Lessons learned from people that developed mechanisms we use today, such as removing spam content and implementing Search Engine Optimization or AI, are precious and should become a rule book for all who want to grow and expand their knowledge, practices, and business. 

Therefore we have decided to bring you the lessons and conclusions of SEO godfather Matt Cutts. He shared with his followers the outline of best practices we can implement in creating our website and how development in different areas can impact us, our business, and our revenues.

Therefore, if you are interested in learning from the best and finding some of the fundamental principles of working on SEO and web development, we invite you to read our following passages. 

 

Who is Matt Cutts?

We are talking about a mechanical engineer who has a Ph.D. degree in this area and has been a recognized member of its community thanks to his work in Google and the U.S. government. 

Student Years

Matt Cutts graduated from UNC-Chapel Hill and pioneered some mechanisms we still use while adapting our websites to browsers, especially Google.

His university journey was not a straightforward experience in any sense. He decided to take a risky move and pursue a career in the internet startup when he didn’t have much experience in this area. What was very suitable for his job was that during his years at UNC-Chapel, he needed to take extra classes to fulfill his university’s criteria. That time was when Matt pursued two courses related to computer science and both on the topic of search engines, which helped him further in his work. 

During his time at Chapel Hill, Matt got interested in term frequency and coding, so he started his professional career with Linux before it went public. Later he continued his mostly known career at Google.

Time at Google

He began his journey in Google, working on how to make regular expressions faster for a string matcher. Besides this, he has also been working on developing programs and regulations to stop the spam culture that has been proven to manipulate rankings. 

His very first major project was related to pornography. Matt is a pioneer of the family-safe version of Google. Namely, he wrote a pornography classifier that would exclude any such content. He worked on this project for around three months, trying to shape porn content and recognize patterns in people’s behavior when searching for pornographic content.

“Porn Cookie” Guy

Since there were many ways for people to Google pornography content, Matt needed help, which led to a famous anecdote about “porn cookies,” which were used as a reward for all who found pornography on the server or helped him with suggestions. Firstly, people thought it was like a trap from human resources, which fostered the idea to ask his wife to bake cookies. Other teams started to do similar things, so they created spam cookies.

The project further fostered Matt’s career and his ideas in other projects, which helped him develop some of the still-used patterns and codes in search engines. Additionally, it inspired him to find a new way of solving problems, and in the next chapter, we discuss what he learned on this journey.

Lessons

Lesson 1:  Sometimes Creativity Makes A Big Difference

One of Matt’s first lessons during one of his lectures emphasized the importance of creativity and its impact on his work in multiple stages. One of the examples he uses is controversy tied to the Digital Millennium Copyright Act (DMCA).

The main issue that Google was dealing with included searches. Namely, if you typed in more evil than Satan, you would have Microsoft as number one on your search results. It was simply a result of the algorithms, and nobody had hard-coded this, meaning nobody meant for that to happen. 

How DMCA Works?

Matt’s quick elaboration on what can be done to fix the issue is connected to DMCA. Also, he points out how this experience is relevant to the lesson he learned.

As you may know, DMCA is an act where one individual can submit a  request to a web host or Google to take down any content that its original author has stolen. An issue in this area is verifying the truth and whether this is something Google should do. Therefore Google has created a mechanism in which issues are resolved in court once the person files a counter-complaint leaving Google out of the dispute. 

The complication comes if the person doesn’t counter-notify, which still means that the search engine or the information provider has to take that information down. One should remember that this is only valid for issues or requests that are copyrighted and related to something that is not considered public domain.

Example of Issues with DMCA

An example of this issue was the case when the Church of Scientology wanted to suppress critics based in Norway running a trademarked website. The guy in Norway eventually did not wish to counter-report since that would put him accountable under U.S. law. 

The Google team was caught in a tough decision since they had a legal obligation to take this content down, and the person who would typically counter notify was not willing to do so.

To resolve this issue, Matt had a creative approach and added a notice to the bottom of the page that said due to a complaint. Eventually, this evolved into taking down content for specific areas, such as the example of the Chinese government and the results of Tiananmen Square or Falun Gong.      

The other search engines in that specific country also started to add that notice, so it helped people get context on the situation even if they weren’t Google users. This creative and out-of-the-box thinking was a great example of a way to implement a new approach to your day-to-day work while fulfilling the needs of all stakeholders involved.

Lesson 2: Be Proactive, and while doing so, ask for what you want

Matt had the opportunity to work in various departments at Google, eventually becoming more creative and proactive because he was working on important issues. That is why he emphasizes the importance of proactivity and working on stuff you love, knowing the benefits and difference it might bring.

What he took from this experience is that no one cares more about you or your career than yourself and that it is imperative. It should motivate you to ask what you want and gamble since it can pay off, and you will end up working on things you like the most.

Experience That Changed Matt’s Perspective

One of his examples shared in the lecture is related to Matt’s work in the advertising group at Google. He was a team member for about a year focusing on front-end programming. Eventually, he got drafted into the ads group, where he learned a lot of skills.

Working in this sphere, especially during work on SafeSearch, inspired Mark. He figured out how people cheat the system and rank their pages higher by working with spam and ads.

He got interested and wanted to improve search quality, which is the breakpoint in his time at Google when Matt decided he wanted to work on spam even more. According to him, this period in his life taught him the importance of time and how it is hard to get momentum right.

Lesson 3: Circumstances Change – Adapt To It

Regarding momentum, Matt pointed out one quote from Paul Graham: many startup ideas initially look bad because something has changed or flipped in the world. 

Matt continues by saying that the circumstances are never the same, and we need to keep up with development and understand the implications that other areas can have on our work.

In addition, Matt mentioned Peter Thiel, who puts the same logic by elaborating that what we believe is that no one else considers because if all you have is the conventional wisdom, it’s hard to set yourself apart from the crowd. It’s hard to get momentum. 

Example of Affordable Care Act

Doing something different is problematic because it causes disruption but also an opportunity, and it is an experiment. So you should take that risk as an entrepreneur or an imaginative person. 

An example of this is the Affordable Care Act. In this context, Matt mentioned how ACA was developed and how you can now get health insurance even if you have preexisting conditions, which made a huge difference. Additionally, you don’t have to get insurance through your employer anymore.

It was a significant game changer in the USA, which impacted many people becoming entrepreneurs and opening their businesses since they didn’t have the conditions before. Matt, in his lecture, uses a great example of how ACA impacted preexisting companies and new ones.

One of the results of ACA, which nobody saw coming, is that companies like Uber or Lyft that rely on independent contractors might be happy about Obamacare due to the ability of people to get their insurance. Matt mentioned that if one of us talked with Travis Kalanick, the CEO of Uber, he would be pleased about Obamacare.

Something like health insurance that looks utterly unrelated to startups could have a considerable ripple effect down to something important to you, which means that your business, online or offline, has more chance of gaining competition. These seemingly unrelated things explained how everything could have a ripple effect that you can use to your benefit.

Lesson 4: You Can Do A Lot Of Cool Stuff With Data

One critical lesson Matt mentions is the time before Google search engines. In this perspective, Matt has shared examples that help us understand how you can achieve a lot using data. He also shares how an innovative approach to old tools can be successful using newly available data but also how we should be careful regarding the information we receive since it can happen that, in practice, we get surprised. 

Improvement and Its Problems

One of the first experiences related to the proper use of material and data that Matt shared is connected to the overall appearance of Google’s machines, which had evolved from giant machines to hard drives that can fit in Legos. 

Google used the opportunity to change giant retailers for your local retailer to be accessible to everyone and could store the same amount of data as before. So the ability to use commodity hardware instead of these massive and expensive things meant that Google could run much more cheaply simply because they had big data about the quality and adaptability of new hardware.

With the development and commercial expansion, Matt continued the lecture with natural anticipation that the system would have bugs. Now the case is that since you have a hundred machines, they’re more likely to go down more frequently, and as Matt elaborates, people in Google have to figure out how to make them more reliable in software.

Issue of Increasing Size of Your Reach

To elaborate on the size issue Google was potentially dealing with, Matt uses the equation example, which relates to the advantages of having data that can preempt the issues in the long run.

Google gets five billion queries a day and 100 billion queries monthly, rounding it to five billion. How many of them could you do before you had some catastrophic bug? 

Maybe one in a million, you know 99.99999% of the time, things work perfectly? Well, it turns out you just have to do the math and divide it, and that means 5,000 catastrophic bugs a day, so you have to deal with things going up and down, corner cases machines that go faulty, and so these were the sorts of machines that we used in the early days like literal PC hardware.

Eventually, programmers created a crawler machine to make it better and avoid issues. These are known as corkboard machines because they have a thin layer of corkboard to keep the motherboards from shorting out. Over time programmers were trying to get it to work reliably, and now we got better. Hence, the machine started to look more and more professional thanks to development, adapting to the situations, and more extensive data span available.

The issue with bugs noted as Google switched to the cheaper hardware made Matt aware of the spillover effect, transferring this issue to the other lines of programming. It brings us to the story when Paul Buchheit, who invented Gmail, said machines had all these deadly crashing bugs, so maybe people in Google should write the code without them. 

Matt continues the story that eventually, they reached a point where they could take a cage in a data center and, in three days, fill it up with devices but still had to work on the software to be reliable and fix the presented issue. 

Inspiring Story About Storing the Data

Continuing with the lessons learned by using an immense amount of data you can access and impractical hardware, Matt shared how hard drives had also been an issue for a while and their ability to store data. Namely, he mentions that a mechanical hard drive takes about 10 seconds to seek data you type in. He uses an excellent analogy where you can imagine it as an arm that physically has to move to the right place on a platter. 

On the other hand, if you have everything in RAM, instead of doing only about a hundred seeks per second, you can do a lot of desires, and you need to do those seeks when people type in users’ queries.

So if you put your entire web index not on a hard disk but in RAM, which is a lot of money, you can still get much higher throughput, so that trade-off might be worth it. 

It imposed the question about commodity hardware, and is that why Google succeeded? Because of the innovative approach using old tools and an extensive data set? According to Matt, it seems so.

Matt’s PageRank Experience

Moving on to the differences encountered while dealing with data that can be inspiring, Matt brings up his personal experience with PageRank. Matt mentioned that he thought he was know-how from his paper about PageRank.

Matt realized that in practice, the PageRank was more sophisticated than had ever appeared in the papers or had been published in the patents. You could have searched for different abstractions spanner, Google file system, MapReduce, etc. That made a difference and fostered enthusiasm to work further.

This example shows that inspiration from data and immense knowledge can occasionally be deceiving. He reminds us that people see great results, but it is not that we have that one shining epiphany, and then everything is fixed. It takes dozens or hundreds of different things to make a successful: company or a successful career.

Lastly, Matt was talking about our tendency to question everything and to remember that circumstances change. So if you have a lot of data, you can do either a sophisticated algorithm or even a dumb algorithm, but even so, you can create pretty cool stuff. 

AI and Development

A great example was the case of artificial intelligence and its development which he experienced at the university. Before, there was no even option to take this course, but now, 15 years later cannot take this course at university because of its popularity.

It is also the moment where it is worth mentioning Matts’s reference to the Brain – a computer used for learning and connected to AI. They programmed a computer to watch YouTube and see what can be concluded from that.

The base for this was a neural network. It had many layers of neurons, and the program developed its neuron to recognize cats. Therefore you can use that same kind of processing to better word recognition. Since 2012, every Android phone after Jellybean has had a much better word error rate using voice recognition because of a deep neural network. So 30% drop in word or error rate was like the most significant improvement in speech recognition they’d had in years.

As Matt said, Jellybean and Android are other indicators that if circumstances change, you must be willing to adapt to them. You can do a lot of stuff with data, so you have to question your assumptions and remember, back in 1999, everybody was like A.I. is stupid, it’ll never be anything, and universities didn’t offer any courses on it, and that’s not the case anymore.

Lesson 5: Murphy’s Law

Matt stated that his time at Google was beautiful and full of fantastic chances, significant growth, and progress, but it wasn’t all cookies and success. He had to deal with several lawsuits as an engineer, such as those by companies for data claims, up to nations that wanted to control the internet and adapt it to their politics.

Matt closes his lecture with one last important lesson that he connects to the entire outline of his time at Google.

Everyone should prepare themselves for lawsuits and every deposition they might encounter. He remembers that Google had the Search King case about whether it was allowed to rank its search results the best way it thought. Another lawsuit is called Kinder start, and because they wanted to win these lawsuits, they also had issues with governments worldwide.

Know Your Team and Adapt

Besides, these issues can also happen in company culture. Knowing that personal conflicts can cause problems, you must be ready to discover your balance, work hard, and have fun.

Despite all this, we live in a world that progressed, where now you can record meetings or wrestle with spammers. Matt remembers pleasant memories that led to the moment he is now, among which he highlighted when he had caught a company that was spamming and threw them out, and they sent in what’s called a re-inclusion request, so they sent a cookie the size of a pizza.

Another great memory is the time Larry and Sergey, founders of Google, acquired to introduce the new Googlers or Nooglers. Then they talked about all the stuff that happened in the last week, and then for about half an hour, you were allowed to grill the company’s founders. 

That is a hilarious opportunity to ask billionaires all you want and question their decisions. Also, ask for what you want and be proactive. 

One Last Thought for the End of the Lecture 

You would be amazed at how much of a difference it makes when you just tell somebody I would like to do X because I’ve been a manager, and if I know someone wants to work on something, they’re going to work about twice as hard and they’re going to be more persistent at it.

If you don’t set yourself apart from the crowd, it’s hard to get enough momentum to succeed no matter what. Some weird bad things are going to happen. But don’t forget to have fun and take pictures a long way.