a supporter of environmental protection, wildlife sanctuary and native horticulture. Drawing energy from nature, a favourite pastime is to side-step the water splash as waves roll in on beaches near his home. Bill lives and works out of Carlsbad, just north of downtown San Diego, CA.
This post is part of a series called ‘Conversations in Search’. I discuss the current state of SEO practice with other SEO experts and discover their views on the future of SEO.
Damien: Hi Bill, how are you?
Bill: Yes, I’m well thank you. It’s been kind of chilly here in San Diego. We don’t usually get much rain, but we’ve made up for it the last week or so.
Damien: How are you and your family coping with COVID-19?
Bill: My family, a lot of them, are on the east coast in places like New Jersey, which are very hard hit by COVID-19, and some of my family members have lost friends. San Francisco was one of the first places that saw many people afflicted with COVID-19. So far San Diego hasn’t been hit too hard.
Damien: It’s great to chat with you and learn more about Bill the person, and of course your views on SEO. Tell us where you are now?
Bill: I’m in a townhouse in Carlsbad, which is part of North County, San Diego. It’s not the city itself; it’s more of a suburban area. I’m a relative newcomer to California, having moved here five years ago. I was not familiar with a lot of the native plants, and I really enjoy them. I didn’t know, for instance, that there are so many eucalyptus trees in San Diego.
They started planting them in the 1880s, thinking that they might be useful railroad ties, the boards to connect tracks. They planted the trees and tried making ties out of them but found out pretty quickly they weren’t perfect for railroad ties. It turns out; they didn’t hold position very well and got too loose when they were down. But there are lots of eucalyptus trees all over the place now.
San Diego is a semi-arid climate, and it’s kind of dry. It doesn’t rain a lot, and last week, we got lots of rain. Usually, when the weather forecasters talk about rain here, they say ‘you’re going to get 200th of an inch’, well we got about 6 inches of rain the last few days.
I’ve been buying plants from local nurseries, they have a section filled with California native plants, and I grow them on my back and my front patios. So it’s good seeing things like Hibiscus, Catalina cherry trees and Coffeeberry trees. I’ve got an orange tree and dwarf lemon tree on my back patio too.
I like having native plants because they are great at drawing in the native birds. I love nature. In the last couple of days, I’ve had hummingbirds birds above my head. They like the flowering plants. It’s kind of good to see as they’re not scared by my presence.
Damien: Where did you grow up, Bill? What stands out in your memory most from that time?
I grew up in New Jersey, which is known as the Golden State because it’s halfway between New York City and Philadelphia, and it’s been a source of produce for both cities. We had a big yard when I was growing up, and my parents had a garden that was 50 foot by 50 foot. We used to grow all kinds of fresh vegetables, and they spent lots of time in the garden, and it rubbed off.
I went to school in Delaware. We were in Central New Jersey, halfway between New York City in Philadelphia. We had a choice between New York news and Philadelphia news, but not New Jersey news because we had no comparable sized metropolitan area like New York and Philadelphia.
When I was young, we lived right by the beach, and I enjoyed that. We used to fish so much and go crabbing too. We were like a 10-minute drive from the ocean.
When I moved out here to California, that’s one thing I tried to make sure I experienced again. I am now just 10 minutes drive from the ocean. I can’t visit right now because they’ve closed the beach during this pandemic, but I like going to the beach and walking down the beach for 2, 3, 4 miles.
There’s a specific type of bird that likes going to the beach, called a plover. The plover will run up on the sand towards the water, and then run the opposite direction as the waves come in. I’m a bit like a plover myself with the waves getting a little bit too close, and trying to get away from them without getting soaked.
Damien: When you think back, who would you say had the most substantial influence on you and your moral compass?
Bill: Of course, my parents had a tremendous influence because they were strong people. My father was Robert Slawska, he was involved in the industrial plastics industry, and he was designing, building and selling plastics machinery.
Industrial plastics machinery company, Sterling Extruder, was where he’d design machines, and he helped to sell them too. He had a reputation for being very honest.
I remember when I was about 13, and he got a phone call from somebody, a potential client, and they asked if the machines he sold would make certain types of parts. He did not hesitate to tell them, ‘No, we don’t make those, you want to go with this other company to do this’.
They were surprised that he told them about a competitor. But they then told him every time they were planning on building new machines and asked him questions about the best solutions. I think the opportunity to be honest and to be open with potential customers made all the difference.
He was presented with a Lifetime Achievement Award, in the plastics industry a few years ago. A number of his competitors showed up on stage and spoke on his behalf, and told us he was a fierce competitor, but he made the industry better, and I always appreciated that.
My mother was Vilma Slawski, and she was a school teacher. I spent lots of time with her testing me with flashcards, and encouraging me to read widely and spend time at the library learning. I found myself doing a lot of that, and I was a real bookworm. I used to read a book a day throughout my time at high school.
Damien: Interestingly, your surname is different from your dad, may I ask how that is? Perhaps we can learn more about your family?
Bill: That’s a good story. It’s a Polish tradition for your last name to end with ‘ski’ if you’re a male, and ‘ska’ if you’re female. When my grandmother went to the hospital to have him, the nurse who admitted her was Polish, so that nurse wrote Slawska on her admission sheet. Then when he was born another nurse, not Polish, simply copied that Slawska name from the admission as his last name so my father ended up with his name misspelt on his birth certificate.
When he went to serve in the Marines, they told him ‘Hey, you can pick one and use just that one’. From then on, he decided to use Slawska for all his business affairs, but Slawski for his personal dealings.
My father was Polish Slovakian. His parents were first-generation Americans. He was second-generation; his grandparents were from Poland and Slovakia. My mother was Italian and Hungarian and was second-generation too. Her parents were both born in the US, and their parents were born in Italy and Hungary.
Damien: So it must have been a real treat for you to come over to Europe, did you get to reconnect with your roots?
Bill: It really was, I got to visit Milan a couple of years ago and last year, which is in northern Italy, and it’s not too far from Bobbio , which is where I know my great grandmother left Italy to come to us. You know it was funny; my father met my mother on a train.
He was in the Marines at that time, and he was dressed up in his uniform. His best friend was dressed up in his uniform too as they were travelling together. His best friend started talking to a girl on the train, and so my father started talking to her friend. My mother told my father that her name was Vilma.
My father said, ‘It’s funny, that’s my mother’s name and my sister’s name’. My mum thought he was lying to her because this is not a common name. But he was telling the truth, as was his trademark.
Damien: Bill, over the years, you’ve faced online bullies and trolls for some of the ideas you put forward. How do you maintain your calm?
Bill: Well, you treat everybody with decency, and if they don’t treat you with decency back, it’s their loss. It’s why I like to do as much research as I do. When somebody tells me that something is a certain way and I know after reading many thousand patents that they’re wrong, I tell them.
For example, somebody said all of the people at Google treat the main algorithm as if it’s PageRank. I’ve read over 1,000 patents from Google about different inventions and algorithms, and in only four or five of them do they call what they’re talking about PageRank. So the hundreds of other search engineers from Google don’t.
Damien: Where do the many myths and seemingly baseless theories about what constitutes good SEO practices come from?
Bill: I think it’s just a lack of understanding. I spend much time looking at patents, and there aren’t too many other people who do that. They’re great sources of information because the reason why they’re published isn’t to market the invention, it’s not to educate the public, it’s not to try to encourage excitement over Google as a search engine. It’s to protect the stuff they’re describing, the processes as intellectual property to keep other search engines from imitating them, from stealing what they’re working on.
Damien: You’ve been translating a language that can often be entirely foreign. Where did you learn that knack?
Bill: I did go to law school, so I am used to some of the legal jargon. You don’t spend much time studying the patents, but you do spend more time looking at some arcane stuff like Admiralty law for instance. In Admiralty law, they often decide cases based upon contracts that are sometimes hundreds of years old. They don’t change the contract language too much because it’s been litigated so frequently. Every line in an Admiralty contract has legislation behind it, multiple cases that have been decided upon single lines. So there’s specific jargon that carries a lot of meaning to it.
In patents, you’ll see phrases like ‘this patent is written for those learned in the art’ which means somebody who is involved in a particular type of research that the patent describes. They have probably come across prior research and have a sense of how to build what the patent describes. If you’re reading a patent about a search engine, how it works, they’re not going to tell you every little aspect of every little thing and how it works together. They’re going to assume that you have specific baseline knowledge for the techniques and prior work.
Damien: Your first career before SEO was very distinguished work in the Superior Court of Delaware. Tell us what memories stand out to you from that time, and what you hold onto?
Bill: Yes, it was challenging. I started working there as an intern to the staff attorney, and during that six-month job one of the administrators there asked me if I’d be interested in working as an administrator of the court, and I agreed to it. I ended up my job title was Assistant Deputy Criminal Prothonotary, which is a first clerk of the court. I’d supervise people in the criminal division, court clerks who went up into the courtroom or processed paperwork and filings for the court like bail paperwork or warrants, things like that. I had clerks who were going to court and doing sentencing orders for cases.
I would get phone calls, sometimes from prisoners. They would ask me to help them get released from prison because their case had already been litigated in the courts where they had pled guilty or were found guilty. The charges that they were prosecuted on in an indictment.
Those charges were the same as those a state police officer had initially charged them before they were indicted and those got entered into the computer system as one type of charge. Still, then they’d get indicted under a different charge.
That would mean they would have two sets of charges against them. The courts would take care of the charges that were in Superior Court work, but not the charges that were entered into the computer system of the state police. Sometimes they would get their case finished in court, and then get immediately arrested again for the earlier versions of the charges.
In those cases I would investigate and send paperwork to the prisons, showing that they’d had those charges taken care of, rather than have them go through a judge and do a writ of habeas corpus. I would end up helping people get released from prison when they were being held wrongly.
I remember after doing that job for several years, and then going into SEO, going to some job interviews and having people ask me, ‘How would you feel about supervising people in high-pressure situations?’ I’d say, ‘Well, I’ve supervised people who would get involved in decisions that might mean the life of a person, because with death sentence cases’, or another case might involve millions of dollars worth of damages, so I have no problem with high pressure.
Damien: I imagine the court and police computer systems were antiquated. Was this the time when IBM manufactured the mainframe?
Bill: Yeah, a little bit after that. My job title at that time was ‘Mini microcomputer network administrator’. I never did see a mini microcomputer, but that was my job.
Damien: So you decided at some point that SEO was a route you’d like to progress, what was the catalyst for your interest?
Bill: There were a few different ones. I was an English student in college, and I was doing lots of stuff with the deconstruction of literature, looking at how things are written and analysing the different aspects of them.
At some point, I remember a professor asked me a question about a piece of work that I thought was interesting, and I said ‘I noticed the author uses the word singular a lot’. The idea of looking at the frequency of usage of certain words was something I was doing as an English student.
When I got to law school, I was a teaching assistant for an environmental law professor — helping him update a paper that you’ve written about electronic database use to assess natural resource damages in legal cases.
He’d written about sources like LexisNexis and other electronic databases that were around before the world wide web. How people are using these electronic databases in a knowledge-based profession like the law was something that I had access to.
This was a little bit before Google. I ended up helping a couple of friends who were starting a business, create a website and promote that, which is what launched me into doing SEO.
A friend of mine was working as a service mechanic at a car dealership, and he hated his job and wanted to do something else. The business idea was to help people manage the technical requirements to incorporate and run the business in Delaware.
I was reading the book on how to set up incorporation in Delaware at the time, and having a mailing address in Delaware to receive service notices and process for people was the main technical requirement.
Effectively, working as the agent, you’d check their mail every day and see whether or not they’re being served. If they were, you’d notify them, and we figured out how to make this work as a website.
I went to a local bookstore and got a ‘Learn HTML in 2 Weeks’ book, and I built the website. I started reading all that I could about the web, on the internet, and I started reading a book by an author named Roger C. Parker, on web design and building eCommerce websites that was helpful.
I built the site for my friends Alan and Larry. Alan worked in the car dealership, and Larry was a friend who was a lawyer who was renting a house, which had an empty office space. Alan started working in that empty room, and that was the start of that business which did reasonably well.
Damien: Alan and Larry are happy with what you’ve done in producing this website for them. You’re still working with the court at the time, and your interest in building sites and how to promote them grew.
Bill: Yeah, thinking of promotion, one of the best links that I got at that point, was from a Polish classified website. We were more concerned about getting links that brought traffic to the site than links that brought rankings because there wasn’t a Google with PageRank attached to it.
That classified website link brought many customers from places like Latvia and Estonia, from companies that were taking cargo ships and sending them from Europe to South America. They were shipping new shipments to South America from Europe weekly, like seven or eight ships and every new ship would have a new corporation attached to it.
We had no reason to try to increase ranking. Alan found a currency converter website which had a currency converter calculator that you can put on your site. He translated the currency converter programme into something that can be placed on the page on this website and put it up. Soon, other eCommerce websites saw the converter and liked it. They didn’t imitate his effort at putting their currency converter on their website but linked to his.
Then, when Google started showing PageRank all of a sudden this currency converter page had a higher PageRank because of so many other eCommerce sites linked to this currency converter as something to use to translate their prices to another currency.
Damien: It sounds like you made a successful transition from working for the government then to freelancing and working for an agency?
Bill: Government work was very convenient. There’s lots of time off. There was no likelihood of being laid off because of a lack of business, I mean, you’re working for the criminal and civil courts for a state. The pay wasn’t as high as it could be, which is true for most government work. I was also working as an administrator at cre8asite forums at a time.
Loren Baker sent a note to the administrators of the cre8asite forums saying he’d like to post about a job that was available at an agency he was working with. He asked permission to post, and I said, ‘Sure, there’d be no problem. I might be interested in that, tell me more’.
So the job was about half an hour drive away from where I was then in Delaware. I went to an interview, got the job and joined up with them for a year. The chance to work for an agency exposed me to clients directly in different niches and topics and areas.
Damien: You started SEO by the Sea in 2005, take us back to the Chesapeake Bay in 2005.
Bill: When I came up with the name for SEO by the Sea, I was looking out the window of Web Ad.vantage, watching sail ships, bouncing up and down on the Chesapeake right where the Susquehanna empties into the Chesapeake Bay.
I had just come back from SES New York, a week or so earlier and I was thinking about starting up a conference. I had in mind creating a free type conference thing like BarCamps ended up being. All the people who were guests would sign up to teach something.
It ended up not turning out that well, it wasn’t heavily attended, but we got to take a tour of the Chesapeake Bay and skipjack which is a type of sailing ship that was used for oyster hunting.
The website promoting the ‘BarCamp’ type conference was a good idea. That started me with the website. It was a WordPress website, and it took me about two hours to put up, and I posted something every day for about a year.
After the conference was over, I thought I’ve got this website. I should do something with it, and I’ll keep on posting to it.
I started writing about patents. Loren Baker, who was my coworker at Web Ad.vantage, and he’s in charge of Search Engine Journal now, and Aaron Wall said nice things about my site that helped bring people to it. Search Engine Journal had a bunch of awards, and one of them was a technical SEO blog, and I ended up either tying or winning that one.
I’d developed a process for writing about new patents which were coming out of Google. It gave me a chance to learn new things and stuff that I could use with clients.
So I was doing a lot of stuff with posting on the forum as administrator creating site forums, about SEO with other people who were SEOs. I was still doing in-house SEO for the incorporation business still.
I started writing more, and I got lots of positive responses. I had opportunities to speak at conferences sent to me. I spoke at SES in New York. I got an offer to speak in Las Vegas at a show where they offered to pay for the flight and hotel. I didn’t do PubCon back then. I’d heard about Pubcon, and I ended up going to Pubcon in 2006 for the first time. The flight was a lot more affordable than I expected. I was able to stay at The Best Western for $43 a night.
Damien: Let’s fast forward to today. You’re a successful SEO with a depth of knowledge, and experience of interpreting patent filings from Google and other companies. What are you most proud of?
Bill: I’m most proud of helping other people learn, that makes a difference.
Damien: Your work has influenced the shape of our industry Bill. You’ve disseminated fact from fiction with your research into patents. As you’ve watched the web grow and evolve, what has shocked and surprised you the most?
Bill: Let’s see if we can put this into words. I see a lot fewer patents from Google coming out. I’m seeing them filing things that talk about technologies like neural networks or word vectors. The technology they describe is requiring more of an actual background in computer science. The idea behind an algorithm is something that’s a process to solve the problem.
Some of the stuff they’ve published as patents the past few years have been very explicitly detailed to address how they would solve problems with things. Some of the new stuff is like ‘Here’s this completely new technology we’ve developed that treats and understands the web differently.’
For instance, there was one, website representation vectors, where they said we’re going to use neural networks to look at lots of websites and try to identify features of those websites that are unique to the industry that they’re in.
There are certain things that each of those does that are unique.
If we have a health site that’s purely about medicine and health, like a CDC website, it’s going to have features that stand out. Other health-related websites which are written by laypeople, which may be more about, ‘how I dealt with fighting diabetes as a diabetic’, Doctors do not write them, but patients do. But they do attempt to solve problems for people, and they are written in a certain way. So the website representation vectors pattern, described how they would classify these websites based upon the niches that they fill and the amount of expertise behind the authors of the site.
It’s not the E-A-T that’s written about in the quality raters guide, but it’s similar in many ways. They talked about how our queries might reflect different needs for different levels of knowledge in various industries. Somebody wants to know about health-related supplements, and they’re going to be the best answer from somebody who is not a doctor.
Damien: Google has thousands of staff who are technically brilliant. What assumptions do you think they make in this new world of search?
Bill: There are people at Google who study why people search. One of them is named Daniel Russell who gave a presentation around 2006 and said something like ‘when we come up with new things, we’re talking about the web, we make certain assumptions and intuitions, and as computer scientists at Google, our intuitions are all bad. They’re not very good. We have a certain perspective. We don’t have the empathy to understand what people actually want.’ So if we base things upon our intuitions, we’re missing the boat many times.’
Notes from the original presentation website wrote up the outcome as:
‘To address these challenges, Russell proposed a model he calls the “3M Points of View.” Each of the three components in this model has its own individual strengths and weaknesses, but when triangulated, together paint a fuller picture of user goals and intent than anyone alone. The 3M’s are micro, meso, and macro. The micro level involves data collected at the lowest, millisecond level, such as eye tracking in a lab study. Meso level involves data collected over minutes or a day, such as field studies. Finally, macro level involves data collected at a high level over minutes, hours, or days, such as log file analysis.’
Damien: Do you think the shifts in search technology, the vectorisation, reliance on internal layers of machine-learned outputs are in some way to compensate?
Bill: Neural networks are something that was around at the start of artificial intelligence but wasn’t conceived the same way. It’s not the same as it was conceived of back in the 1950s. It’s changed a lot.
There are other ways to do artificial intelligence that aren’t neural networks that are based more upon knowledge. We see that with Google’s Knowledge Graph.
There was the ‘Fact Based Repository’ part of the ‘Annotation Framework’ in the mid-2000s that was the predecessor to Knowledge Graph. There were a dozen or so scientists, engineers from Google who were working on this annotation framework. One of the things they came up with was the ‘Fact Based Repository’, and another one they came up with was local search at Google.
Local searches serve a semantic web type, knowledge base thing where instead of looking at links, they’re looking at citations, which are mentioned so the names of local entities or businesses but not linked. They also really paid attention when they found address information in a postal format. Most web pages are unstructured data, so the postal format was something that they could recognise using ‘wrappers’ that would help them say this is a format; this is an address for business.
Damien: At a Google Webmaster Conference in 2019, Google software engineer, Paul Haar, said when we search it is not just for the thing we searched for, but also its synonymous relationships. As both a form and a contributor to knowledge, what does this mean for the future of SEO?
Bill: Knowledge is the understanding of the relationships between entities. When you have, like, a baseball player, and you’re saying, ‘this baseball player is related to this baseball team’, which is also an entity, and the relationship is he’s a player for the team. That’s knowledge. A knowledge graph is all the edges pointing towards each other, towards different entities and their classes and their attributes.
When somebody searches and they’re looking for an answer to a question, like ‘When was George Washington born?’ They’re not looking for a web page that shows the date that he was born, they’re looking for the date that he was born and an answer wouldn’t be a list of links to the date George Washington was born, the answer would be the date.
When somebody searches for ‘What are the best restaurants in New York City’ they want a list of links to the best restaurants.
Some queries are answer-seeking queries where a person searching expects an answer right away.
Damien: There’s an evolution there; people are searching either for knowledge or searching for information on the web.
Bill: If you have zero-click search results, people finding what they’re looking for in the search results page, they do not necessarily intend to click on anything. They don’t want to spend time searching through a list of websites. They just want to answer.
Damien: So we have to adjust, to think differently and begin to consider how we optimise better to serve those users’ needs rather than our own?
Bill: If your business model is based upon simple facts like ‘Abraham Lincoln was six foot four’ you need a new business model.
Damien: You spend lots of time in the knowledge graph these days, what’s your view on how this changes SEO?
Bill: LilyRay who is a DJ in New York City was a drummer on the West Coast before she moved out to New York and she got into Knowledge Graph by entering some information into MusicBrainz as a musician. Google added her to Knowledge Graph and just recently noticed that she also does SEO, she writes for Search Engine Land.
They changed her Knowledge Graph entry to show that she does SEO now, in addition to being a musician and all of this is algorithmic.
Google acquired a company called Wavii in 2013. Created at the University of Washington, and one of the people behind it was Oren Etzioni, who’s now CEO of Allen Institute for AI. The idea behind Wavii was to read websites and to learn knowledge from those websites and put the information in something that now looks like a Knowledge Graph.
That’s an approach that Google is likely moving in at this point. It’s reading websites using something like BERT, which pre-trains natural language processing. So it helps to understand and recognise entities, it tags different parts of speech when it reads stuff, it tries to understand those relationships between entities which brings the Knowledge Graph information together. Knowledge Graph is a game-changer for SEO.
Damien: Most searches have a specific context; that is, we’ve got our surroundings and background knowledge to inform the meaning of our search. How do you think Google orchestrates understanding this?
Bill: One of the clients I’ve worked within the past, was Baltimore.org, which is the visitor and conference centre website.
I was working with a remote copywriter at that point. I sent her a message saying, ‘Could you try to put together a travel guide, a walking tour of the city of Baltimore with information about the famous churches, the universities, the celebrities, and places people could go in Baltimore?’ I wanted all these different entities, all these different people that Baltimore is well known for, like Billie Holiday, or Cab Calloway.
My copywriter put together a lengthy document, and we put that up on the website. Within three months or so, it was like the sixth most visited page on the site. People were driving into Baltimore following this walking tour visiting all these places, which was the purpose behind the website as a visitor centre website; it fulfilled its promise to users.
Damien: Let me play this scene from the Wizard of Oz for you, the man behind the curtain, is that partly what you hope, to show the version of Google behind the curtain?
Bill: It makes sense to try to do that. Lewis Carroll was a mathematician too, right? There’s a lot behind the curtain at Google, lots of things that I’ve noticed because I spend the time with patents that I don’t think a lot of other people have seen.
I’ve written about them, I’ve tried to share things I’m seeing. It’s funny, sometimes things I write about I won’t see too much of that from Google until 10 or 12 years later.
For instance, Google released a white paper on semantic topic modelling, which is the process behind phrase-based indexing, which I first wrote in 2005, and I’ve written about maybe a dozen times since then. It’s an algorithmic ranking process at Google.
It does not use PageRank or information retrieval that most SEOs try to use on websites, but uses a slightly different approach. The person who came up with it, Anna Patterson, had built one of the most significant search engines of the 21st century before she joined Google, she created ‘Recall’ a beta search engine at the Internet Archive.
You could use it to search all the different pages on the Internet Archive, and there were billions of them because they had the same sites multiple times at different points in time.
Google acquired that technology and then hired Anna, and then the first phrase-based indexing patent came out.
Damien: And you were paying attention.
Damien: Because of your focus on patents and company acquisitions, you’ve caught many things others have missed. Where have SEOs dropped the ball and effectively taken the red pill?
Bill: Some things get distracting, it’s like misdirection. One of the most popular blog posts I wrote was one where I read about rank modifying spammers patent. What the patent was about was something they called ‘transition rank’.
Google might see that you’re making changes to the website and in response to those changes, take some random action. Either increase your rankings, decrease your rankings, or make no changes at all. The point behind it was they’re trying to use social engineering to encourage you to make other changes, maybe make more desperate changes.
If you make changes that are spammy, it might end up getting you penalised. Google wanted to see if you are genuinely spammy. If you were doing something a little bit spammy, they might decrease rankings significantly to try and encourage you to try and do more of the right things.
Damien: What should SEOs be taking notice of right now and focused on?
Bill: Google, with these neural networks or machine learning approaches, is doing an excellent job of recognising features and patterns on websites. When Amit Singhal wrote a paper in 2011 of questions, you should ask about whether or not your site is a high-quality site.
Questions like ‘Would you be comfortable giving your credit card information to this site?’ is one of the items. There are 22 questions like that in the blog post from him geared towards making your site high quality.
It’s the type of thing that you want to understand so that your site fits. High quality is something that will help you rank, will help you show up, and results will make people happier to use your website and enhance your reputation with that audience.
Some patents focus on quality that uses different measures. One of them is the concept of ngrams. These are collections of short passages, several words long like a two-gram, three-gram, four-gram, or five-gram.
Google might look at your website, break all the words, all the sentences on your pages down into four-word ngram combinations. Then build statistics based upon these four-word combinations, and compare them to four-word combinations on other websites that are similar from their website vectors work.
They’re giving your content a quality score based upon those ngram combinations and trying to get a sense of how well you do against something like the New York Times or The IRS which they’ve scored on these ngram statistics and determined to be high quality.
Something I see that people fail at when they talk about SEO: they don’t speak about audiences enough, understanding the audience. Who is your customer? What is it that they desire? How do you make that affinity with them? How do you listen to them? How do you learn about them?
You’ve got clients who are subject matter experts in the industries that they’re in. They have people who talk to their audience every day in their sales departments or the customer service departments. If you’re not asking if you’re not interviewing those people, you’re missing opportunities.
Damien: Preaching to the choir there, Bill. Gosh, we’ve covered some ground, one last question, what’s next in search?
Bill: I see lots of different things. One thing I wrote recently in the Search Engine Journal was author vectors, which I thought was kind of fun as an English major in college and somebody who did a deconstruction of literature. It was interesting seeing Google saying we can understand what people have written by training machine learning on the stuff they’ve written lots of times and creating vectors about these things.
Then when we see something else written by the same person, we can recognise it from the writing style we’ve assessed. I saw a patent about ngram and statistical models and understood its potential. I wondered to myself, when will they apply similar techniques to do that with authors?
Damien: Thanks so much for your time, Bill, you’ve been very generous with your time. Speak soon.