Tag Archive | "Learning"

SearchCap: Eric Schmidt changes role, enterprise SEO & machine learning for PPC

Below is what happened in search today, as reported on Search Engine Land and from other places across the web.

The post SearchCap: Eric Schmidt changes role, enterprise SEO & machine learning for PPC appeared first on Search Engine Land.



Please visit Search Engine Land for the full article.


Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Posted in IM NewsComments Off

Machine Learning and It’s Impact on Search

The terms machine learning (ML) and artificial intelligence (AI) have been cropping up more often when it comes to organic and paid search. Now a recent report by Acquisio has confirmed just how effective machine learning for search results.

According to Acquisio, paid search accounts that have been optimised for machine learning have 71% higher conversion rates and have lower cost-per-click (CPC). But these were not the only benefits that accounts using machine learning enjoyed. The web marketing company also revealed these accounts were also able to reach their target spending levels and had lower churn rates.

The data implies that small marketing teams and CMOs now stand on an even playing field with more established companies now that ML is more affordable, effective and accessible to everyone.

This doesn’t mean that marketers should ignore organic search and original, value-laden content. Paid search might be the easiest way to rank high in search engines, particularly since AI will be doing the bulk of the work, developing campaigns that have greater odds of being seen by the right searchers at the proper time. However, organic search is more authentic and will last longer than paid searches.

The goal now is to understand how ML impacts the search system and how to take advantage of the technology’s evolution that made paid and organic searches more effective.

Paid vs Organic Search: Which Wins in the End?

There’s been an ongoing debate as to which is better – paid or organic searches. Interestingly, both have come out on top, but at different times and conditions. The results have depended on the type of research done and other outside factors. For instance, a study conducted in 2011 showed that organic search was more effective. However, paid search has outpaced its counterpart from 2013 onwards. But this appears to be due to the changes Google has made to its algorithm.

So which is better? Andy Taylor, the Associate Director of Research at Merkle, believes that flexibility is the best option. Instead of just sticking to one approach, companies should determine what search strategy is ideal for their business at the moment and the technology that’s currently available. After all, the ideal marketing strategy for your company now will probably change in a few months as customers change their expectations and technologies expand.

Machine Learning is Changing More Than Search

The rise of machine learning has also resulted in a shift to data-driven models instead of the conventional attribution models. This multi-touch attribution model (MTA) relies on an analytics scale that’s more descriptive and takes into account various touchpoint outputs, like ad interactions, ad creative, or exposure order. It also allows marketers to have a better understanding of how factors, like a distinct set of keywords and ad words, can affect a conversion.

But it’s not just search capacities that machine learning has an impact on. The technology is also being used to refine and make algorithm changes. It has been theorized that Google’s RankBrain utilizes machine learning to assess if the company has to revise its own rankings based on what the consumer searches for and whether the user was satisfied with the result.

Machine Learning Will Push for More Sophisticated Content

Because machine learning technology is developing more advanced SEM capacities and sophisticated algorithms, search engines are pushing marketers and content producers to deliver more refined content. This would eventually lead to search engines becoming more discerning to the quality of online content a company is putting out. This means producing high-quality content that particularly targets what the consumer is looking for becomes more vital than ever before.

Machine learning and AI are impacting every aspect of marketing. Companies should start understanding them and how to utilize ML-optimized tools effectively in their marketing campaigns.  

The post Machine Learning and It's Impact on Search appeared first on WebProNews.


WebProNews

Posted in IM NewsComments Off

Learning to Re-Share: 4 Strategies to Renew, Refresh, and Recycle Content for Bigger Reach

Posted by jcar7

In the nearly three years the MeetEdgar blog went live, we’ve published more than 250 posts, written over 300,000 words, searched for hundreds of .gifs, and used our own tool to share our content 2,600 times to over 70,000 fans on social media.

After all that work, it seems silly to share a post just once. Nobody crumples up an oil painting and chucks it in the trash after it’s been seen one time — and the same goes for your content.

You’ve already created an “art gallery” for your posts. Resharing your content just lets the masses know what you’ve got on display. Even if hundreds or thousands of people have seen it all before, there’s always someone new to your content.

In a social media landscape that’s constantly changing, building a solid foundation of evergreen content that can be shared and shared again should be a key part of your social media strategy.

Otherwise, your art gallery is just another building in the city.

But wait… aren’t we supposed to be writing fresh content?

Yes! One of the biggest misconceptions about resharing is that it’s a spammy tactic. This is just not true — provided that you’re resharing responsibly. We’ll explain how to do that in just a moment.

Resharing actually does double-duty for your brand. It not only gets the content that you spent your valuable time creating in front of more eyeballs (and at optimal times, if you want to get fancy), it also frees you up to have more authentic, real-time social interactions that drive people to your site from social media — since you’ve got content going out no matter what.

Did we mention that resharing is good for SEO? Moz Blog readers know that the more people engage with a post, the better your blog or site looks to search engines. And that’s only one facet of the overall SEO boost (and traffic boost!) resharers can see.

How resharing impacts SEO

Big brands are probably the most prolific content resharers. Heck, they don’t even think twice about it:

BuzzFeed is a perfect example of the value of repeating social updates, because they don’t necessarily NEED to.

So why do they do it anyway? Because it gets results.

Social sharing alone has an impact on SEO, but social engagement is really where it’s at. Quality content is totally worth the up-front time and cost, but only if it gets engagement! You up your chances of engagement with your content if you simply up your content’s exposure. That’s what resharing does awesomely.

With literally zero tweaks to the content itself, BuzzFeed made each of those social posts above double in value. Chances are, the people who saw these posts the first time they were shared are not the same people who saw them when they were reshared.

But simply resharing social posts isn’t the only way to get more engagement with your content. This post covers how companies large and small do resharing right, and highlights some of the best time-saving content strategies you can implement for your brand right now.

1 – Start at the source: Give old posts a new look

Lots has changed in five years — the world got three new Fast & Furious movies and LKR Social Media transformed from a consulting service into social media automation software.

We’ve done the math: three months is one Internet year and five years is basically another Internet epoch. (This may be a slight exaggeration.) So when we transferred some of our founder’s older evergreen blog posts to the new MeetEdgar blog, we took stock of which of those posts had picked up the most organic traffic.

One thing that hadn’t changed in five years? A blog post about how Vin Diesel was winning the social media game was still insanely popular with our readers:

Screen Shot 2017-07-24 at 11.53.06 AM.pngScreen Shot 2017-07-24 at 11.54.34 AM.png

Writing blog posts with an eye toward making them as evergreen as possible is one of the smartest, most time-saving-est content marketing strategies out there.

There weren’t a ton of tweaks to make, but we gave this popular post some love since so many people were finding it. We pepped up the headline, did a grammar and content rundown, refreshed links and images, updated social share buttons, and added more timely content. The whole process took less time than writing a brand new post, and we got to share it with tens of thousands of followers who hadn’t seen it when it was originally published.

So… check your metrics! Which evergreen posts have performed the best over time? Which have lots of awesome organic traffic? Make a list, do a content audit, and start updating!

2 – Find your social sharing “sweet spot” by repackaging your content

When you read studies that say many social media users reshare social posts without ever clicking through to the content itself… it can be a little disheartening.

Okay, a LOT disheartening.

You’ve probably spent tons of time creating your content, and the thought that it’s not getting read NEARLY as often as it could be is a recipe for content marketing burnout. (We’ve all been there.)

But it’s not all for naught — you might just need to experiment until you find the “sweet spot” that gets people to read and share. One way to do that is to simply repackage content you’ve already written.

The tried-and-true “best of” post offers a reprieve from the content-creation grind while still delivering tons of value to your fans and readers.

Repackaging is best when it reframes your content with a new focus — like rounding up similar posts based on a theme. (You can do this in reverse, too, and turn one great post into a bunch of fresh content to then share and reshare!)

If you can get people to your site, a “best of” post encourages readers to stay longer as they click links for the different articles you’ve gathered up, and engage with content they may never have thought to look up separately.

Most fun of all, you can repackage your content to target new or different subsets of your audience on social media. (More on that in the next section.)

3 – Social shake-up: Reaching and testing with different audiences

“What if the same person recognizes something that I’ve already posted in the past?” you might be asking right about now. “I don’t want to annoy my followers! I don’t want to be spammy!”

Forget about people resharing social posts without reading the content behind the links — most people don’t see your social posts at all in the first place.

This is just one of those uncomfortable facts about the Internet, like how comment sections are always a minefield of awful, and how everyone loves a good startled cat .gif.

That doesn’t mean you should repeat yourself, word-for-word, all the time. Chances are, you have more than one type of reader or customer, so it’s important not just to vary your content, but also to vary how you share it on social media.

Savvy marketers are all over this tactic, marketing two sides (or more) of the same coin. Here are a couple of examples of social sharing images from a Mixpanel blog post:

Option A

Option B

Both Option A and Option B go to the same content, but one highlights a particularly juicy stat (problem statement: “97% of users churn”) and the other hits the viewer with an intriguing subheader (solution statement: “behavior-based messaging”). In this way, Mixpanel can find out what pulls in the most readers and tweak and promote that message as needed.

Pull a cool anecdote from your post or highlight a different stat that gets people excited. It can be as easy as changing up the descriptions of your posts or just using different images. There’s so much to test and try out — all using the same post.

4 – Automate, automate, automate

Remember, your best posts are only as good as the engagement they get. That fact, however, doesn’t mean you have to keep manually resharing them on social media day in and day out.

Unless, of course, you’re into that boring busywork thing.

Automating the whole process of resharing evergreen content saves tons of time while keeping your brand personality intact. It also frees you up to have real-time interactions with your fans on social media, brainstorm new post ideas, or just go for a walk, and it solves the time crunch and the hassle of manually re-scheduling posts, while actually showcasing more of your posts across the massive social media landscape. Just by spacing out your updates, you’ll be able to hit a wider range of your followers.

(This is probably a good time to check whether your social media scheduling tool offers automatic resharing of your content.)

Now, social media automation isn’t a substitute for consistently creating great new content, of course, but it does give your existing evergreen content an even better opportunity to shine.

Win with quality, get things DONE with resharing

It’s noisy out there. The law of diminishing returns — as well as declining social reach — means that a lot of what you do on social media can feel like shouting into the void.

And there’s not a huge ROI for shouting into voids these days.

Responsible resharing is an important part of your overall content marketing strategy. As long as you keep your content fresh, create new quality content regularly, and talk to your fans where and when they’re most active, chances are people won’t see the same thing twice. The data shows you’ll get more clicks, more traffic, and better SEO results — not a bad bonus to that whole “saving lots of time” thing.

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don’t have time to hunt down but want to read!


Moz Blog

Posted in IM NewsComments Off

Apple Shares Source Code For Machine Learning Framework at WWDC 2017

Apple’s recent WWDC (Worldwide Developers Conference) saw the unheralded release of Core ML, which will reportedly make it easier for developers to come up with machine learning tools across the Apple ecosystem.

The way this works is that developers need to convert their creations into an API that is compatible with the Core ML. They then have to load their programs into the Apple Xcode development before it can be installed on the iOS.

Developers can use any of the following frameworks: Keras, XGBoost, LibSVM, Caffe, and scikit-learn. To make it even easier for them to load their models, Apple is allowing them to come up with their own converter.

According to Apple, the Core ML is “a new foundational machine learning framework used across Apple products, including Siri, Camera, and QuickType.”

The company explained that this new machine learning tool would be “the foundation for domain-specific frameworks and functionality.”

One of the primary advantages of the Core ML is that it speeds up the artificial intelligence on the Apple Watch, iPhone, iPad, and perhaps the soon-to-be-released Siri speaker. If it works the way that is billed, any AI task on the iPhone, for instance, would be six times quicker compared to the Android.

The machine learning tools supported by Apple Core ML include linear models, neural networks, and tree ensembles. The company also promised that private data by users won’t be compromised by this new endeavor. This means that developers can’t just tinker with any phone to steal private information.

Core ML also supports:

  • Foundation for Natural Language Processing
  • Vision for Image Analysis
  • Gameplay Kit

“Core ML itself builds on top of low-level primitives like Accelerate and BNNS, as well as Metal Performance Shaders,” the company added.

But Apple is reportedly not content with just releasing the Core ML. According to rumors, the company is looking to fulfill its promise of helping to build a very fast mobile platform. In fact, if the rumors are true, the company is also building a much better chip that can handle AI tasks without compromising performance.

Though Core ML seems promising, Apple is certainly not blazing the trail when it comes to machine learning. In fact, Facebook and Google have already unveiled their own machine learning frameworks to optimize the mobile user’s experience.

The new machine learning framework is still part of Apple’s Core Brand, which already includes Core Audio, Core Location, and Core Image as announced earlier.

The post Apple Shares Source Code For Machine Learning Framework at WWDC 2017 appeared first on WebProNews.


WebProNews

Posted in IM NewsComments Off

SearchCap: AMP images, Google Maps promoted places & machine learning

Below is what happened in search today, as reported on Search Engine Land and from other places across the web.

The post SearchCap: AMP images, Google Maps promoted places & machine learning appeared first on Search Engine Land.



Please visit Search Engine Land for the full article.


Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Posted in IM NewsComments Off

Search Buzz Video Recap: Google Mobile Index, Penguin Destruction, Machine Learning & More

This week in search, Google said they are launching their mobile only index within months and moving the desktop index as a secondary index. Google Penguin 4.0 is completely live now. Gary Illyes said Penguin can act to discount all your links…


Search Engine Roundtable

Posted in IM NewsComments Off

Machine learning for large-scale SEM accounts

Can machine learning be applied to your PPC accounts to make them more efficient? Columnist David Fothergill describes how he utilized machine learning to find new keywords for his campaigns.

The post Machine learning for large-scale SEM accounts appeared first on Search Engine Land.



Please visit Search Engine Land for the full article.


Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Posted in IM NewsComments Off

SearchCap: Machine learning, content marketing & search rankings

Below is what happened in search today, as reported on Search Engine Land and from other places across the web.

The post SearchCap: Machine learning, content marketing & search rankings appeared first on Search Engine Land.



Please visit Search Engine Land for the full article.


Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Posted in IM NewsComments Off

Why Learning to Write Is the Toughest and Best Thing You’ll Do

why better writing is worth the effort

Trigger warning: I’m about to list some terms that might give you nightmares. Do you remember these?

  • Gerunds
  • Participles
  • Sentence diagrams
  • Split infinitives
  • Absolute modifiers

Just talking about them might cause you to flash back to middle school. You’re sitting in a sweaty classroom, listening to the chalk squeak as your teacher writes the definition for each term on a dusty chalkboard.

You, in the meantime, are mentally calculating how many minutes are left before lunchtime.

Here’s the thing about learning to write: It’s not about the terms above. Yes, you need to be aware of them. But if you think learning to write well is about mastering grammar, you’re missing the point.

Learning to write goes beyond masterful handling of the parts of speech. They’re just the paper that wraps the gift.

Today, we’re going to cover what writing well really looks like and why it might be the hardest and best skill you’ll ever master. It’s the gift that keeps on giving: read on to learn why.

Well-written ideas are easier to circulate

You’re reading Copyblogger. And you probably read paper books, ebooks, news sites, long posts on social media, and more.

When we want our ideas to spread, we start by making them look good in writing.

With the surge in popularity of podcasting and the widespread use of visual platforms like Instagram, Pinterest, and even YouTube, you might wonder if the written word matters as much as it used to.

But most podcasts and videos start out as words in one form or another. They begin life as a written outline, a thoroughly-planned script, or notes on an index card.

When you’re a proficient writer, those first-draft-quality notes will do a better job getting your ideas out of your head and into a new format.

Jerod Morris, co-host of both The Showrunner and The Digital Entrepreneur podcasts, starts 75 percent of his episodes with some type of written outline. Written outlines help you plan, pace, and express your information.

And any medium will benefit when you write well.

That headline you want to add to your Pinterest image? That quote for the image you plan to post on Instagram?

When you know how to write well, you can count on finding the perfect words more easily and expressing them in a way that’s compelling and gets noticed.

Your ideas stand a better chance of spreading when they’re well-written.

Writing builds discipline (and not just for writing)

Here’s the worst-kept secret about becoming a better writer: To get good at it, you have to write — more than you think and on a regular basis. And you’ll need to keep it up for longer than you may expect.

You may find that in order to keep your writing chops in the best possible shape, you need to write almost every single day.

Our own Sonia Simone, for example, has written something every day for thirty years, with the exception of a short stint in the hospital while she recovered from major surgery. (We’ll let that one slide.)

There aren’t too many things in life that promise the kind of return that writing on most days will give you. (More on that below.)

And the discipline you’ll build from steadily working to improve your writing will build your character.

You may even find yourself looking around for more to write about once you’re in the habit of writing most days.

Clearer thoughts are born from your writing structure

The process of writing clearly usually involves starting with some sort of basic outline.

But since “outline” is another one of those scary words from English class, I want to offer you the phrase I use to describe the initial stage of writing — building the backbone.

Building the backbone refers to the process of working out the basics of the idea you want to express by deciding on a topic, then hashing out the underlying structure of how you’ll present your information. It forces you to bring your ideas into focus and clarify them so they are strong enough to support the concepts you’ll hang on them.

There’s nothing like figuring out your supporting arguments to help you clarify your ideas.

This process can spill over into many other areas of your life.

Structuring your thoughts before you share them in writing will get you into the habit of structuring your thoughts before you share them anywhere else as well. It will help you clarify your message and put it into a form that’s easier to understand.

How can you become a better writer?

Start with the posts below. They’ll cover the basics and help you establish a strong writing habit that you can use to structure and share your ideas.

You can also download and print out this poster (3.3 MB) to help motivate you to write on a regular basis.

And consider joining us inside Authority: it’s where people who want to become better writers get weekly education, support, and encouragement so they can get there faster.

Become a better writer inside Authority

Authority is our content marketing training and networking community designed to help you build the skills you need to profit online.

Put your name on the Authority interest list by clicking on the button below. We’ll let you know when we open our doors.

Join the Authority interest list

The post Why Learning to Write Is the Toughest and Best Thing You’ll Do appeared first on Copyblogger.


Copyblogger

Posted in IM NewsComments Off

The Machine Learning Revolution: How it Works and its Impact on SEO

Posted by EricEnge

Machine learning is already a very big deal. It’s here, and it’s in use in far more businesses than you might suspect. A few months back, I decided to take a deep dive into this topic to learn more about it. In today’s post, I’ll dive into a certain amount of technical detail about how it works, but I also plan to discuss its practical impact on SEO and digital marketing.

For reference, check out Rand Fishkin’s presentation about how we’ve entered into a two-algorithm world. Rand addresses the impact of machine learning on search and SEO in detail in that presentation, and how it influences SEO. I’ll talk more about that again later.

For fun, I’ll also include a tool that allows you to predict your chances of getting a retweet based on a number of things: your Followerwonk Social Authority, whether you include images, hashtags, and several other similar factors. I call this tool the Twitter Engagement Predictor (TEP). To build the TEP, I created and trained a neural network. The tool will accept input from you, and then use the neural network to predict your chances of getting an RT.

The TEP leverages the data from a study I published in December 2014 on Twitter engagement, where we reviewed information from 1.9M original tweets (as opposed to RTs and favorites) to see what factors most improved the chances of getting a retweet.

My machine learning journey

I got my first meaningful glimpse of machine learning back in 2011 when I interviewed Google’s Peter Norvig, and he told me how Google had used it to teach Google Translate.

Basically, they looked at all the language translations they could find across the web and learned from them. This is a very intense and complicated example of machine learning, and Google had deployed it by 2011. Suffice it to say that all the major market players — such as Google, Apple, Microsoft, and Facebook — already leverage machine learning in many interesting ways.

Back in November, when I decided I wanted to learn more about the topic, I started doing a variety of searches of articles to read online. It wasn’t long before I stumbled upon this great course on machine learning on Coursera. It’s taught by Andrew Ng of Stanford University, and it provides an awesome, in-depth look at the basics of machine learning.

Warning: This course is long (19 total sections with an average of more than one hour of video each). It also requires an understanding of calculus to get through the math. In the course, you’ll be immersed in math from start to finish. But the point is this: If you have the math background, and the determination, you can take a free online course to get started with this stuff.

In addition, Ng walks you through many programming examples using a language called Octave. You can then take what you’ve learned and create your own machine learning programs. This is exactly what I have done in the example program included below.

Basic concepts of machine learning

First of all, let me be clear: this process didn’t make me a leading expert on this topic. However, I’ve learned enough to provide you with a serviceable intro to some key concepts. You can break machine learning into two classes: supervised and unsupervised. First, I’ll take a look at supervised machine learning.

Supervised machine learning

At its most basic level, you can think of supervised machine learning as creating a series of equations to fit a known set of data. Let’s say you want an algorithm to predict housing prices (an example that Ng uses frequently in the Coursera classes). You might get some data that looks like this (note that the data is totally made up):

In this example, we have (fictitious) historical data that indicates the price of a house based on its size. As you can see, the price tends to go up as house size goes up, but the data does not fit into a straight line. However, you can calculate a straight line that fits the data pretty well, and that line might look like this:

This line can then be used to predict the pricing for new houses. We treat the size of the house as the “input” to the algorithm and the predicted price as the “output.” For example, if you have a house that is 2600 square feet, the price looks like it would be about $ xxxK ?????? dollars.

However, this model turns out to be a bit simplistic. There are other factors that can play into housing prices, such as the total rooms, number of bedrooms, number of bathrooms, and lot size. Based on this, you could build a slightly more complicated model, with a table of data similar to this one:

Already you can see that a simple straight line will not do, as you’ll have to assign weights to each factor to come up with a housing price prediction. Perhaps the biggest factors are house size and lot size, but rooms, bedrooms, and bathrooms all deserve some weight as well (all of these would be considered new “inputs”).

Even now, we’re still being quite simplistic. Another huge factor in housing prices is location. Pricing in Seattle, WA is different than it is in Galveston, TX. Once you attempt to build this algorithm on a national scale, using location as an additional input, you can see that it starts to become a very complex problem.

You can use machine learning techniques to solve any of these three types of problems. In each of these examples, you’d assemble a large data set of examples, which can be called training examples, and run a set of programs to design an algorithm to fit the data. This allows you to submit new inputs and use the algorithm to predict the output (the price, in this case). Using training examples like this is what’s referred to as “supervised machine learning.”

Classification problems

This a special class of problems where the goal is to predict specific outcomes. For example, imagine we want to predict the chances that a newborn baby will grow to be at least 6 feet tall. You could imagine that inputs might be as follows:

The output of this algorithm might be a 0 if the person was going to shorter than 6 feet tall, or 1 if they were going to be 6 feet or taller. What makes it a classification problem is that you are putting the input items into one specific class or another. For the height prediction problem as I described it, we are not trying to guess the precise height, but a simple over/under 6 feet prediction.

Some examples of more complex classifying problems are handwriting recognition (recognizing characters) and identifying spam email.

Unsupervised machine learning

Unsupervised machine learning is used in situations where you don’t have training examples. Basically, you want to try and determine how to recognize groups of objects with similar properties. For example, you may have data that looks like this:

The algorithm will then attempt to analyze this data and find out how to group them together based on common characteristics. Perhaps in this example, all of the red “x” points in the following chart share similar attributes:

However, the algorithm may have trouble recognizing outlier points, and may group the data more like this:

What the algorithm has done is find natural groupings within the data, but unlike supervised learning, it had to determine the features that define each group. One industry example of unsupervised learning is Google News. For example, look at the following screen shot:

You can see that the main news story is about Iran holding 10 US sailors, but there are also related news stories shown from Reuters and Bloomberg (circled in red). The grouping of these related stories is an unsupervised machine learning problem, where the algorithm learns to group these items together.

Other industry examples of applied machine learning

A great example of a machine learning algo is the Author Extraction algorithm that Moz has built into their Moz Content tool. You can read more about that algorithm here. The referenced article outlines in detail the unique challenges that Moz faced in solving that problem, as well as how they went about solving it.

As for Stone Temple Consulting’s Twitter Engagement Predictor, this is built on a neural network. A sample screen for this program can be seen here:

The program makes a binary prediction as to whether you’ll get a retweet or not, and then provides you with a percentage probability for that prediction being true.

For those who are interested in the gory details, the neural network configuration I used was six input units, fifteen hidden units, and two output units. The algorithm used one million training examples and two hundred training iterations. The training process required just under 45 billion calculations.

One thing that made this exercise interesting is that there are many conflicting data points in the raw data. Here’s an example of what I mean:

What this shows is the data for people with Followerwonk Social Authority between 0 and 9, and a tweet with no images, no URLs, no @mentions of other users, two hashtags, and between zero and 40 characters. We had 1156 examples of such tweets that did not get a retweet, and 17 that did.

The most desirable outcome for the resulting algorithm is to predict that these tweets not get a retweet, so that would make it wrong 1.4% of the time (17 times out of 1173). Note that the resulting neural network assesses the probability of getting a retweet at 2.1%.

I did a calculation to tabulate how many of these cases existed. I found that we had 102,045 individual training examples where it was desirable to make the wrong prediction, or for just slightly over 10% of all our training data. What this means is that the best the neural network will be able to do is make the right prediction just under 90% of the time.

I also ran two other sets of data (470K and 473K samples in size) through the trained network to see the accuracy level of the TEP. I found that it was 81% accurate in its absolute (yes/no) prediction of the chance of getting a retweet. Bearing in mind that those also had approximately 10% of the samples where making the wrong prediction is the right thing to do, that’s not bad! And, of course, that’s why I show the percentage probability of a retweet, rather than a simple yes/no response.

Try the predictor yourself and let me know what you think! (You can discover your Social Authority by heading to Followerwonk and following these quick steps.) Mind you, this was simply an exercise for me to learn how to build out a neural network, so I recognize the limited utility of what the tool does — no need to give me that feedback ;->.

Examples of algorithms Google might have or create

So now that we know a bit more about what machine learning is about, let’s dive into things that Google may be using machine learning for already:

Penguin

One approach to implementing Penguin would be to identify a set of link characteristics that could potentially be an indicator of a bad link, such as these:

  1. External link sitting in a footer
  2. External link in a right side bar
  3. Proximity to text such as “Sponsored” (and/or related phrases)
  4. Proximity to an image with the word “Sponsored” (and/or related phrases) in it
  5. Grouped with other links with low relevance to each other
  6. Rich anchor text not relevant to page content
  7. External link in navigation
  8. Implemented with no user visible indication that it’s a link (i.e. no line under it)
  9. From a bad class of sites (from an article directory, from a country where you don’t do business, etc.)
  10. …and many other factors

Note that any one of these things isn’t necessarily inherently bad for an individual link, but the algorithm might start to flag sites if a significant portion of all of the links pointing to a given site have some combination of these attributes.

What I outlined above would be a supervised machine learning approach where you train the algorithm with known bad and good links (or sites) that have been identified over the years. Once the algo is trained, you would then run other link examples through it to calculate the probability that each one is a bad link. Based on the percentage of links (and/or total PageRank) coming from bad links, you could then make a decision to lower the site’s rankings, or not.

Another approach to this same problem would be to start with a database of known good links and bad links, and then have the algorithm automatically determine the characteristics (or features) of those links. These features would probably include factors that humans may not have considered on their own.

Panda

Now that you’ve seen the Penguin example, this one should be a bit easier to think about. Here are some things that might be features of sites with poor-quality content:

  1. Small number of words on the page compared to competing pages
  2. Low use of synonyms
  3. Overuse of main keyword of the page (from the title tag)
  4. Large blocks of text isolated at the bottom of the page
  5. Lots of links to unrelated pages
  6. Pages with content scraped from other sites
  7. …and many other factors

Once again, you could start with a known set of good sites and bad sites (from a content perspective) and design an algorithm to determine the common characteristics of those sites.

As with the Penguin discussion above, I’m in no way representing that these are all parts of Panda — they’re just meant to illustrate the overall concept of how it might work.

How machine learning impacts SEO

The key to understanding the impact of machine learning on SEO is understanding what Google (and other search engines) want to use it for. A key insight is that there’s a strong correlation between Google providing high-quality search results and the revenue they get from their ads.

Back in 2009, Bing and Google performed some tests that showed how even introducing small delays into their search results significantly impacted user satisfaction. In addition, those results showed that with lower satisfaction came fewer clicks and lower revenues:

The reason behind this is simple. Google has other sources of competition, and this goes well beyond Bing. Texting friends for their input is one form of competition. So are Facebook, Apple/Siri, and Amazon. Alternative sources of information and answers exist for users, and they are working to improve the quality of what they offer every day. So must Google.

I’ve already suggested that machine learning may be a part of Panda and Penguin, and it may well be a part of the “Search Quality” algorithm. And there are likely many more of these types of algorithms to come.

So what does this mean?

Given that higher user satisfaction is of critical importance to Google, it means that content quality and user satisfaction with the content of your pages must now be treated by you as an SEO ranking factor. You’re going to need to measure it, and steadily improve it over time. Some questions to ask yourself include:

  1. Does your page meet the intent of a large percentage of visitors to it? If a user is interested in that product, do they need help in selecting it? Learning how to use it?
  2. What about related intents? If someone comes to your site looking for a specific product, what other related products could they be looking for?
  3. What gaps exist in the content on the page?
  4. Is your page a higher-quality experience than that of your competitors?
  5. What’s your strategy for measuring page performance and improving it over time?

There are many ways that Google can measure how good your page is, and use that to impact rankings. Here are some of them:

  1. When they arrive on your page after clicking on a SERP, how long do they stay? How does that compare to competing pages?
  2. What is the relative rate of CTR on your SERP listing vs. competition?
  3. What volume of brand searches does your business get?
  4. If you have a page for a given product, do you offer thinner or richer content than competing pages?
  5. When users click back to the search results after visiting your page, do they behave like their task was fulfilled? Or do they click on other results or enter followup searches?

For more on how content quality and user satisfaction has become a core SEO factor, please check out the following:

  1. Rand’s presentation on a two-algorithm world
  2. My article on Term Frequency Analysis
  3. My article on Inverse Document Frequency
  4. My article on Content Effectiveness Optimization

Summary

Machine learning is becoming highly prevalent. The barrier to learning basic algorithms is largely gone. All the major players in the tech industry are leveraging it in some manner. Here’s a little bit on what Facebook is doing, and machine learning hiring at Apple. Others are offering platforms to make implementing machine learning easier, such as Microsoft and Amazon.

For people involved in SEO and digital marketing, you can expect that these major players are going to get better and better at leveraging these algorithms to help them meet their goals. That’s why it will be of critical importance to tune your strategies to align with the goals of those organizations.

In the case of SEO, machine learning will steadily increase the importance of content quality and user experience over time. For you, that makes it time to get on board and make these factors a key part of your overall SEO strategy.

Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don’t have time to hunt down but want to read!


Moz Blog

Posted in IM NewsComments Off

Advert