Social media statistics: correlation versus causation

Most online statistics that you read are bullshit. They are weak and misleading numbers that persevere in the form of infographics, unsound “studies,” and in our everyday business storytelling. Not only can these statistics be misleading, but if misinterpreted they can have negative effects on your business.

In this post I’m going to discuss the relationship between correlation and causation, and why these are important when deciding what data to believe. At the end I’ll describe a way to look at comparative social data and make better decisions around its results.

Correlation and causation defined

Let’s start with a very straightforward definition of correlation and causation:

Correlation is the degree to which two or more quantities are linearly associated. (source: Wolfram Alpha)

Causation is the act or process of causing something to happen or exist. (source: Merriam Webster)

It’s important to understand the distinction between the two. Correlation shows an association between variables but can never show that one thing causes another. Here are a couple of graphic representations of correlation between two variables.

Correlation coefficient - positive
This is an example of positive correlation between two variables. The correlation coefficient is the slope (m) of the line. In this image the slope is positive, which means that there is a positive correlation between variables. The correlation coefficient can range from -1 (perfect negative correlation) to 1 (perfect positive correlation)
Negative Correlation coefficient
This is an example of negative correlation between two variables. The correlation coefficient is the slope (m) of the line. In this image the slope is negative, which means that there is a negative correlation between variables. The correlation coefficient can range from -1 (perfect negative correlation) to 1 (perfect positive correlation)

Online statistics will never rarely ever show causation

You’ve probably heard the phrase “correlation doesn’t necessarily equal causation.” It’s an overused expression that people generally use before they inappropriately infer cause from correlation. Put more simply, because two things happen at the same time doesn’t mean that they are causing the other to happen.

As an example, my wife and I have two kids. We had the same (horrible) nurse in the delivery room for both of their deliveries. Although there is 100% correlation between the nurse’s presence and our kid’s births, neither caused the other.

Online statistics tend to infer causation from correlation, despite brandishing the “correlation is not causation” cliche. This is wrong to write and wrong to accept.

Consider how difficult it is to determine what causes anything else. All variables would have to be isolated and then one would have to be explicitly shown to cause a specific outcome. In science this can be done through well-constructed, rigorous, expensive trials. For the type of data that informs infographics and most “studies” it is close to impossible.

Which is not to say that correlation is a bad thing….

Some correlation is actionable

We wouldn’t take the time to observe and measure correlation unless it was helpful to us.

Consider the Google algorithm. There are many variables and conditions that inform what results come back to you on a SERP.  Searchmetrics and Moz (and I’m sure others) do thorough analyses of SERPs to attempt to gauge how much correlation there is between certain variables and search engine rank. Let’s take a look at the top twelve ranking factors and their correlation coefficients from Searchmetrics 2014 data:

  1. Clickthrough rate (.67)
  2. Relevant terms (.34)
  3. Google +1 (.33)
  4. Number of backlinks (.31)
  5. Facebook shares (.28)
  6. Facebook total (.28)
  7. Facebook comments (.27)
  8. Pinterest (.27)
  9. SEO-visibility of backlinking URL (.26)
  10. Facebook Likes (.25)
  11. Tweets (.24)
  12. % backlinks = “rel=nofollow” (.23)

You may be tempted to look at the list and determine that you need to focus more on Google +1s and less on Facebook posts in order to raise your SERP position. That conclusion assumes that this is a report of causality, and the relationships these studies establish are not causal at all. From a statistical standpoint, the correlation coefficients in this study are weak for every factor other than clickthrough rate. Recall also that Google can’t see a large portion of Facebook and said explicitly that +1s don’t influence SERPs, so they can’t cause anything to happen. These results show (weak) correlation where there isn’t a high probably of causation.

An alternative way to look at this study of correlation is to ask yourself how many of these items you can improve for your website? Is there a way to devote your resources that might impact a few of these items? Since we really can’t understand the specifics of the Google algorithm (which x causes y), we can look at correlation here as a list of things that when improved may improve our SERP position to some degree. This is why many SEO practitioners will focus more effort on link building (#4) than on soliciting +1s.

If you consider how hard it would be to isolate a specific variable from others, you can safely assume that comparative data measures correlation. By disregarding the statistical possibility of causation, you can be properly critical of the data and make better decisions around the data.

Conclusion

Many writers pay lip service to correlation and causality and then continue to incorrectly infer causality from their data. Don’t fall down this rabbit hole.

If you take nothing else away from this post understand this: All comparative social data and statistics will be correlative. You’re unlikely to ever isolate an explicit cause for your outcomes. And while I know this is unsettling, treating your comparative data as correlative allows you to make better decisions about it.

Let me know your thoughts on this one.

How to easily capture cropped screenshots using Greenshot

If you write a lot of “how-to” articles or posts, you likely have come across the same issue that I have: cropping full-screen screenshots is time-consuming and stupid. After all, the technology that captures an entire browser window should be able to capture a partial screen with similar ease (amen?).

The technology does exist and one of the most popular pieces of (free) software to do this is Greenshot.

About Greenshot

Greenshot is an app that enables you to capture a partial screenshot. This eliminates the need to crop a screenshot in WordPress or another third party app.

It is straightforward to download and set-up (You can download it for free at http://getgreenshot.org/). Once installed, it offers you a menu to directly initiate screenshots or to set your preferences:

greenshot preferences

Greenshot preference options

In this next section I’m going to show some screenshots of preferences (in kind of a meta-twist captured by Greenshot).

There are a lot of really useful preference options for Greenshot. One of them is the Hotkey options. You can set this to whatever you like – you can see that I have Greenshot running when my computer is running, and anytime I hit control-down it triggers a Greenshot window. You can set this up however you like, this works well for me.

greenshot preferences1

One of the more useful options to set is the “destination.” If you check “select destination dynamically” the destination prompt will pop-up everytime you take a screenshot. I prefer to set it ahead of time depending upon how I intend to use the image.

  • For applications like OneNote, Evernote and Gmail where I want to paste an image into a note or email, I check the “copy to clipboard” option.
  • For writing (specifically in WordPress), I choose the “save as” option. This saves the file to my downloads so that I can batch upload to WordPress and publish them easily.

greenshot3

Alternatives

Greenshot isn’t a unique app – it’s just very user-friendly. I’ve also tried PicPick at the recommendation of somebody writing for Lifehacker, and found it equally useful but not as easy to setup and use (it may be that my application has a limited scope).

In any event – Greenshot isn’t the only free software that can do this for you.

Conclusion

If you use a lot of screenshots (or even a few) in your writing – Greenshot is a great app to have in your arsenal. It is easy to use, does what it says, and is free. It’s hard to argue with that triumvirate of benefits.

If you have experience with Greenshot or an alternative please share in the comments.

How to grow Twitter followers using SocialBro

One of my favorite tools for Twitter management is SocialBro. I initially became a fan because of the Excel file export feature, and have grown to appreciate many features (the crazy easy follow/unfollow mode comes immediately to mind).  Let me disclose upfront that SocialBro provides me with a free premium account, so the features that I describe may not be the same features that you would see with a free account.

As much affinity as I have for SocialBro, I also would hate to advocate for an app based upon my admiration rather than on tangible fact. So I wanted to put SocialBro to the test and see how effectively I could grow my Twitter audience using only their app.

My mission (Should I choose to accept it)

socialbro mission icon

In the upper right hand corner of SocialBro, there is a little rocket icon that links to a “mission” page. Currently they have four “missions” for users to follow to help accomplish certain tasks using the app. One of them is this one:

socialbro mission

What follows below is my attempt to follow these recommendations to grow my personal Twitter account @jimdougherty.

1. Change Twitter bio to describe myself in 160 characters. Update website, location and use a recognizable image.

So here’s a screenshot of my Twitter profile before:

dougherty profile twitter

I wordsmithed the heading, changed the header image (using a template in the awesome cloud graphic design tool Canva). And this is what I came up with:

new twitter profile

Okay, so I’m not a graphic designer. But I think it’s an improvement and hope you do, too.

2. Assess the current state of my Twitter community

Next, SocialBro recommends assessing the state of your community. The mission tool takes you step-by-step through key parts of the app to do this.

First step: People who follow me that I don’t follow back

socialbro followback filter

You probably know enough about Twitter to understand that two groups of people follow you. Some follow you because you’re interesting and the others follow you to try and coax a reciprocal follow so they can unfollow you and appear to have a legion of fans that adore them.

My general rule is to try and follow as many people who take the time to follow me. If I can spot a spammy or inactive account I’ll skip it, but I followed back the majority of those 64 people.

socialbro follows

Second step: Influencers

Here’s an embarrassing fact: when I look at influencers for the last 30 days, the only account that I mentioned is YouTube (in my defense I purposefully keep my Tweets sparse on this account).

In any event, if you’re Tweeting a lot or using a tool like Triberr to promote your posts. The “influencer” filter is a pretty awesome way to make sure you’re following everyone that your Tweeting to and about.

Third (and fifth) step: Tags

One of the neat features of SocialBro is the “bio tagcloud” tool. It is a collection of the most common tags in bios of people that you follow or that follow you.

The third check in this mission is the tagcloud for the people you follow and the fifth is for your followers. The check is meant to make sure that your audience interests are what you want them to be, AND for you to take note for future follows.

socialbro bio tagcloud

Fourth step: Content check

The fourth point in this mission is to assess Twitter content. SocialBro recommends a mix of  your own posts and other people’s content. I don’t Tweet enough from this account, but for the period that I’m trying to grow my audience I will post two pieces of content per day, 25% my own content.

The check of the follower tagcloud is meant to give an idea of the interests of people who follow you now, so I will make a point to use these terms in each Tweet that I send out:

socialbro bio tagcloud content

Sixth step: make a list

I don’t know that making a list of all of my Twitter followers would be particularly helpful, so I filtered out the high and low outliers and focused on people who have a pretty good sized following and who follow back. Seems to me they’re doing something right.

socialbro list creation

The seventh point (step) is to monitor this list daily and find content to RT. I will resolve to RT one Tweet from this list per day while I’m trying to grow my followers.

Eighth step: Target my competitor’s community

Yikes, who is my competitor? Because I want to show how awesome this tool is, I’m going to target a good Cincinnati-based company, Ahalogy. If you’re unfamiliar with them, they are doing some creative and cutting-edge stuff around Pinterest and are worth knowing about.

socialbro ahalogy

You can see that SocialBro gives you an entire dashboard for your competitors, with information analogous to your own.

Steps nine and ten – analyze the competitive data

One of the cool things about the competitive analysis tool is the “source followers not followed by you and not following you” widgets.

 

socialbro competitive followers

I pulled up the first list, and followed everyone in Cincinnati that is following Ahalogy (following appropriate followers of your competition is step eleven). It took me about 20 seconds.

Step twelve – find the ideal times to Tweet

SocialBro (like many services) creates a best time to Tweet report based upon the activity on your account. So I ran the report from the dashboard:

socialbro best time to tweet

You can’t schedule Tweets from SocialBro proper, but two of the awesome partnerships that they maintain are with Hootsuite and Buffer. You can export best time to tweet data from SocialBro to Buffer or Hootsuite (if you have a pro account for either):

socialbro hootsuite and buffer export

(Cheapskate alert) I don’t have a pro account for either, so I configured Buffer based upon the information in the SocialBro report. Maybe not as slick as the automated tools, but got the job done:

socialbro best time to tweet

And that’s that. My account is configured properly to grow my Twitter followers using SocialBro.

Conclusion

Will I be able to grow my Twitter audience demonstrably in a month? I don’t know. But you can clearly see that SocialBro is a powerhouse for Twitter and I suspect that these tactics will work (note: you can also manage Instagram with SB). As of this writing I have 1,346 Twitter followers. I will update this post in 30 days to share how successful these tactics were to increase my follower count, and you can always check out my progress on my Twitter page.

I would be remiss not to mention the author of this “mission,” Leticia Polese. Her understanding of Twitter and the SocialBro tools is pretty impressive. You may want to follow her, and you can do that easily by pressing the button below:


Should tech tools embrace niche users?

If you’re familiar with the story of Novocaine, German chemist Alfred Einhorn discovered it and (thinking it had medical application) fought to keep it from being used in dental surgeries. It wasn’t until Einhorn died that Novocaine became widely used in dental applications, due in large part to his resistance to allow its use in the dental niche.

I think of this story in the context of tech tools. Some have potential for mass market application and haven’t achieved it (Google Plus and Twitter come immediately to mind). Some tech tools are upstart companies whose features may have very strong application for specific niches (Haiku Deck Zuru, which automatically populates slide decks, and Canva, a cloud-based graphic design tool come immediately to mind). Like Google Plus and Twitter have found niche audiences with much higher adoption than the general public, I suspect there are professional niches that will find unique utility from Haiku Deck Zuru and Canva.

Which leads me to IQTell. I’ve written promotional pieces for IQTell after being drawn to the app for its unique utility as a universal inbox, Evernote manager, calendar manager and Getting Things Done (GTD) database. When I had the opportunity to talk with one of the creators of IQTell, Sahaf Flam, I shared that I thought many more people would find his app helpful beyond GTD disciples who use Evernote (there are literally dozens of us). Sahaf said I sounded a bit like Jason Vichinsky.

Jason is a lawyer, and has become a champion for IQTell in the law profession. I got in touch with him, and asked if I could ask him some questions about the utility of IQTell with him and he agreed.

Back to the Novocaine example, I feel unease using independent tech tools without understanding their funding and their prospects for the future. I love IQTell but don’t want to get a note one day in my inbox saying that the product is discontinued. If you remember the untimely passing of Behance’s Action Method project management tool, you understand it’s a real concern. I love that IQTell embraces niches where it is extraordinarily useful and that folks like Jason find great professional utility in it. I find IQTell indispensable and hope it becomes the biggest thing since Novocaine.

If you’re not familiar with IQTell – this is the app what we’re discussing:


I think how Jason took this tool and found specific utility for it in his niche rather fascinating and hope you do as well. Note that I changed his responses in a couple of places where he misunderstood my relationship to IQTell (I’m just a fellow user).

Q: What aspects of IQTell are useful for you as a lawyer?

IQTell  is the best project management software I have ever used.  If you are an attorney with a small firm or a solo and want to have a completely paperless office IQTell and Evernote are the only tools you will ever need.

The way the product allows me to handle email is absolutely phenomenal. The ability to immediately link that email to an action or project is indispensable.  I am now able to deal with all of my email and get my inbox empty in a fraction of the time it used to take.

That I am able to use the product on my Mac, my iPad, and my smart phone is fantastic. This means that I  literally carry my entire practice with me wherever I go. Moreover, because the data is stored both locally and in the cloud I have four backups of my data. I never worry about losing anything.

Q: What tools were you using to manage your information before using IQTell?

I have test driven Rocket Matter, MyCase, Cleo, Remember the Milk, Toodledo.  I also experimented with trying to use Google products together. Nothing I have used comes close to IQTell.

Q: From your perspective what are the strongest aspects of the IQTell app?

The strongest aspect of IQTell is simply this: the user truly has everything she needs in one place in a clean robust interface that works on any computer or mobile device with an Internet connection. It is the holy grail of the paperless mobile law practice.

Q: Can you talk about the utility of Evernote in your profession?

I use Evernote to store all of my documents, Web clippings, audio recordings, photographs, and notes. I also use it with penultimate to store handwritten notes for those occasions when I cannot use voice typing.  That I am able to access all of these from any computer with an Internet connection is phenomenal. Evernote truly is the best at what it does. Combined with IQTell an attorney needs nothing else to run a completely paperless and highly organized practice.

Q: What aspects of time management are critical for lawyers and how does IQTell address these?

Because it allows me to carry my entire practice with me wherever I go, I am able to make use of time that might otherwise be lost. There are often occasions when I am sitting in court waiting to be called. Before I had your product that time Was lost. Now I can use that time to work on almost anything. Moreover, it is not uncommon to encounter another attorney who is in court for a separate matter. If that attorney and I have a case together we can go sit in the conference room and discuss that case while we are waiting.

IQTell is free for a 60 day trial and then is $5.95 – $9.95 per month depending upon the number of email accounts that you need to sync. I should probably mention that neither Jason or I was compensated for mentioning IQTell or any other product in this piece.

Why you should use Cloudflare on your website

When you start a WordPress site it seems very straightforward and easy: choose a theme, write, publish, proofread (not necessarily in that order). As you get into it more, other considerations cascade on you: social sharing, plugins, hosting, speed and security among them.

One really powerful implementation that you can do to make your site more secure, faster and to decrease your hosting costs is to use Cloudflare. Cloudflare is a powerful (freemium) content delivery network (CDN) that filters out malicious users and delivers your content faster by utilizing its network of servers to serve your content.

Long story short, Cloudflare is pretty awesome.


How Cloudflare makes your site more secure and fast

Cloudflare works by adding itself as an intermediary between your domain and your host. For example, when you type leaderswest.com to go to my site you are routed through Cloudflare rather than directly to my host. Suspicious traffic is filtered by Cloudflare, and a CAPTCHA screen is presented to flagged visitors before they are permitted on to your site. This means that malicious bots have a lower likelihood to ever see your site.

I channeled my inner Sal Khan and drew a diagram of how Cloudflare works. Implementing Cloudflare onto your site is simply a process of signing up, redirecting your DNS, and choosing some options.

cloudflare diagram

You can see that domain traffic that is routed through Cloudflare may be served by either the host, the Cloudflare CDN, or a combination of the two (depending upon the caching options that you choose).

Success that I’ve had with Cloudflare

I implemented Cloudflare when I was using WPEngine as my host. For those familiar with WPEngine, they are a great host but they charge a ton for user overage and the traffic they charge you for runs two to three times greater than what Google Analytics measures. Long story short – my hosting costs got very expensive and I wanted to try anything I could to keep my WPEngine hosting (I couldn’t justify it, but that’s another story).

WPEngine recommended Cloudflare to try and mitigate my traffic. I was livid about the overages and didn’t see the utility, but I tried Cloudflare anyhow. It immediately filtered out hundreds of malicious attempts every day. It also serves thousands of pages from the CDN (rather than the host) every month, which keeps my hosting costs lower. It has been super helpful.

I also just implemented Cloudflare on a nonprofit website and cut almost two seconds off of its average load time.

This is a screenshot of a Pingdom speed test from my site. When this (image-intensive) content is explicitly delivered by the host, load time is about 1.5 seconds. This test shows the content delivered in 389ms from the Cloudflare CDN.

cloudflare pingdom

The freemium model

I use the free version of Cloudflare and I use the highest caching settings that I can manage. There are three levels up from the free version: Pro, Business and Enterprise. Each has more speed and services than the free version (a list of all features can be found here), but for my purposes free is a pretty great price.

There are a lot of benefits of Cloudflare that I deliberately haven’t discussed. For bloggers who want increased security and faster load times, Cloudflare is an excellent low-cost way to achieve these two objectives. It takes a little time to set-up, but once it’s humming it can deliver your content very fast to its intended audience. I think a lot of bloggers would benefit from a Cloudflare implementation.

You should also note that Cloudflare recommends that WordPress webmasters use the Cloudflare WordPress plugin. In tests with the P3 Plugin Profiler, this plugin consistently added less than one-hundreth of a second to page load time. In other words, if you’re using Cloudflare with WordPress you should install this plugin as well.

I’m curious to know your thoughts if you have experience with Cloudflare. Please feel welcome to share in the comments.