Analytics

Sarah Worsham / Jun 10, 2014

Book Review: The Visual Organization by Phil Simon

The Visual Organization: Data Visualization, Big Data, and the Quest for Better Decisions (Wiley and SAS Business Series)We are inundated and surrounded by Big Data. So much so, that it is very difficult to wrap your mind around how to use all the information that pelts us from all directions every day. Understanding how to use Big Data is becoming imperative for organizations and data visualization is the method to turn data into understandable information. In Phil Simon’s latest book, The Visual Organization, he uses easy-to-understand explanations and real-world examples from a variety of organizations to help you visualize (pun intended) how your organization could use data visualization. Starting with an example of how a data visualization company made it big, Phil shows how the rapid innovation and quickly changing industry of d.v. has opportunities for big disruptions in every field. Organizations of every kind and size would find this book a helpful primer and resource on the way to becoming a visual organization.

Why & How to Use Data Visualization

Divided into four sections, The Visual Organization is a pleasant and interesting read straight-through, but also allows more experienced individuals to skip to the most important sections. You will get an understanding of what data visualization (d.v.) is, why you should care, why some level of d.v. is vital for every organization and how higher levels of d.v. can improve your business strategy by better informing key decisions. One size does not fit every organization, especially for tools. Phil discusses a variety of data visualization tools from large enterprise vendors, open source tools and design firms.

What is a Visual Organization?

Key to becoming a visual organization is understanding what one actually looks like, beyond just concepts and tools to making d.v. an integral part of how the company operates. Phil uses real organizations in his case studies, which include large companies, small companies, non-profits and show many different ways to leverage d.v. to improve how the organizations operate.

Become a Visual Organization

Becoming a visual organization goes beyond just purchasing some d.v. tools, and Phil discusses steps, strategies, tips and insights to help you put d.v. into practice with a 4 level framework. Understanding what a visual organization would do when making business decisions is key to properly implement data visualization and Phil will help you navigate mistakes, myths and challenges in a real world execution.

Data Visualization Tools

As more data visualization tools come to market, the ability to analyze the wealth of information organizations collect will not only become easier, it will be vital to staying competitive. The easier it becomes to get good information from so much data, the more companies will start to leverage data visualization. Get ahead of the curve by reading Phil’s book to understand the how, what, and why of using data visualization for your organization.

Buy Now:

(links to the book on Amazon are affiliate links — feel free to use them, or not)

Sarah Worsham / Apr 25, 2014

Understanding Facebook Boosted Post Metrics

Cardboard rocket (Photo credit: Matt Biddulph)

As you may be aware, Facebook is making it more difficult for companies and organizations to engage directly with their intended audiences. To counter this (and to make revenue), Facebook offers the ability to boost an individual post so that it shows up in the news feed of your intended audience. These boosted posts can be fairly low-cost, with a minimum boost of $5 per day. So how well do these boosted posts perform and what sort of metrics does Facebook provide? To find out, I boosted a post on Lady Paragon’s Facebook page (a site I run with my sister for Women in STEM careers).

Facebook Post Pre-Boost

Here’s what the post looked like before I boosted it:

The metrics we see are:

1 person liked it (red circle)
There was 1 share (green rectangle)
976 people saw the post (blue rectangle)

I boosted this post for 1 day at a budget of $5 and targeted fans & friends of fans of Lady Paragon’s Facebook page.

Facebook Boosted Post Metrics

Here are the metrics after the boost:

The metrics provided are:

4 people liked it (red circle) — 1 was from before, which Facebook properly reports in the red circle in the How people engaged with your post section.
1 share (green rectangle) — this was from before the boost
3102 saw the post (blue rectangle) — Facebook reports that 2079 were from the boost in the Paid Reach box. You can also see the percentage of paid to organic in the box with the 3102 — blue was organic, green was paid
4 link clicks (purple circle)
Engagement of 7 — this is the number of link clicks added to the number of post likes

Facebook Post Insights

When you look at the post in the page Insights, you see the following metrics (more recent data):

The orange bar shows the number of people who viewed the post, divided into lighter orange for organic, darker for paid. 3.1K is pretty close to the 3102 mentioned above. 218 is the number of post clicks and 116 is the number of likes, comments and shares. This is very interesting. Either the boosted metrics didn’t include some of the stats, boosting the post helped increase the organic reach and engagement, or the post received an unusually high number of engaged traffic from some of the people who saw it (remember that when someone likes a post, their network sees that they liked it, at least for a short time period).

Hypothesis: Boosting a Post Improves It’s Organic Reach & Engagement Too

I boosted another post on the same page (same budget $5) and got the following results:

1331 Paid Reach
5 Engagements – 3 link clicks, 2 post likes

According to the post insights, the post got 15 post clicks and 4 likes, comments and shares. Not nearly as high, so there probably is a difference in the influence of the people who engaged with each post.

If we look at the Google Analytics traffic to the actual post on the website (April 2-April 22), the April 2nd post (Jessica Kirkpatrick) had 338 pageviews (20 from Facebook), while the April 9th post (Kate Synder), had 93 pageviews (77 from Facebook).

Conclusion: Unclear, More Results Needed

The results do tend to show that a boosted post receives more organic engagement, especially if there are people with good influence that do engage with the post. Using good targeting to reach the right audience to improve engagement on a boosted post may provide the most beneficial of results. More testing is needed — I’ll continue to monitor my efforts.

What have you found with Facebook boosted posts?

Sarah Worsham / Apr 1, 2014

What You Want to Know Will Dictate What You Measure

Google Analytics Hacks (Photo credit: Search Engine People Blog)

With so much data, it’s so easy to get caught up in all the numbers. Looking at the wrong numbers will result in faulty analysis and recommendations — you may fix things that aren’t broken, or not fix things that are. Or you may think you have the right solution to a problem, but not even be looking in the right place. While it may seem obvious, taking a step back to understand what you want to know first will help you choose the right measurements.

Step 1: State What You Want To Know

The first step is to state what it is you want to know — without using any measurements or metrics at all. For example, if a website has several links to its Careers page on the homepage, ‘We want to know what place on the homepage is sending the most traffic to our Careers page’. This is quite different from ‘We want to know where the most traffic is coming from that enters the site on our Careers page’. One is about the design of the homepage and the marketing there — the other is about external marketing efforts to the Careers page. We’re going to stick with the first for our example…

Step 2: Refine Your Data Needs

Now that we understand what we want to know, we can further refine our data needs to see if we have the right measurement in place. When we look at the homepage, we can see that there are actually 4 places that someone could click through to the Careers page: 1) Menu at the top of the page 2) Linked text in the middle of the page 3) Ad box in the sidebar 4) Menu in the footer of the page. Ok, so now we know there are 4 possible links a visitor could click, so in order to answer our ‘what we want to know’ question, we have to be able to tell the difference between each of these 4 links.

Step 3: Know Your Technologies

Unfortunately, the next step is fairly technical. In order to know if you can distinguish between the 4 links, you need to know 1) how your analytics package collects data and 2) how the links have been coded. In the case of Google Analytics, it treats all data that goes from one page to another as the same, if the links are the same (with a caveat explained in a second). This means that to Google Analytics, it can’t distinguish between the 4 links on the homepage in terms of how much traffic each sent to the Careers page. But there is hope… Google analytics allows you to add tags to links that can help you distinguish where traffic is coming from to the same web page. Which means that if the links were coded with these tags, the data will already be available. And if not, it can be if they are added. Other analytics tools may collect data differently and your content management system (CMS) can also impact how this works.

Step 4: Zero In on the Right Information

So now that we know what we’re trying to measure, what data refinements we need, and how our web technologies work, we can zero in on the right information in our analytics tool. In Google analytics, we’d look for traffic to the Careers page from each of the 4 tags on the homepage to provide information about what place on the homepage is sending the most traffic.

Good Measurement is In the Details

While this may seem complex, the first step — knowing what you want to know — is really vital for communicating your measurement needs to those that may help provide you with the metrics. Without this refinement, you may get back the wrong metrics, or your technologies may not be setup properly to provide them in the first place.