Taxonomies are important because they help us speak our customers’ language and help us organize our content in common ways. When we tag co...
Taxonomies are important because they help us speak our customers’ language and help us organize our content in common ways. When we tag content with centrally controlled values, we help content management systems and other applications serve more relevant content to our audiences. But using the T word often causes executives’ eyes to glaze over and prompts those execs to start remembering meetings that they are already late for. Please don’t close the book yet. Taxonomy management is really not that hard.
A taxonomy is all about naming things as a company. That is a central part of marketing. If you asked brand marketers if naming products is essential, they would of course say yes. A taxonomy just organizes these names, making it possible to understand how two names are related, which is especially important when a name needs to change. Most importantly, taxonomies help us name things consistently. You allow taxonomies to do their job when you follow three essential approaches: Use data to decide wording. When we have debates about wording, we get sucked into the maelstrom of subjective semantics. If your taxonomy systems are informed by data rather than opinion, debates don’t regress; they are informed by the data. A taxonomy is just a mapping between the common words in our vocabulary and the way we want to put our stamp on them in our corporate nomenclature. If the core word you use comes from an external data source, such as keyword research or an accepted industry classification scheme, arguments about semantics evaporate. Also, external data sources tend to change less often than brand names. For these reasons, data-driven taxonomies are easier to manage and more effective.
Use ontologies to organize your taxonomies. An ontology describes the relationships between different names in a taxonomy rather than the meanings of the names themselves. Ontologies are as to links as taxonomies are to pages. Links provide context for web pages, so that Google can understand the purpose and relative authority of a page in context. Similarly, if you use ontologies, you won’t have to focus so much on the meaning of individual names, and you can build relationships between related names to use different names in different contexts. The mappings between core data-driven words and corporate brand names are examples of ontologies.
Use objects, not rigid hierarchies. We have a natural tendency to want to build hierarchies of names: Products go up to brands, brands go up to categories, and so on. The problem is that a lot of names fit under multiple branches of a tree. Is cloud security under the cloud category or the security category? The correct answer is “yes.” But hierarchies force us to choose to put them in either one branch or another —or have multiple variations of the same value. Object-oriented ontologies solve this problem by focusing on relationships that are not hierarchical. These relationships become part of the intelligence that you build into each node of your taxonomy.
It’s not easy. But if you know what kind of tools you need and you hire really smart people to use them, you can have a taxonomy that works for your digital content efforts. When we say it works, we mean it is relatively easy for authors to use the appropriate tags for their content. When they do, it’s possible to build adaptive, personalized, iterative, reusable, shareable, and findable content. Without good tagging, it’s virtually impossible to do any of this.
A taxonomy is all about naming things as a company. That is a central part of marketing. If you asked brand marketers if naming products is essential, they would of course say yes. A taxonomy just organizes these names, making it possible to understand how two names are related, which is especially important when a name needs to change. Most importantly, taxonomies help us name things consistently. You allow taxonomies to do their job when you follow three essential approaches: Use data to decide wording. When we have debates about wording, we get sucked into the maelstrom of subjective semantics. If your taxonomy systems are informed by data rather than opinion, debates don’t regress; they are informed by the data. A taxonomy is just a mapping between the common words in our vocabulary and the way we want to put our stamp on them in our corporate nomenclature. If the core word you use comes from an external data source, such as keyword research or an accepted industry classification scheme, arguments about semantics evaporate. Also, external data sources tend to change less often than brand names. For these reasons, data-driven taxonomies are easier to manage and more effective.
Use ontologies to organize your taxonomies. An ontology describes the relationships between different names in a taxonomy rather than the meanings of the names themselves. Ontologies are as to links as taxonomies are to pages. Links provide context for web pages, so that Google can understand the purpose and relative authority of a page in context. Similarly, if you use ontologies, you won’t have to focus so much on the meaning of individual names, and you can build relationships between related names to use different names in different contexts. The mappings between core data-driven words and corporate brand names are examples of ontologies.
Use objects, not rigid hierarchies. We have a natural tendency to want to build hierarchies of names: Products go up to brands, brands go up to categories, and so on. The problem is that a lot of names fit under multiple branches of a tree. Is cloud security under the cloud category or the security category? The correct answer is “yes.” But hierarchies force us to choose to put them in either one branch or another —or have multiple variations of the same value. Object-oriented ontologies solve this problem by focusing on relationships that are not hierarchical. These relationships become part of the intelligence that you build into each node of your taxonomy.
It’s not easy. But if you know what kind of tools you need and you hire really smart people to use them, you can have a taxonomy that works for your digital content efforts. When we say it works, we mean it is relatively easy for authors to use the appropriate tags for their content. When they do, it’s possible to build adaptive, personalized, iterative, reusable, shareable, and findable content. Without good tagging, it’s virtually impossible to do any of this.
Audience Feedback Systems
Systems that measure the performance of your content are perhaps the most mature of the three types of systems that you need to integrate. But that doesn’t mean it’s any easier to integrate them into your content ecosystem. Because feedback systems have been around so long, you are probably already drowning in relatively big data and can’t figure out what to focus on. Just when you think you get a handle on it, you get another deluge, as the data comes at you faster and faster, as the executives demand that you analyze it in real time. The biggest requirement for your audience feedback system is probably simplification.
Simplification really is possible. Most marketers do not need to understand the complexities of all the different measurement systems out there. It was once a simpler life for marketers, when all they needed to understand was their web analytics system. And as long as all of your content was on your website, that’s all marketers needed.
But with social media quickly moving from a technical curiosity to a marketing imperative, more analytics systems loom. Suddenly every social platform where you want to place and promote your content has its own lovely little analytics system. Twitter, Facebook, LinkedIn, YouTube—you name it—they all have separate analytics systems that track Likes and clicks and many other metrics.
Rather than pulling all these metrics in and trying to analyze them together, you need to understand your key performance indicators (KPIs) by which you will measure the results of all your efforts. KPIs are like filters for data that allow you to simplify the data that comes in and analyze it when it does. KPIs are up next.
Simplification really is possible. Most marketers do not need to understand the complexities of all the different measurement systems out there. It was once a simpler life for marketers, when all they needed to understand was their web analytics system. And as long as all of your content was on your website, that’s all marketers needed.
But with social media quickly moving from a technical curiosity to a marketing imperative, more analytics systems loom. Suddenly every social platform where you want to place and promote your content has its own lovely little analytics system. Twitter, Facebook, LinkedIn, YouTube—you name it—they all have separate analytics systems that track Likes and clicks and many other metrics.
Rather than pulling all these metrics in and trying to analyze them together, you need to understand your key performance indicators (KPIs) by which you will measure the results of all your efforts. KPIs are like filters for data that allow you to simplify the data that comes in and analyze it when it does. KPIs are up next.
Content marketing analytics KPI's
Content marketing analytics KPI's are the foundation of an analytics strategy. What you choose to measure will determine every other decision you make in designing a system that works. Here we want to give you a simplified model that you can give to any analyst to help them understand what you need to know:
Impressions
Your first question for any piece of marketing is “Did they see it?” For some kinds of content, such as a blog post or a web page, your web analytics system probably does a great job of giving you this number (which it calls page views). YouTube dutifully reports the number of views of your video. But for other kinds of content, such as tweets or Facebook shares, we know how many people could have seen it but not always how many did. For emails, you might need to settle for knowing how many people opened your email. Nevertheless, when you can determine how many people saw your content, it’s a number worth knowing.
Selections
Your next question for any piece of marketing is “Did they choose it?” Whether it was clicked with a mouse or stabbed with a finger, you want to know how often people engaged with your content by going deeper than just seeing it. Most kinds of content can be measured this way, and when it can, it tells you which content is resonating.
Conversions
Your next question is “Did it persuade them?” Exactly what your conversion is will vary by your business, but whether you want to see an online shopping cart checkout or white paper download, you want to know which content is being seen by the people who end up converting.
Revenue
Your next question is “How much did we sell?” For ecommerce, this is a simple calculation, but you might need more help to determine offline sales by tying your online conversions to a point-of-sale coupon or a customer relationship management (CRM) lead, for example. In any case, your analyst is the one who can help you connect the dots.
Return on Investment (ROI)
Your final question is “Was it worth it?” Most marketers know how much they spend, but they don’t know how much they make. If you can crack the revenue question, you can calculate your ROI to prove that what you are doing is effective.
what is content marketing and why are KPI's important?
Your KPIs are critical to running content marketing as a business, but it isn’t the only feedback that you need to collect. In addition to knowing how you are doing, you need other numbers to help you understand exactly which content asset got (or didn’t get) the results you were hoping for. Through KPI's performance managers can education top management on
why is content marketing important.
Attribution Modeling
Many buyer journeys require several pieces of content before the prospect is ready to talk to a sales rep or directly purchase a product. In those cases, each piece of content plays some role in the purchase and should get some credit for it. Attribution is about assessing the relative contribution to revenue for each individual piece of content.
Attribution is an old concept that goes back to offline direct marketing, when matchback systems tried to determine what percentage of credit for a sale different individual catalog mailings should receive. It made sense to them that if it takes multiple viewings of ads to persuade consumers to act, then being exposed to the same products month after month must have some cumulative effect also. Many studies were done to understand how to put together a matchback model with the best calculation to attribute credit. For digital marketing, we still use models to do this, but we call them attribution models.
Some attribution models give stronger weight to the early-stage content because it is what brings the customer into the digital front door in the first place. Some attribution models ascribe greater weight to the piece of content immediately prior to purchase because it is the “closer.” And there are models that give equal weight to each piece.
Whatever attribution model you choose, its accuracy depends on measurement systems that track users through the buyer journey. As we examined the plethora of analytics systems above, it’s critical that your marketing analyst know how to automatically pass the baton from system to system so that you know that the user who looked at your YouTube video is the same user who is now visiting your website.
Several methods can tie visitor data across systems, including cookies, tracking codes, and device identifiers. Which ones make sense will likely depend on the situation you find yourself in; you will unfortunately find some situations in which none will help. Impress upon your analyst the criticality of tying as many individual paths through social networks, search, and all the way through purchase even offline purchase. The fanciest attribution model on the planet is limited by the inability to chain visitor data across audience feedback systems.
As you become more sophisticated in your use of attribution modeling, you will find that these systems can do more than just give content credit for the part they play in purchase decisions; they help you understand your user paths so that you can optimize them. In this way, they are both reporting mechanisms and diagnostics that can help you present common next steps. Think of how Amazon presents products that are typically purchased together; much like that, if you know frequent paths through content, you can recommend common next steps to users to ease their paths through your content.
For more content content marketing tips, please explore more on our blog.
Attribution is an old concept that goes back to offline direct marketing, when matchback systems tried to determine what percentage of credit for a sale different individual catalog mailings should receive. It made sense to them that if it takes multiple viewings of ads to persuade consumers to act, then being exposed to the same products month after month must have some cumulative effect also. Many studies were done to understand how to put together a matchback model with the best calculation to attribute credit. For digital marketing, we still use models to do this, but we call them attribution models.
Some attribution models give stronger weight to the early-stage content because it is what brings the customer into the digital front door in the first place. Some attribution models ascribe greater weight to the piece of content immediately prior to purchase because it is the “closer.” And there are models that give equal weight to each piece.
Whatever attribution model you choose, its accuracy depends on measurement systems that track users through the buyer journey. As we examined the plethora of analytics systems above, it’s critical that your marketing analyst know how to automatically pass the baton from system to system so that you know that the user who looked at your YouTube video is the same user who is now visiting your website.
Several methods can tie visitor data across systems, including cookies, tracking codes, and device identifiers. Which ones make sense will likely depend on the situation you find yourself in; you will unfortunately find some situations in which none will help. Impress upon your analyst the criticality of tying as many individual paths through social networks, search, and all the way through purchase even offline purchase. The fanciest attribution model on the planet is limited by the inability to chain visitor data across audience feedback systems.
As you become more sophisticated in your use of attribution modeling, you will find that these systems can do more than just give content credit for the part they play in purchase decisions; they help you understand your user paths so that you can optimize them. In this way, they are both reporting mechanisms and diagnostics that can help you present common next steps. Think of how Amazon presents products that are typically purchased together; much like that, if you know frequent paths through content, you can recommend common next steps to users to ease their paths through your content.
For more content content marketing tips, please explore more on our blog.
COMMENTS