Building Better Content By Improving Upon Your Competitors
Posted by Bill.Sebald
In rock n’ roll music, stealing is expected. Led Zeppelin allegedly lifted from lots of earlier blues and folk artists. The famous I-IV-V chord progression of The Wild One’s song “Wild Thing” was used only a couple years later on “Mony, Mony.” My favorite example of musical larceny – “Let It Be” by The Beatles, “Farmhouse” by Phish, and “No Woman, No Cry” by Bob Marley are built around the exact same chord progression. Yet in all these cases, the songs were tweaked enough to stand on their own in meaning, served as distinct entities, and inspired unique feelings from the listener. Granted record company execs often disapproved, but some artists were often flattered to see interpretations of their riffs and progressions. At the end of the day, this is what spawned (and advanced) the rock music genre. Sometimes stealing is the engine of innovation.
�Your idea isn�t new. Pick an idea; at least 50 other people have thought of it. Get over your stunning brilliance and realize that execution matters more.� �Mark Fletcher of Bloglines.com.
In marketing, we don’t just “steal” the minds of consumers, we sometimes steal – and interpret – from our competitors. Sometimes we’re lazy about it, and sometimes we’re perceived as originals. Remember one of the immutable laws of marketing – always appear to be first. Well then why not be first to make someone’s content strategy more effective (for your own gain)?
Wait – so do I condone being a pickpocket, cat burglar, or politician? No. What I�m suggesting is reviewing what inspires you, analyzing why it was successful, and inspiring yourself to make something better. Better for us, better for our clients, and better for their customers.
Oh no; is this another “Content Is King” post?
I’m not a huge fan of that phrase anymore. SEO has gone through some serious developmental stages in its lifetime. Once the hype was all about “keyword density,” then “anchor text,” then “duplicate content;” now I feel like our latest bandwagon concept is the semi-vague “content is king.”
These are certainly all valid concepts in SEO, but without proper context, they often fall short of sound advice. They become blind directives. So here we are in 2014, with many business executives nodding along, “yes – content is king. I’ve read that a trillion times. We need to crank out 100 posts a month. Go, go go…” But I think this is a problem. Now that SEO is mainstream, there’s so much “good content” that the noise ceiling has simply been raised. I’ve said it before, “Fair-quality copy is becoming the new Google spam.” I go into pitches now where businesses can’t understand why their legacy content isn’t getting searches. In other words, they ask why “content is king” isn’t producing results. It’s usually because content was treated as a homogeneous tactic where a marketing or SEO strategy wasn’t put in place to link the pieces together.
I think it’s time SEOs put that phrase to rest, and start thinking in terms of how a traditional content marketer would think about it. “Content that is unique in value, strong in expertise, provides a necessary point-of-view, and leads the pack in terms of usefulness is more than king – it’s fundamental to success.” A bit of a mouthful (and less sexy), not to mention harder to develop, but it really needs to be adopted.
So if you would, please keep that in mind during this post. Continue on!
What are your competitors doing?
Content ideas come from lots of sources. Some are vapid (like content topic generators) and some are interpreted (like reviewing customer poll results). Often a simple interview with your sales or service team can teach you plenty about the mindset of your consumer. Studying on-page product reviews can also be inspiring. Focus groups, experiments; all this and more can help produce pieces of content that can be strung together and tracked in order to build a truly converting funnel.
We all know the most effective content is inspired by data, versus �crazy ideas� with no concrete evidence quickly thrown against the wall. While this occasionally has some SEO benefit (arguably less and less with Panda updates), it rarely does much for your conversion funnel. It takes that extra digging that some aren’t quick to execute (at least in my experience). But what happens when your competitor is willing to do the work?
That’s where you can learn some interesting things. Marketing espionage!
Granted, most competitors don’t want to share their data with you, no matter how much beer you try to bribe them with (believe me, I’ve tried). We have tools like SEMrush to estimate search metrics, and services like Hitwise and Compete to get more online visitor data. While that is certainly helpful, it’s still directional. But we’re marketers – so what do we do? We get creative.
How to get a birdseye view of a content play (with common SEO tools)
It’s time to lift the hood. I like to start with Screaming Frog. Most SEOs know this tool. If you don’t, it’s a spider that emulates what a search engine spider might find. In my experience there’s no better way to find the topics a website is targeting than with a “screaming” crawl.
Filter down to HTML, and you’ll find the URL, Title Tag, Meta Description, H1, and sometimes the Meta Keyword data. If you already have your own keywords and entities in mind, and want to see what a competitor is doing with them, it’s as simple as searching for them in Screaming Frog (or an excel export) and scanning for it.
Click for a larger image:
Consider this totally random “shammy” example in the screenshot above. If I worked in the shammy business, through a quick scan I might be interested to know that at least one of my competitors found value enough in creating a section around an iPad cloth. Is that a segment I never considered?
Don’t have Screaming Frog? The site:operator is a less powerful option. You can’t export into a spreadsheet without a scrape.
Ubersuggest or keywordtool.io can be used in clever “quick and dirty” way – put in a keyword you think there’s opportunity for, and add “who,” “what,” “where,” “why,” or “how” to the query. Your fragmented query will often show some questions people have asked Google. After all, plenty of great content is used to answer a query. Search some of these queries in Google and see what competitor content shows up! At the very least, this is a nice way to find more competitors who are active with creating content for their users.
At this point you should be taking notes, jotting down ideas, observations, potential content titles, and questions you want to research. Whether in a spreadsheet or the back of a napkin, you’re now brainstorming with light research. Let your brain-juice flow. You should also be looking for connections between the posts you are finding. Why were they written? How do they link together? What funnels are the calls-to-action suggesting? Take notes on everything, Sherlock!
Collect the right data
Next, step it up with more quantifying data.Time to trim the fat.
Search data
By entering and measuring your extracted in Google’s Keyword Planner, you’ll see not only is there interest in an iPad cleaner (where an “iPad Shammy” might make sense with its own strategy), but some searcher interest in the best ways to clean an iPad. That could be fun, playful content to write – even for a shammy retailer. It could tie directly to products you already sell, or possibly lead you into carrying new products.
Click for a larger image:
Estimated searches don’t tell the whole story. We know plenty of keywords and metrics from this tool are either interpolated or missing. I’ve found that small estimated searches can sometimes still lead to more highly-converting volume than expected. Keep that in mind.
Social data
What searches enter into Google’s search box isn’t the only indicator of value. Ultimately if nobody likes a certain topic or item your content, they aren’t going to share or link to it. Wouldn’t it be great to have another piece of evidence before you get to structuring a strategy and writing copy? That evidence may lie with your competitors’ social audience.
At this point you have keyword ideas, content titles, sample competitor URLs, and possible strategies sketched out. There are some great tools for checking out what is shared in the social space. Topsy, Social Crawlytics, and Buzzsumo are solid selections. You can look up the social popularity of a given URL or domain, and in some cases drill down to influencers. If it’s heavily shared, that may suggest perceived value.
Click for a larger image:
Look at the image above. If my agency is a competitor of yours, you might be interested that one of my posts got 413 social shares. It was a post called “Old School SEO Tests In Action (A 2014 SEO Experiment)”. You can dig in to see the debates boiling through the comments or the reactions through social media. You can go so far as see who shared the post, how influential these people are, and what kind of topics they usually share. This helps qualify the shares.
With these social metrics I believe It’s reasonably safe to infer people in the SEO space care about experiments, learning about things that move rankings, and that most believe older tactics aren’t worth pursuing. With very little time at all, you might be able to come up with ways to improve upon this post or ideas for your own follow up. Maybe even a counter argument? Looking at who the post resonated with, you could presume my target audience was SEOs with a goal of providing industry insights. With a prominent lead generation form on this post, you might even suspect a secondary interest was as a source of new client leads.
If you surmised any of these things from the social data, you’re 100% right! This was certainly a thought out post with those goals in mind.
Backlink data
Let’s examine link popularity and return to the shammy industry. Specifically let’s look at a pretty unique item – a shammy for Apple products – https://www.klearscreen.com/detail.aspx?ID=11.
- Open Site Explorer found 1 link from a retailer.
- Ahrefs found 8 links from 8 domains, one being a forum conversation on Stackexchange.com, and the others from a retailer.
- Majestic found 13 links from 6 domains. Similiar to what Ahrefs found.
- WebMeUp found 30 backlinks from 9 domains.
From this data it looks like the iPad shammy market isn’t exactly on fire. Now it doesn’t appear iKlear (or Klear Screen) is doing much marketing for this particular product – at least not according to Google. Their other Apple product cleaners seem to get more attention, but perhaps iKlear simply knows this isn’t a high demand product. It could be true – after all it hasn’t gone viral. It hasn’t generated much in the way of online discussions. But it also hasn’t been marketed much.
This is why all the data needs to be collected, correlated, and analyzed. You want the best hypothesis you can get before you start committing your time to a content strategy. Did this just kill a possible content strategy for an iPad Shammy, or is this a huge untapped opportunity? It entirely depends on how you interpret all the data you collect.
You’ve got some ideas; now what’s the execution?
You just did a lot of work. You can’t go off half-cocked throwing up willy-nilly content. Jeepers, no! The next step is the most crucial!
At this point you should have uncovered some great ideas based on your competitor’s clues. Now comes the part where you thoughtfully determine how to implement these ideas and craft a strategic roadmap. The options are endless, which could provide a decision-making struggle. From new microsites to overhauling existing content, there’s so much you can do with the gems you’ve dug up.
Remember to examine what your competitors did. How did they plug everything together?
But sometimes your competitors don’t have a discernible content strategy. Instead just fragmented content floating like an island. This is even better for you. Now you have opportunity to not only outshine in the actual content, but put together an actual experience that your users will value, thus providing a likely positive SEO result. Here are three options I tend to build a strategy around most often:
- Create a new funnel
- Create content for off-page SEO
- Create emphasis content
With fresh metrics, the new funnel is often necessary. Chances are you discovered uncharted territory (at least from your website’s perspective). All future or existing content should have pre-conceived goals – there’s a top and bottom to every funnel, and maybe some strategic off-ramps leading to forms, contact pages, or products. Remember, you’re goal is to be driving the reader through an experience, eliciting emotions and appealing to their needs of which you’ve already built a hypothesis upon. This new funnel can dip into your current website or run parallel (ie, a microsite, sub-domian, or otherwise disconnected grouping). The greatest thing about digital marketing is that nothing is in stone. It’s so easy to test these funnels and redesign with collected data when necessary.
Off-page is also very common (right link builders?). Find something that is popular, and go share it with sites more popular than yours. Maybe you can even start generating new popularity and create a segment of its own. Build a strategy to take this burgeoning topic and let the widest audience know about it. Get branding, mind share, links, and ideally profit like a beast.
The “emphasis content” (as I call it) has been a solid go-to plan for me when I discover small pockets of opportunity; notably the stuff that may have a smaller impact and isn’t worth a month long content strategy. If I were to create my own iPad shammy play, based on what I’m seeing so far, I’d probably think about a page or two as emphasis content.
This content is like an independent port of entry or landing page, either to an existing funnel or a direct money maker. In a previous post I talked about creating niche collection pages for eCommerce. That could serve as emphasis content to a parent collection, but I’m usually thinking of heavier use of text in this case. Where you really take your goal, slice it up, and provide nice, beefy communication about it.
This play can be nuclear. By creating these one-off pages based on all the metrics discussed above, it’s usually much easier to do targeted outreach and social marketing. A well placed page, providing well placed internal links (ideally off popular pages), can pass PageRank and context like a dream, A tool like Alchemy API can help you see the relevance of pages and help you determine the best place to publish this page
Summary
A content strategy doesn’t go far if it’s phoned in. Take all the help you can get, even if it’s from a competitor. Learn from businesses who took steps before you. They may have very well discovered the holy grail. Competitive research has always been a part of any marketing campaign, but scratching the surface only gets you superficial results. Look deeper to uncover more than just a competitor’s marketing plan, but the very reason why the competitor may be beating you in search. Then, hopefully you’ll become the rock star others are trying to copy from. That’s a good problem to have.
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!
Continue reading →Demystifying Data Visualization for Marketers
Posted by Annie Cushing
I presented on wrangling and demystifying the data visualization process for marketers at MozCon this year, and it turns out there was far more to talk about than could fit into that half-hour. For the sake of those who couldn�t make it and those who could but want to learn more, I pulled together this overview of my presentation, offering more detail than I could in the slides.
To see all of the links shared in this post, check out my MozCon Bitly bundle.
You may want to open the SlideShare file in another tab or browser window, so you can easily toggle between the post and the SlideShare.
I�m going to go through the presentation slide by slide to bring the narrative to print.
Slide 3
I have a confession: Although it�s probably safe to say I�m a fairly advanced Excel user � at least among marketers � until recently I had no real charting strategy. In fact, I signed up to do this presentation partly to force me to carve out a strategy, particularly with Google Analytics data.
Slide 4
In this presentation I have focused on Google Analytics data for a couple reasons:
- If you can wrangle Google Analytics, other marketing data is a walk in the park.
- It has naming conventions that map beautifully to Excel, making it an ideal tutor.
Slide 5
My approach may seem a bit Karate Kid-esque, but if you can grasp the interplay between Google Analytics and Excel, you�ll never be left wondering how to visualize your data.
Although there are many aspects to data visualization, I focus primarily on charting.
Slide 6
In Excel there are two components to charts that are critical to understand: data series and categories. They are always used together.
Think of categories as buckets for your data and data series as the data itself.
Slide 7
If you dumped a pile of Legos in front of a group of kids and told them to organize them by color into their corresponding, labeled containers and then count them, the containers would be categories. And the data series would be the count of Lego bricks.
Slide 8
First let�s peek under the hood on a PC by cracking open the Select Data Source dialog. You get to it by right-clicking on your chart and choosing Select Data.
Slide 9
Excel for Mac also has data series on the left and categories on the right. And that�s about all they have in common.
Slide 10
But, as with most features in Excel for Mac, the functionality of the Mac�s Data Source dialog is far inferior to that of the PC.
Slide 11
This sort option is helpful if you have a stacked chart and want to sort the individual data series. I like to put the larger series on the bottom and smaller ones on the top. But if you have a stacked chart on the Mac and you want to reorder the data series, you actually have to delete the series you want demoted and manually add it back in.
It�s kind of like that game, Hand on Hand, you might have played as a kid where kids go around in a circle putting their right hands in the middle, followed by the left hands. Then they go around the circle moving the bottom hands to the top of the pile as fast as possible.
Although in this case, you�re moving the data series to the bottom of the pile.
Slide 12
To move the Sessions data series to the bottom of the pile, first select it from the Series list.
Slide 13
Then click the Remove button to delete it from the list.
Slide 14
Then click the Add button to add it back to the list of data series.
Slide 15
Click the data selector button to the right of the Name field and select the series name, as directed in the screenshot.
Slide 16
Click the data selector button to the right of the Y values field and click-and-drag over the values. If the column is long, just click the first cell and press Ctrl-Shift-Down Arrow (Mac: Command-Shift-Down Arrow) to select the entire column without scrolling. (We are nothing if not efficient.)
Slide 17
And finally you need to click-and-drag over the category axis labels. Which brings us to the Mac�s other issue �.
Slide 18
In the PC version, there�s one place for the category axis labels. On the Mac you have to choose the axis labels for each series. It�s counter-intuitive.
Slide 19
Categories end up along the horizontal axis � or the vertical axis for horizontal bar charts.
Slide 20
The data series ends up in the legend and is usually a metric (from GA). But there are a couple exceptions, which we�ll get to in a minute. The categories populate to the horizontal axis label or vertical axis label with the bar chart.
Slide 21
Transition to Google Analytics.
Slide 22
The two major players in Google Analytics � that we�ll be mapping to Excel � are dimensions and metrics. They�re (practically) inseparable.
Slide 23
Dimensions are the buckets your data is broken up into. These come into Excel as text � even if they�re values � like you get with the Days to Transaction dimension (which you can get from Conversions > Ecommerce > Time to Purchase). They are always the far-left column of the table.
- Add a secondary dimension in any report (standard or custom).
- Create a custom flat table with two dimensions. Learn how in this post.
- Use the API. This is the only option that will allow you to use more than two dimensions. You can pull up to seven dimensions in one API call.
Slide 24
Metrics are anything that can be measured with a number.
Slide 25
If you�re in a custom report (or have clicked the Edit link at the top of most standard reports), metrics always show up to a party in blue.
Slide 26
And dimensions show up as green.
You can learn more about custom reports from the video tutorial I created to help marketers.
Now it�s time to marry Google Analytics and Excel.
Slide 27
In most cases dimensions in Google Analytics map to categories in Excel.
Slide 28
And metrics map to data series in Excel.
Slide 29
I�m going to break this down systematically, based on the number of dimensions and metrics you�re wanting to visualize.
Slide 30
Dimensions: 0
Metrics: Multiple
You want this if you want to know aggregate numbers, e.g, sessions for the month, or revenue, or goal completions.
Slide 31
I hate to start on a downer, but you need the API to do this. The GA interface requires at least one dimension.
Slide 32
As with most things, if you prod enough, you�ll discover hacks and workarounds. But the name of the game here is to come up with a dimension that will only have one bucket. Going back to the Legos analogy, it would be kind of like saying, �Put all the plastic Legos in this bucket and count them.�
Slide 33
Workaround: Set dimension to something that will encompass all of your data, meaning you�ll only have one row in the report. One example of that would be the User Defined dimension (under Audience > Custom > User Defined).
As you�ll see in the screenshot, all of the values are consolidated as (not set) since this profile (now called view) doesn�t use the User Defined dimension.
Slide 34
If you�re still using the User Defined dimension (and, therefore, have multiple rows reporting), you really need to update.
If you�re using classic GA, you should be using custom variables and custom dimensions if you�re using Universal.
Slide 35
Another option is to use the Year dimension with a custom report. This is ideal if you are gathering data for a single month. You can aggregate data beyond one month, as long as the date range you choose doesn�t straddle more than one year.
Slide 36
Here’s what the custom report looks like under the hood. Learn how to create custom reports in Google Analytics in a video tutorial I did.
Slide 37
You can access this report here while logged in to Google Analytics.
Slide 38
This data isn�t conducive to charting, but you can sexy up a table with sparklines and conditional formatting.
Slide 40
Dimensions: 1
Metrics: 1
An example of this might be revenue segmented by country or bounce rate segmented by device category.
Slide 41
Pie Chart Basics
Here are some highlights about the pie chart:
- They use angles to show the relative size of each value.
- You should put data in descending order to put the most significant data point at 12:00 and radiate clockwise.
- Avoid 3D pie charts. They distort data.
- Data points must add up to 100%. So you can�t take traffic from 5 of your 8 campaigns and chart them.
- Microsoft says no more than seven categories; I say no more than five.
- None of the values in your data series can be negative.
- Learn more
Pie Chart Tricks
Ways to trick out your chart:
- You can grab a piece of the pie to isolate it and drag it out slightly to draw attention to it. This is called exploding pie pieces.
- You can also change the values to percentages in the data labels or even add the categories, thereby negating the need for a legend.
Slide 42
Donut Chart Basics
Here are some highlights about the donut chart:
- Donut charts show data in rings, where each ring represents a data series
- It uses the length of the arc to indicate the size of the value.
- You should put data in descending order to put the most significant data point at 12:00 and radiate clockwise.
- Data points must add up to 100%. So you can�t take traffic from 5 of your 8 campaigns and chart them.
- Microsoft says no more than seven categories; I say no more than five.
- None of the values in your data series can be negative.
- Learn more
Donut Chart Tricks
Ways to trick out your chart:
- You can put the title or the value you want to highlight in the center.
- I don�t recommend using the donut chart for multiple series or dimensions. They�re more difficult to interpret.
- Like the pie chart, you can pull one out to draw attention to it.
- You can use a donut chart to create a speedometer chart.
- You can fill it with an image that resembles the surface of a donut to make it look like a � Okay, yeah, never mind �
Slide 43
Column Chart Basics
- Should sort in descending order.
- The axis should start at 0.
- Categories don�t have to add up to 100%
- Learn more
Column Chart Tricks
- You can add a trendline to make trends stand out.
- Consider going totally minimalist with the techniques I demonstrate in this video tutorial. (You can skip to the 15:53 mark.)
- Don�t be afraid to move the legend around.
- Excel�s default axis tends to be dense. I typically double the Major Unit, so if the major unit is set to 100, I typically up it to 200. Learn more about the major unit from the Microsoft site. (But I also show how in the above-mentioned video tutorial.
- You can use a column chart to create a bullet graph to show current data vis-�-vis goals or projections.
- You can use a column chart to create a waterfall chart.
- You can add a target line to your chart.
- If you have many categories to chart, you can use a scrollbar.
- You can use a column chart to create a thermometer chart.
- Just remember safety first when working with column charts.
Slide 44
Bar Chart Basics
- You need to sort your data in ascending order to put the longest bars at the top.
- Bar charts are good for categories with longer labels.
- You shouldn�t use bar charts if your dimension is time based (date, month, etc.).
- Learn more
Bar Chart Tricks
- You can use all of the tricks (except the last two) listed in the Column Chart Tricks list.
Slide 45
Radar Chart Basics
- Category labels are at the tip of each spine.
- You can use a fill with your radar charts.
Radar Chart Tricks
- Radar charts can be compelling when you compare multiple entities at once. For example, I saw a set of 50 radar charts that compared metrics like crime rates for different types of crime for each state.
- If you don�t want the axis labels to show, you can set the number formatting to ;;; to hide them altogether. You can then include an annotation on your chart that lets viewers know the intervals.
Slide 46
Notes about the Heat Map
Learn how to create a heat map in this video tutorial I did.
Slide 47
And now let�s look under the hood at a typical chart that uses 1 dimension and 1 metric. Let�s say we have this table of analytics data �.
Slide 48
If we create a column chart from this table, this is what it�s going to look like (with some cleanup).
Slide 49
Now if we look at the data source this is what we�ll see �.
Slide 50
The mediums show up over here in the categories �
Slide 51
And the sessions values show up here in the data series �
Slide 52
Which populates to the legend. But you can delete the legend when you only have one metric (or data series). You�ll then want to include the metric in the chart title.
Slide 53
And the mediums populate the horizontal axis labels.
A little piece of Excel trivia: The Select Data Source dialog still says Horizontal Axis Labels, even for bar charts where the labels are on the vertical axis. #pedantic
Slide 54
Example of 1 dimension and multiple metrics: Sessions, goal completions, and revenue broken down by Device Category (mobile, tablet, desktop)
BTW, the Device Category dimension is one of the most important in Google Analytics. By itself it�s pretty useless, but in the context of other data, it�s very useful. You should be segmenting all of your data by it.
Slide 55
Notes about the Clustered Column Chart
- Clustered column charts are good for showing comparisons (e.g, sessions vs revenue for each month or ROI vs Margin by campaign (or keyword).
- You could transform the clustered column chart into a combination chart by adding a line chart on the secondary axis that adds a percent value.
Slide 56
Notes about the Stacked Column Chart
- The stacked column chart is good for showing how each data series contributes to the whole.
- An example might be revenue broken down by medium.
- If you want to order the columns by overall height, you can create a total column for the series. You just won�t chart that column.
Slide 57
Notes about the Clustered Bar Chart
- All of the notes in the above-mentioned stacked column chart.
- Like the [single] bar chart, the clustered bar chart is better for categories with long labels.
- You can hack the clustered bar chart to create a double-sided bar chart. You can view a video tutorial I did on how to do this.
Slide 58
Notes about the Stacked Bar Chart
- If you want to sort the bars so that the longer bars are on top, create a totals column and sort it in ascending order.
- You shouldn�t use the stacked bar chart if your dimension is time oriented (date, month, etc.).
Slide 59
Notes about the 100% Stacked Column Chart
- Use the 100% stacked column chart when you are working with percentages.
- The data series must add up to 100%.
- For example, if you wanted to see what percentage of social referrals came from desktop, tablet, and mobile devices.
Slide 60
Notes about the 100% Stacked Bar Chart
All of the notes under the 100% stacked column chart apply here.
Slide 61
Notes about the Radar Chart
- Category labels are at the tip of each spine.
- You can use a fill with your radar charts.
- Radar charts can be compelling when you compare multiple entities at once. For example, I saw a set of 50 radar charts that compared metrics like crime rates for different types of crime for each state.
- If you don�t want the axis labels to show, you can set the number formatting to ;;; to hide them altogether. You can then include an annotation on your chart that lets viewers know the intervals. See the screenshot under the Slide 45 note above.
Slide 62
Notes about the Combination Chart
Learn all about combination charts in this post I wrote on the Search Engine Land site.
Slide 63 � 69
Self-explanatory as they follow the same dialog as slides 46 � 52.
Slide 71
Notes about the Line Chart
- In a line chart, category data is usually distributed evenly along the horizontal axis and value data is distributed evenly along the vertical axis.
- Line charts can show continuous data over time, so they’re ideal for showing trends in data at equal intervals, like months, quarters, or fiscal years.
- You can add markers and set the lines to none to use them in ranking charts.
- Avoid using stacked line charts. It�s not always apparent that the data series are stacked. If you want to stack, use an area chart instead.
- You can add interesting line markers like the ones I created in this video tutorial to replicate the charts in Moz�s tool set.
Slide 72
Notes about the Stacked Area Chart
- Ideal for showing stacked data series over time, especially if you want to demonstrate a fluid trend. Stacked column charts should be used if you want to keep each of the categories more disparate.
- You should order the data series so that the larger series are at the bottom of the stack with the smaller series being clustered together at the top because people�s eyes naturally travel from the horizontal axis upward with stacked area charts.
- If you keep the gridlines, make them significantly lighter. A light gray works well.
- Make sure you have adequate contrast between contiguous data series. Sometimes Excel puts two colors next to each other that blend.
- If you have smaller data series that are difficult to see, use stronger colors to make them easier to view.
- If you have all larger data series and you want to add some finesse, give your data series a line (what would be called a stroke in graphic design programs) that�s slightly darker than the fill.
- You can create a combination chart with a stacked area chart. Just don�t use a line chart for the other style. I like to use a chart style that stands out from the area chart, such as a column chart. You may want to increase the transparency of its fill so that you can easily see through to the stacked area chart.
Slide 73
Notes about the Clustered Column Chart
- You use the clustered column chart to show comparisons between data series (as opposed to how they contribute to the whole).
- The clustered column chart is especially effective for showing year-over-year data. The categories would just have the name of the month (I abbreviate to three letters, which you can learn how to do in this tutorial), and one column would be used to show data from one year and the other colored column would indicate the previous year. To show the month from each year as a disparate data series, you would have to make each year a separate column in your data.
- You can add a line chart on the secondary axis that highlights the percent change between values.
- You can play with the gap width and overlap settings to adjust the series. You get to those by selecting a column, pressing Ctrl-1 (Mac: Command-1), and navigating to the Series Options (Mac: Options) area of the Format Data Series dialog.
- Excel doesn�t provide the option to add a data label that indicates the total of all the data series for each column. You can hack one by adding a total column that you include in the clustered column but then change to a line chart. From there, remove the line and add data labels above the line.
Slide 74
Same as Slide 60.
Slide 75
Same as Slide 58
Slide 76 � 77
Self-explanatory.
Slide 78
Things get more complicated when you want to chart two dimensions. There are three ways to get 2 dimensions:
Slide 79
So here we have two dimensions (Device Category and User Type). I picked these dimensions to demonstrate because they have a finite number of options. I LOVE the device category dimension and use it frequently to segment my data in Google Analytics.
Note: When you chart two dimensions, you can only use one metric (or data series in Excel).
Slide 80
Here�s an example of what a clustered column chart might look like.
Slide 81
We now have a dimension in the legend � or category in Excel.
Slide 82
Using the Switch Row/Column button �.
Slide 83
This is what the chart would now look like. Notice we now have three data series and two categories.
Slide 84
Now let�s take a peek under the hood.
Slide 85
Again, here you see we have dimensions, not metrics, in the data series. The metrics should be included in the chart title.
Slide 86
And now the Device Category dimension is in the category area.
Slide 87
Your chart options are the same as when you had one dimension and multiple metrics. These options are not exhaustive.
Slide 88
Slide 89
The data in this table is in report format. If only marketing export data came in this format. (It doesn�t.)
Slide 90
This is how marketing data actually comes out of different marketing tools. It�s called tabular format.
Slide 91
Just as in a database, rows in tabular data are called records.
Slide 92
Columns are called fields.
Slide 93
And the column headings are called field names. But if I were to select two dimension columns and one metric and select a chart, here�s how Excel digests the data �
Slide 94
Gross, I know. I�m a child.
Slide 95
Here�s what it actually looks like. A royal mess.
Slide 96
Excel requires that data be in a report format in order to chart two dimensions. And the one metric (sessions, revenue, impressions, whatever) goes into the green area. There’s only one way to corral an export with two dimensions and one metric into report format …
Slide 97
Pivot tables sound scary and intimidating but not if you think about what pivot means.
Slide 98
When a soldier pivots, s/he very simply goes from standing facing one direction to turning at a 90 degree angle. That’s what a pivot table does. By moving one of your dimensions into the Columns field (Mac: Column Labels field), Excel puts that dimension’s values across the top of your data.
Once you have your data in report format, and you can chart it. You typically want to put the dimension with fewer values into the columns area.
Learn how to create pivot tables in this comprehensive video tutorial I did.
Slide 99
Although pivot tables come with a lot of junk in the trunk, you can see the pivot table puts the data into report layout, which Excel can then use to chart the data. If you�re on a PC, you can create a pivot chart. If you�re on a Mac, you can create a static chart from the pivot table because Excel for Mac still doesn�t support pivot charts. Still. Ridic.
Slide 100
Now you�re ready to look at GA data � nay, all marketing data � with a more strategic eye� And spend less time tooling around in Excel trying to figure out how to visualize your data!
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!
Continue reading →The New Link Building Survey 2014 – Results
Posted by JamesAgate
Many of you may have seen Skyrocket SEO’s Link Building Survey results that we published here on Moz around this same time last year. The reception was fantastic, so we decided to push ahead with turning this into an annual series to see how this strand of the industry is developing and evolving over time.
Firstly, “link building”…
Yep, we’ve not changed the name to a “content marketing survey” or “inbound link acquisition survey;” we still feel link building is a vital part of an SEOs arsenal of tactics, and therefore it deserves its own survey.
As a company we’re investing just as much in link building for our clients (granted, we’ve adapted what we are doing), but the fact remains that if you want to score big with decent organic search visibility then you need links.
Now that that’s out of the way, let’s get down to the details:
Who took the survey?
A massive thank you to the 315 or so people who took the survey. That number is slightly down from last yeah, which I feel is partly due to fewer people considering link building to be a part of their day-to-day roles. I’d argue that’s a missed opportunity, and this year we had a few duplicate entries and submissions that needed a bit of tidying up, so we trimmed it back to these 315 submissions.
The makeup of the respondents was broadly similar to last year, as expected, although based on user feedback from our inaugural survey, we added a few more categories for respondents to self-classify�so it is hard to make specific comparisons.
How much does your company spend on link building per month?
In the 2013 survey, 10% of respondents said their company spent $50k+ per month on link building, so it appears that the upper limit to link building spend may have decreased slightly across the industry.
That being said, there now appears to be a much larger number of companies in the $10-$50k per month bracket when you compare this year’s 37% with last year’s 11%.
I would attribute the changes year-on-year to two factors;
- Reclassification of the term “link building:” Many companies have shifted budget that they would previously classified as link building budget into content projects that more than likely still have an impact on link building efforts.
- Recognition of opportunity: Based on our own experiences we see a number of website owners and businesses pushing harder with their content promotion and link building as they recognise an opportunity to invest when their competitors are running scared.
Warren Buffett once said “Be fearful when others are greedy and greedy when others are fearful.” Based on conversations alone that I’ve had with a wide range of businesses, many are now fearful when it comes to building links. In fact, we gathered some data later in the survey that revealed that one of the biggest challenges people face is not knowing which links will help and which will harm them. Google’s widespread action against websites (and dare I say it webmaster propaganda) has had a dramatic impact on some people to the point of paralysis.
There are clear opportunities that, with a sound strategy, can be seized in today’s market.
You can build links like it’s 1999 for a microsite or second level property, keep it super-clean and identify link opportunities that would be valuable irrespective of Google, or somewhere in between those extremes. The fact is the links still form the backbone of the internet and of Google’s algorithm and that isn’t going to change for a very long time.
What percentage of your overall SEO budget is allocated toward building links?
Thanks to John-Henry Scherck for this one as he made the suggestion following the 2013 survey that having data on the percentage would be really interesting. Looking back we don’t have a point of comparison but not of course moving forward we will have so we should get a clearer picture of whether online marketing budgets are just increasing in general (and therefore link building gets allocated the same percentage but of a bigger pie) or whether folks are seeing the value from building links and therefore allocating a larger percentage of the same sized pie to link building activities.
Would you say you’ve increased or decreased your spend on link building over the past 12 months?
This aligns with our data on more people entering the $10-$50k per month investment bracket this year:
Why the increase/decrease in spending?
We asked people why they decided to increase or decrease their spending on link building over the past 12 months.
Responses could be categorized into the following areas:
Common reason for increases:
- Increased costs related to moving away from older style and often “cheaper” link building
- Increased costs related to production/creativity
- Good links are just as important as ever; links still move the needle in terms of search engine visibility and performance therefore it makes sense to increase investment in this area.
Common reasons for decreases:
- Moving link building budget into content marketing projects (to be fair, this budget will probably indirectly fund link acquisition of some kind even if it is seen as a secondary goal for the content campaign.)
- We wanted to scale back and assess the impact that Google’s manual actions etc have on our website.
In the next 12 months, will you look to increase or decrease your spend on link building?
Why the planned increase/decrease in spending?
- Link building continues to get more expensive
- To raise the bar on existing efforts, and to beat competitors with increasingly sophisticated content assets
- Unsure where to invest/which links are working so concentrating budget into other activities.
Which link building tactics do you utilise most often?
(Numbers listed are votes rather than percentages)
When we compare with responses from the 2013 survey, there is a clear shift towards content-led initiatives and a reduction in some tactics for example close to 50% said in 2013 that guest blogging was their staple tactic, in 2014 fewer than 15% listed it as one of their staple activities.
Another interesting bit of data is the fact that paid links have seen somewhat of a resurgence in popularity, presumably as companies look for tactics where they can maintain greater control. In 2013, just 5% listed paid links as their staple linking tactic whereas in 2014 over 13% reported paid linking and blog networks as one of their main link building tactics.
What is currently your biggest link building challenge?
- Getting links to pages that aren’t particularly linkworthy (money pages)
- Lack of scalability (time, process, training, spreading time between clients)
- Avoiding Google penalties
These are similar challenges to those reported in 2013 in the sense that there is still concern over which links are helping and harming organic search performance as well as difficulties relating to processes and the lack of scalability.
The interesting thing is that SEO is full of challenges so as soon as one is overcome, the next appears. In 2013, 28% of respondents said that “finding link prospects” was a key challenge but this year not a mention of link prospects being an issue. This arguably suggests that we as an industry were adjusting to the “new world” back in 2013 and that now we have advanced our capabilities enough for this to now longer be the primary challenge in our day to day work. Now the main problem doesn’t seem to be getting links as such but more about getting links into the pages that we all need to rank to stay in business � the money pages.
Which link building tactics do you believe to be most effective?
(numbers below are “votes” rather than percentages)
Which link building tactics do you believe to be least effective?
(numbers below are “votes” rather than percentages)
Which link building tactics do you consider to be harmful to a site?
(numbers below are “votes” rather than percentages)
See the complete visual below:
Thank you for everyone who took part in the survey! See you all again next year.
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!
Continue reading →Get Ahead of Google with Insight into Semiotics
Posted by Isla_McKetta
Write it and they will come. That’s the drum we’ve been beating for a long time now. We optimize our pages and our content to please search engines and cross our fingers and hope that customers will convert.
We can do better.
But to do it, we have to think beyond Google. Yes, you still need to check all your standard SEO boxes to make your site crawl friendly. Then it’s time to stop catering to the bots and start catering to the users instead.
That means we have to�no, we get to�think bigger when we think of SEO. As Rand said in his Whiteboard Friday last week, “SEO is really any input that engines use to rank pages.” That’s why we have to reexamine the way we design, the way we create, and the way we optimize. Most importantly, we’re going to have to reconsider the underlying logic we use to approach all three of those activities as we learn to think of the user first and the bots second.
This idea of blending search and user optimization isn’t new. But when Gianluca Fiorelli called for a shift from semantic to semiotic thinking on State of Digital, he got me thinking about whether semiotics are the next step in earning the audience you want.
What the heck is semiotics?
Semiotics is the study of the creation of meaning. Semioticians look at everything�words, images, traffic lights, kinship structures�and study what those signifiers (signs or anything that signifies anything) mean and how people create meaning from those signs.
Semiotics is composed of three parts: syntactics, semantics, and pragmatics. When we’re approaching user optimization from a semiotic point of view, we’re shifting from a focus on semantics to an incorporation of all three elements.
Let’s get to know them.
Syntactics (form)
Syntactics (more commonly called “syntax”) is the study of the formal relationship between signs. Think of syntax as dealing with grammatical rules, form, and spatial order. Syntax is why you place “inurl:” before the url in a query instead of after. Syntax can be as arbitrary as the order of lights in a traffic light, but it is unchanging.
In grammar, syntax is why you say “oranges are good” but Yoda says “good are oranges.”
Syntax is so embedded in search these days that we don’t even talk about it, and as long as your code is in the right order and the content on your pages is written for users who aren’t Yoda, you’ve mastered syntax. Hooray!
Semantics (meaning)
Semantics is the study of conventional meaning. Let’s take the word “orange.” It can mean either the fruit or the color.
Whether or not you use semantic markup, search engines are usually capable of reading the context on a page and returning a result for either the fruit or the color, depending on the parameters you entered. Crawlers have been using things like context, synonyms, taxonomy, and information architecture to determine the relevance of search results for a very long time. When Hummingbird came along, the semantic nature of search became more obvious because we could see that Google is looking at queries and not just keywords.
If you’re keeping score, we’re already thinking about and optimizing for two elements of semiotic thinking. And we’ve caught up with the latest algorithm updates. But syntax and semantics aren’t the whole story when it comes to how humans create and understand information.
Enter pragmatics.
Pragmatics (use)
You (and your customers) bring a whole life’s experiences into any interaction whether it’s reading a website or chatting someone up at a cocktail party. Those experiences shape the way you interpret images and words.
For example, if you’re a soccer fan, the way you fell about the word “orange” could be affected by how much you like or hate the Dutch national team whose nickname is “Oranje.”
And if you’re color blind, “orange” could mean any of these colors depending on the exact type of color blindness you have:
“Orange” also has political connotations:
Photo of Orange Revolution courtesy of Wikipedia user Irpen.
The point is that search engines know the dictionary definition of a word. They can even learn about the associations you have by the search terms you enter. But they do not inherently understand (yet) the richness of your personal relationship with a word and the myriad other factors that go into creating meaning for you.
Pragmatics is your opportunity to create a site that engages with all of those connotations in order to create a stronger bond with your customers.
Knock, knock.
Who’s there?
Banana.
Banana who?
Banana.
Banana who?
Orange.
Orange who?
Orange you glad I didn’t say “banana?”
Pragmatics in action
Pragmatics is also a way of describing how complicated our relationship with information inputs is.
Say you see something crazy in your Facebook feed like an article claiming, “Solar Panels Drain the Sun’s Energy, Experts Say.” Your job is to decide whether to share, comment on, or ignore that link. First you have to understand what it means, which in this case is figuring out if it’s good science, bad science, or satire.
Here is the process a human might go through as you use pragmatic interpretations to figure out how not to sound like a dope when replying to this post.
1. Consider the source
The article is from the National Report, which is not a household name. If it was from The New York Times, it might be time to panic, but in this case, you’ll want to dig a little deeper.
2. Evaluate the content
Human thought is remarkably complex and here are just a few of the signs you might consider while trying to make sense of this article:
Signifier | Conservative? | Parody? |
---|---|---|
Name of publication | Seems staunch enough. | Never heard of it, but it sounds a lot like the National Review. |
Tagline | Lots of people think they’re independent. | But calling it out? |
Overall look | Clean without spammy ads. | Wait, how do they make money? |
Endorsers | Conservative darlings. | But if you were going to parody someone� |
Article title | Fuzzy science? | Too crazy to be real. |
Source of study | Privately-owned think tanks produce all kinds of results. | Their site has even more crazy “science.” |
Tone | Straightforward reportage. | Too straightforward. |
3. Check the internet
It seems like this article is probably satirical, but to be safe, you can do what a lot of us do�Google “National Report” (and no, the irony of using to a search engine to prove that human users can make better connections than search engines is not lost on me). And then ask Wikipedia.
You could have made a decision about this article on a syntactic level (the sentences made sense even though the content seemed farfetched). You could even have interpreted it on a semantic level (both Googling the article and the Wikipedia search).
But what many readers need to fully understand this article is the pragmatics of assessing the signs.
So that’s a pretty deep dive just to decide to ignore a Facebook post. But the point is that your customers do this all the time, and the huge number of factors that go into showing us whether we should engage with your site and its content are more than search engines can currently look at.
That’s semiotics. The whole bundle of syntax, semantics, and pragmatics. And we’re doing pretty well with two parts of it, but there’s still a lot of opportunity in pragmatics.
Incorporating semiotic thinking into your web design and content
To recap: search engines aren’t sophisticated enough to know what pragmatic associations your customers bring into a search, but your customers are naturally bringing in layers of context, preferences, and life experience. Which means there are many layers on which you can engage with a customer that search engines can’t yet understand.
Here are some examples of ways to use pragmatics to connect with your audience.
1. Use satire or other humor
As with the solar panels article, some stuff on the internet seems too crazy and stilted to believe until you put it in context. The Onion has mastered this (and they have the engagement to show it). Robots don’t get humor, but humans do, and being funny (when appropriate) makes your site memorable.
2. Build a lexicon for your content
Use a lexicon (a list of commonly used words, slang and/or jargon specific to your audience) to understand the (rapidly evolving) way that your customers speak and communicate with them in their own language. Think about your users and what the words you’re using signify for them. Are they hearing the same things you are saying? If not, fix it.
3. Consider culture in your design
Connect with your audience by designing a site that speaks to their ideas of beauty and the way they process information. See how the US version of Shu Uemura’s site is clean and spare like many American sites (or, for that matter, Wyoming)?
Meanwhile the Japanese version showcases more information in a compact space (kind of like downtown Tokyo).
What I love about this example is that the brand aesthetic carries across cultures�only the way that brand is interpreted that changes. Cultural considerations can include anything from views on gender to perceptions of color. For example, in parts of Asia, purple is associated with luxury, while in the US it’s associated with low prices. Check out this excellent slide deck by Smith Prasadh to learn more about how differently humans can see the world (and how you can use that to connect with your audience).
4. Capture tangential relationships
Engagement doesn’t have to be about your product. Just take a look at what Emirates, a major sponsor of the World Cup, did in customizing their hero image for each target market. The global English version is pretty straightforward.
Things get more personal for Chilean visitors as Emirates tailors not just the flag, but also the copy (using the English version for consistency).
But the best, most customized version of this campaign is the one created for Brazilians. It’s so tailored, in fact, that I had to look up a couple of things. The stripes on the flight attendant’s cheeks are not the Brazilian flag, but instead represent the colors of the Brazilian team. And “Little Canary” is a nickname for the team.
I’ll bet that Google doesn’t care one single bit about these customizations. Even if they can read the text on the images. But my guess is that Emirates has scored a major goal in terms of customer “team” feeling with this campaign which should increase their direct traffic.
5. Incorporate metaphor into your design
Tired of the same old templates and stock photos? Your customers are too. Use images to evoke metaphor like Austin-based Write Bloody Publishing does here to capitalize on the do-it-yourself feeling of the Wild West.
Think about what makes your company unique and own that story with your design. It will make you stand out from the crowd.
Another way to do this is to reconsider your site nav with an eye toward metaphor. Maybe you’re a game company like 2K Games and you want your customers to feel like they are already immersed in your game, say BioShock, as they interact with your site. The first step would be to build a navigation that encourages that kind of feeling. Have your user enter the site as they would enter Rapture�through the bathysphere. Showcase game add-ons as plasmids. And use cutscenes to hint toward exciting features on the site just as you would in the game.
As long as you don’t throw your SEO training out the window, it’s okay to try something new and see if it speaks to your customers. If it doesn’t, try something else. As Lindsay Wassell said yesterday at MozCon, “The internet rewards innovation. Search engines reward innovation.” Be that innovator.
Those are just a few examples. The opportunity in thinking semiotically as you design, create, and optimize is to engage with your customers on a human level. This naturally builds your brand affinity, which should increase your traffic.
I’d love to hear about how you’re using pragmatics to build nuanced relationships with your customers.
Your mission
Let loose your creative team. No one wants to be an SEO copywriter or an SEO designer. When you’re optimizing a site in any way, think first about the user�the one with the most sophisticated relationship�then make sure that your standard SEO boxes are checked. Anything less is like dumbing down a parallax experience to a simple sketch to make sure Google fully understands it fully.
Now go off and use pragmatics to relate to your customers in such a way that so many customers come to your site and engage in such great numbers that the search engines chase you trying to figure out how you did it. You’ll be prepared if Google’s algorithm ever learns how to account for pragmatics, and it beats you chasing rankings any day.
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!
Continue reading →