Creating high-quality marketing content is difficult – and not just because the act of creation itself is hard. The real reason: every marketer has different definitions of ‘quality’.
In this article, I’m going to introduce a framework that helps clarify exactly what high-quality content is (the kernel–chaff rubric). I’m then going to break down how you can create quality content, explore how Google and LinkedIn define quality, and explain how to deal with any low-quality content that you encounter.
- The Problem: No Clear Definition of ‘Quality’
- The Solution: The Kernel–Chaff Rubric
- Why Does the Kernel–Chaff Rubric Work?
- How to Create High-Quality Content
- How Do Search Engines Determine Quality Content?
- How Does LinkedIn Approach Content Quality?
- How Do I Deal With Low-Quality Content?
The Problem: No Clear Definition of ‘Quality’
As a marketing leader, you know high-quality content when you see it. It’s the precisely worded article that effortlessly builds a convincing argument. The snappily edited Instagram reel that elicits instant FOMO. The high-energy podcast episode stacked with aha moments.
But being able to recognise high-quality content and knowing how to consistently produce it aren’t the same things. One reason it’s so difficult: no-one can agree on exactly what ‘high-quality’ means. And, without a shared objective, getting your team to pull in the same direction isn’t easy.
‘Quality’ is a subjective term, and ‘high-quality’ and ‘low-quality’ are assumed to be different ends of a spectrum. That creates problems. Where do we draw the line between the two? What are the determinants of ‘quality’ anyway? How do we objectively measure a subjective concept?
The Solution: The Kernel–Chaff Rubric
The solution to the confusion around ‘high-quality content’ is actually quite simple: abandon the term ‘high-quality’ altogether. If you want to give your team consistent goalposts, you need a quantifiable objective, a concept that everyone has a shared understanding of.
That concept is the kernel–chaff rubric. It’s a framework under which all content is classed as ‘kernel’ or ‘chaff’ – an objective dichotomy instead of a subjective spectrum. Kernel content is the good stuff (which we’ll talk more about in a minute). Chaff content, on the other hand, is content that’s not fit for purpose – assets that need to be discarded, recycled, or significantly reworked.
Kernel content has four traits: it’s technically competent, it’s unique, it adds value for consumers, and it supports marketing outcomes. Here’s what each of those traits means.
Technically Competent Content
Content that is technically competent is exactly that: it reaches a basic level of technical proficiency, the kind of standard that makes you nod and think, “Oh, that looks like it’s been done professionally.”
Technically competent textual content:
- is free from grammar errors;
- is stylistically consistent;
- uses conventions appropriate for the medium in question; and
- is easily comprehendible at both a sentence and a structural level.
Technically competent acoustic content:
- is free from audio artifacts;
- has crisp, evenly balanced sound;
- does not have overlapping dialogue or noises that make comprehension difficult; and
- is pleasant to listen to.
Technically competent visual content:
- is visually consistent;
- has a clear visual hierarchy;
- uses conventions appropriate for the medium in question;
- is accessible for all users; and
- is optimised in terms of file size and image quality.
Exactly what technical standards you decide to apply to your content may vary. Ultimately, though, they should be objective – every stakeholder knows, for example, that bad grammar is a sign of poor quality (even if they might quibble over certain grammatical rules).
No content asset will be wholly unique. That’s simply not possible, especially in fast-paced marketing environments.
Instead, the ‘unique’ trait relates to whether a content asset has some original element not found elsewhere on the distribution channel in question. That could include a different perspective, new information, or a fresh style/medium/format.
For example, TikTok might be riddled with videos of a specific dance sequence – but there might be almost no coverage of that dance on Google, so creating a ‘how to’ article on your website with step-by-step instructions and demo videos would tick the ‘unique’ box.
Another example could be an article about how to become a better copywriter. While that particular topic might be done to death in the SERPs, writing it with the addition of your personal perspective (“Here’s how I became a better copywriter”) would add a unique aspect.
Note: ‘Uniqueness’ is very similar to information gain. The main difference is that ‘uniqueness’ looks at more than just the raw information – it focuses more broadly on bringing something new to the table, even if that’s the same information presented in a fresh way.
Content is, by definition, any communication with the primary purpose of adding value for the consumer. (I say ‘consumer’ because content isn’t just textual – it can be acoustic, visual, tactile, olfactory, gustatory, or any combination of the above. Marketers tend to focus on digital content to the exclusion of other, potentially more impactful mediums.)
That value addition typically falls into three categories: education, entertainment, and inspiration. To be classed as adding value under the kernel–chaff rubric, content must fulfil at least one of those categories.
Educational content provides information that consumers can use for a purpose. Often, that purpose relates to solving or preventing a problem.
Examples of educational content include a video explaining how to change your car’s oil, a news article warning about rising crime in your local area, and a dictionary.
Entertaining content evokes emotions to hold consumers’ attention. Those emotions can be positive, such as humour, awe and excitement, or negative, such as anxiety, anger, or morbid curiosity.
Examples of entertaining content include fictional TV shows such as Billions, performances such as Les Misérables, and artwork like ‘Laocoön’.
Inspiring content incentivises consumers to take action by fuelling desire for a particular outcome.
Examples of inspiring content include travel TV shows, Instagram Reels showcasing idealised body types, and motivational entrepreneurship videos on YouTube.
The reality is that many content assets are not mono-functional – in fact, the most effective ones rarely are.
Think about outrage news programs. They blend a smattering of facts about hot-button issues with emotional manipulation and superficial ‘solutions’, creating a vicious cycle of consumption that keeps viewers coming back for more. Despite being mentally toxic, patently pointless, and damaging to public discourse, outrage news now dominates journalism – proof of its effectiveness as a content style.
A more positive example is nature programs. They combine education with awe-inspiring visuals and mini-narratives that build empathy for their subject creatures. When we watch, we learn – but we’re also emotionally engaged. That’s much more compelling than the same information presented in a Wikipedia-style article.
The final metric in the kernel–chaff rubric is that the content asset in question supports marketing outcomes. That can occur through two mechanisms: moving buyers further along in their buying journeys, and/or increasing brand awareness.
To make sure we’re on the same page about what ‘buying journey’ means, here’s a visualisation of our demand farming framework.
The best content assets pull both levers (buying journey and brand awareness), but you can have kernel content that acts on just one.
An example is the Bored YouTube series by New Zealand comedy group VLDL, which is centred around employees working in a real computer hardware store called Playtech. While most of the episodes don’t bring buyers any closer to purchasing hardware from Playtech, almost all of them contribute strongly to brand awareness.
Why Does the Kernel–Chaff Rubric Work?
The kernel–chaff rubric helps teams deliver high-quality content for three reasons.
- It’s applicable to all content types, not just written content.
- It’s relatively objective. Each of the four metrics is quantifiable (to a point).
- It’s achievable. It isn’t a rubric for ‘highest-quality content’, but the baseline that content needs to reach to be effective (that is, kernel content).
While there are other frameworks for high-quality content, too many of them are mapped to a specific medium or channel. Here’s an example from Entrepreneur.
Originality and relevance might be channel-agnostic, but readability, voice and grammar are all text-centric. Timelessness is also a very SEO-focused metric – why does content have to be evergreen to be high-quality? What about social posts? Podcast episodes? News articles? The kernel–chaff rubric, by contrast, can be applied to everything from e-books to live performances.
Too many frameworks are also incredibly vague or subjective. Everyone wants their content to be ‘engaging’, but what does ‘engaging’ actually mean? The kernel–chaff rubric has clear definitions for each metric, making it easy for any stakeholder to objectively assess content quality.
How to Create High-Quality Content
There are almost infinite ways to create high-quality content. Rather than trying to provide a definitive guide about what to do and what to avoid, I’ll list some basic principles for creating kernel content.
- Use the kernel–chaff framework.
- Hire technicians who know what they’re doing.
- Hire editors and strategists who know what they’re doing.
- Give your team the budget and time that they need to do good work.
- Give them access to subject matter experts, high-quality customer research, and well-developed brand infrastructure.
- Start by thinking about how you can add value for your ICP in a way that supports marketing outcomes. If you’re starting with tactics or channels, you’re doing it wrong.
- Don’t have a committee of stakeholders weighing in on content creation. Consensus creates mediocrity. Use a decision-making framework like DACI or DARE.
- Have a content strategy. If you’re committing random acts of content, expect positive outcomes to be equally random.
- Don’t be timid. If you’re not trying new things with your content, it’s going to be harder to stand out.
- Have a decent measurement system in place that tracks both leading and lagging indicators for channels. If you don’t measure performance, you can’t improve.
- Create processes for content production. The more mature your systems are, the more easily you can optimise them for efficiency.
- Give your content time to work. Most distribution channels don’t deliver instant results – they require an accretion of kernel content to make a tangible impact.
How Do Search Engines Determine Quality Content?
Because search engines play such a major role in content distribution, it’s important to understand how they evaluate content quality. Of course, different search engines have different criteria. For the purposes of this article, I’m going to focus on Google’s determination of content quality.
Luckily, there’s a 176-page document that explains exactly how Google does approach content: the Search Quality Evaluator Guidelines. Written for human quality raters (people who manually complete rating tasks to help measure algorithmic effectiveness), the guidelines provide a clear window into what the search giant considers ‘high-quality’.
From Google’s perspective, high-quality content has three main traits:
- It has been created with a high level of effort, originality, talent, or skill such that it achieves its purpose well.
- The website and content creator both have positive reputations in relation to the topic.
- It has a high level of EEAT.
Let’s break down what each of those traits means.
High Level of Effort, Originality, Talent or Skill
‘Effort’ relates to the thoughtfulness you’ve put into crafting your content. Is it well-formatted and free from distractions? Is it mostly free from grammar errors? Is information delivered in a cogent, logical way?
‘Originality’ refers to whether the information and visuals you provide are unique to your website or scraped/compiled from other sources.
‘Talent or skill’ relates to whether you, the author, have the talent/skill to create content that fulfils the page’s purpose. For example, a skilled chef would be able to create an effective, easy-to-understand recipe – something that a layman might not be able to do.
While the absence of reputational information won’t necessarily affect your chances of being rated as ‘high-quality’, Google does advise its raters to assess the online reputation of either the website or the content creator (in cases where the brand behind the website didn’t create the content).
According to Google, “a website’s reputation is based on the experience of real users and the opinions of people who are experts.” Importantly, that reputation must relate to the content topic – having a positive reputation in relation to marketing content wouldn’t translate into a positive reputation for nutritional content.
Website reputations can be assessed based on:
- independent reviews;
- recommendations from subject matter experts;
- news articles;
- Wikipedia articles;
- magazine articles;
- forum discussions;
- ratings from independent organisations;
- industry awards;
- a history of high-quality content; and
- customer reviews.
Creator reputations can be assessed based on:
- educational degrees;
- peer validation;
- expert co-authors;
- academic citations;
- relevant employment history;
- training and credentials;
- news articles;
- commentary from influencers; and
- comments from platform users.
Reputational information is most important for YMYL content.
While we don’t necessarily know how much (if any) influence third-party references have on real-world search rankings, there are two ‘reputational’ factors that play a big role in SEO: GMB reviews and, of course, backlinks. As per the guidelines, sentiment and keyword presence matter in reviews, and topical context is critical for link impact.
Unlinked brand mentions are a third reputational factor that may also affect rankings. While Google has historically connected mentions and topical entities, there’s little data to indicate what sort of impact they have.
Experience, expertise, authority and trust (EEAT) is Google’s updated metric to evaluate overall credibility. In the figure below, you can see that trust sits at the intersection of experience, expertise and authority – all three are essentially building blocks that underpin trust.
Here’s a definition of all four terms:
- Experience: How much first-hand experience does the creator have in relation to the subject matter?
- Expertise: How much knowledge or skill does the creator have in relation to the subject matter?
- Authoritativeness: To what extent is the creator known as a go-to source on the subject matter?
- Trust: Overall, how trustworthy is the creator in relation to the page’s purpose?
It’s worth noting that EEAT, positive reputation, and ‘high level of effort, originality, talent or skill’ all cover similar ground and have similar requirements, which makes complying with them easier.
The big takeaway from EEAT is that you don’t necessarily need to be an expert; first-hand experience – as a product user, for example, or as a hobbyist – counts towards quality just as much as knowledge-based expertise does.
How Do the Guidelines Align With the Kernel–Chaff Rubric?
There’s significant overlap between the Search Quality Evaluator Guidelines and the kernel–chaff rubric.
‘Effort’ correlates closely with ‘technical competence’. ‘Originality’ and ‘uniqueness’ are synonyms. ‘Talent or skill’ is very close to ‘value addition’, at least in terms of their respective meanings – both relate to fulfilling a specific purpose, and, in doing so, adding value for the reader/searcher.
Positive reputation isn’t included in the kernel–chaff rubric, mostly because, from a non-search standpoint, it isn’t relevant to a content asset’s quality. Someone with no reputation or a poor reputation can create content that is just as high-quality as a subject matter expert (even if consumers might regard the former with more suspicion). Using reputation as a quality metric is an argument from authority – not exactly a sound inclusion for any reliable framework.
EEAT is omitted for similar reasons. Are experience, expertise, subject authority, or trust actually necessary for high-quality content? Of course, it would be difficult to create kernel content without either experience or expertise (whether that’s first-hand or derived from research), but neither is essential in and of itself.
As such, using the kernel–chaff rubric is a good way to align your content with Google’s own quality guidelines.
Do the Guidelines Matter for SEO?
It’s worth noting that the Search Quality Evaluator Guidelines aren’t an algorithm or ranking factor. While there’s almost certainly overlap with signals like the helpful content system, the guidelines themselves are purely to help human evaluators classify content. (These classifications don’t affect the ranking of individual content pieces, either.)
With that said, the guidelines do represent an aspirational standard for Google. Ideally, content would appear to searchers based on query relevance and the quality classifications described in the guidelines – that it doesn’t consistently do so yet is purely because Google’s ranking algorithms haven’t caught up. If you create content in line with the guidelines, you’re essentially proofing it against future algorithm updates.
How Does LinkedIn Approach Content Quality?
As I was writing this article, Entrepreneur’s editor-in-chief Jason Feifer conducted an interview with Dan Roth, LinkedIn’s editor-in-chief, and Alice Xiong, one of the platform’s directors of product management. The result was an excellent article about LinkedIn’s algorithm that you can read here.
Importantly, Jason shed light on some key quality metrics that brands operating on LinkedIn need to be aware of. Here’s a summary of the relevant concepts and how they relate to the kernel–chaff rubric.
According to Jason, “LinkedIn’s system is now evaluating whether a post contains knowledge and advice, and then showing it to other users who are likely to find the information relevant and useful” (emphasis is mine).
The LinkedIn algorithm uses multiple different metrics to assess ‘knowledge and advice’. The three that Dan and Alice shared are:
- The post speaks to a distinct audience. Does it have a clear class of people who will find it relevant?
- The author is writing in their core subject area. How much of an overlap is there between your skills and professional background and the content you produce?
- The post has a perspective. Does the post just offer generic information, or does it include opinions and/or advice?
Dan and Alice also cite having ‘meaningful comments’ as a quality metric. This serves a threefold purpose: to defuse so-called ‘engagement pods’ (LinkedIn users banding together to engage with each other’s content and hack the algorithm), to combat influencers who post viral chaff content, and to gauge how other users are receiving a particular post.
How Do I Deal With Low-Quality Content?
Preventing Low-Quality Content
The idiom “prevention is better than cure” is as true for content quality as it is for anything else. The best way to deal with chaff content is to stop it from being created in the first place.
The easiest way to do that? Follow the checklist under ‘How to Create High-Quality Content’, and make sure your team has internalised the kernel–chaff rubric. Your editors are especially important. They need to be objective enough to spot and call out chaff content, but you also don’t want pedants – people who get so hung up on small details that they demoralise creators and slow down production.
Fixing Low-Quality Content
If your content is posted on chronological channels (such as Facebook, Instagram, or LinkedIn), you don’t need to worry too much about fixing chaff content. After all, the likelihood of average users going back to view posts from two or three years ago is incredibly low. If you do come across anything that could damage your brand, such as inaccurate claims or outdated viewpoints, delete the posts in question.
If you’re dealing with content retrievable via search engines (such as on your website or YouTube), you’ll probably need to conduct a content audit, which should be followed by an action plan. Allocating budget for a full-scale content overhaul may not be your leadership’s priority, so use a prioritisation framework to make changes over a staggered timeline (for example, four articles fixed per week).
We’ve covered a lot of ground in this article – thanks for staying until the end.
By now, you should have a clear understanding of the factors that Google and LinkedIn use to determine content quality. We’ve also explored how to create kernel content and how to prevent and fix chaff content.
Most importantly, you now have a framework to assess content quality, regardless of the medium.
The kernel–chaff rubric’s four metrics – technical competence, uniqueness, value addition, and contribution to marketing outcomes – are all ideal for creating kernel content across a range of different platforms. They’re not tied to any particular marketing ideology or process, and they’re more objective than many existing frameworks.
You can also apply the rubric to any organisation’s marketing, and it’s simple enough that your editors/reviewers shouldn’t have trouble implementing it. Once you can align your team (and leadership) around a set of shared goalposts, creating consistently good content will be much simpler.
There are plenty of excuses for failing to create high-quality content. Lack of resources. Lack of buy-in. Ineffective creators and poor strategy. But not knowing what high-quality content is shouldn’t be one of them.
Make kernel content your goal. Learn how to curb the creation of chaff content. Better, more effective marketing will follow.