For many people, “content quality” is defined the same way the Supreme Court defined obscenity:
“I know it when I see it.”
The problem with content quality being merely a matter of taste and subjectivity is that it tends to slow down the system, limit creativity, and promote endless debates on what is and isn’t quality.
It would benefit our community of word nerds to get just a little bit tighter on this definition.
Here’s my take:
a) Quality is in the eye of the consumer
At CXL, I wrote 5,000-word research and experience-backed guides about semi-technical topics like A/B testing statistics.
HubSpot also wrote guides on A/B testing, though toned down about 5 grade levels.
Was one higher quality than the other? My personal opinion is: yes, CXL was higher quality.
But guess what? My personal opinion barely matters.
HubSpot’s audience isn’t primarily a bunch of pedantic data nerds wading in the esoteric territory of parametric vs nonparametric statistics tests. They were largely small business owners who wanted to grasp marketing 101 fundamentals.
Quality is not for you and me. It’s for the reader.
How do you figure out what the reader wants? If you’re starting from scratch, basic customer research and hypotheses will do.
My favorite methods are to scour forums or communities where the target audience hangs out, do audience research using traditional SEO tools as well as tools like Sparktoro, and do 1:1 customer interviews.
Ultimately, you’re trying to find an answer to the question, “what do[es the reader] want?”
If you’ve been pumping out content, it’s best to iterate based on performance data as well as continuous qualitative feedback.
Your audience will tell you, explicitly and implicitly, if your stuff is resonating. I like to put surveys in drip email sequences, add on-site HotJar polls to content pages, and of course, look at quantitative analytics to see which page types resonate and convert highest.
b) Quality is an input, not an outcome
It may seem paradoxical, but a good decision can result in a bad outcome and vice versa.
If you measure quality only on outputs, you risk “resulting,” a logical fallacy “where we evaluate the quality of our decisions based on the outcome they achieve.”
In the game of poker, and in the game of life, you’re dealing both with skill and luck or chance. To win over the long run, poker players create rules and systems by which they play to avoid confusing luck for skill and to avoid making decisions based on illusory patterns.
When you look at the level of a single blog post, a great one can fail to rank and a seemingly poor one can rank. But program-wide, great blog posts will, on average, perform better.
Therefore, quality standards can predict performance, but quality is an input and performance is the output.
c) Quality is predictive of a positive output
As an input, quality standards should predict a positive output or outcome.
Content marketing/SEO/all of the stuff we’re doing fits into a business context, where we have objectives that map to enterprise value. We’re not artists, even though what we do is often creative. It’s not creative in a vacuum. It’s creative in pursuit of a goal.
The quality standards you set for your brand should lead to a higher likelihood (on average) that you reach your business objectives, and they should map to the specific strategy, channels, and audience that you’ve decided on.
For example, there’s a massive amount of causal and correlational data on what leads to success in SEO. There are also explicit Google guidelines as to how to rank. If your content is designed to rank in search, these quality guidelines should be followed with each piece of content.
And because search is merely the channel by which you reach your target audience, the guidelines should also be set up so that they resonate with this audience once they click through to your page.
If you can’t map back your quality standards or editorial dimensions to a higher likelihood of success, in the context of your strategy, consider the thought experiment, “What if we got rid of this standard? What would change about our results?”
If your outputs are unlikely to change as a result, then it’s likely that the quality standard is superfluous and increases cognitive load for writers and creatives.
d) Quality standards should be objective
Other than your north star vision (which I’ll explain below), your quality standards shouldn’t be subjective or vague.
“We’re formal yet friendly.” What does that mean? 20 different people could have 20 different definitions of that.
“We cite every claim we make with research or experience.” This is good. It’s something an editor can tangibly use to address (and communicate) feedback on a draft.
It’s hard to make quality standards objective, but it’s the only way they’ll be useful. If they’re subjective, it opens up endless back-and-forth with nitpickers with benign intentions: they just have a different definition of what quality means than you.
When quality standards are subjective, that’s when quality gets in the way of quantity and they’re truly dichotomous. They shouldn’t be. Quality standards should enable production scale, not get in the way of it.
e) Quality standards should reduce cognitive load and debate
In addition to being objective, quality standards should be clear and easy to understand.
A lot of what we did at CXL was simply ask, “So what?” and say “Prove it,” in our content.
Here are two dimensions in our style guide that spoke to those editorial heuristics:
Now when you’re writing a draft and you have a sentence like, “Email marketing is the most impactful thing a marketer can do,” you have to answer, “What’s the proof?” and “Okay, so what? What should the reader do about that?”
f) Quality standards are operational components
A style guide should be an artifact that is used to increase operational efficiency no matter the shape of your organization, the tools you use to get the job done, or the strategy you’re working towards.
Whether you’re using AI tools to produce a portion of the content, you’re writing thought leadership to be distributed by your employees via LinkedIn, or you’re writing gated whitepapers to distribute via paid ads on syndicated networks, a style guide is a core component that guides the production of these assets.
It should make the boat go faster. It should make content planning easier, content writing faster, and content editing smoother.
It’s a component of the process, but importantly, it should be audited to ensure it isn’t weighing down your program and limiting your outputs, but rather expanding them and making things feel easier.
3 Components of Content Quality: My Framework
A complete quality guide needs three things:
1. North star vision
Ignore what I wrote above for just 30 seconds. Your content strategy should have a north star vision that is completely subjective and unattainable.
This is a subjective “principle,” and not an objective criteria in your style guide.
At CXL, the core principle and north star vision was that “every single blog post…aims to be the best in the world content on whatever topic it is.”
When I wrote or edited a piece, I always had this question in the back of my mind. I wanted it to be the last piece someone ever needed to read on that topic. Research ends at this blog post.
Was it feasibly attainable? Probably not. Was it a good anchor and inspirational vision for what our content could be? 100%. Does your vision need to be the exact same? Definitely not.
2. Table stakes guardrails
Next, you need checklist items that editors can quickly clean up. Most of this is via negativa, i.e. things not to do.
Think of these as the lower bound on quality. If a piece doesn’t reach these, it’s a binary pass/fail score, and the piece is fixed or thrown out.
At CXL, we listed this as a TLDR on the style guide. It included word count (controversial, sure, but for our strategy and audience, you couldn’t usually produce best-in-the-world content in under 2k words).
It also included readability standards like breaking up copy into readable chunks (we specified how many lines a paragraph should be) and adding photos. Of course, it had to pass an originality checker and we added items about claims backed up by research.
3. Features that differentiate
Finally, time for your personal flair or things that make your brand unique.
The best way to figure out these features is to take your strategy, figure out what works best within those parameters, and then look at your competitors and figure out what they’re relatively worse at or not doing, and then index on the things that can differentiate you.
For example, you could require an executive summary at the beginning of every long-form post. Or require the inclusion of at least 2 original subject matter expert quotes. Or custom images and infographics or GIFs.
The thing with this category is, I can’t tell you what to add here. It’s unique to you, and that’s the point.
Want more insights like this? Connect with me on LinkedIn.
- “High Quality” Content Isn’t Real – great breakdown from Margaret Kelsey and Devin Bramhall. I swiped a lot of points from their chat for my essay above (but hey, here’s the credit).
- Defining Objective Content Quality Standards & Why Content Marketing is Cheesy with Margaret Jones (Airtable) – my initial “aha moment” that quality standards should be objective came from this convo with Margaret from Airtable. Useful and entertaining discussion all around.
- Brand POV – a previous newsletter devoted to your Brand Point of View.I personally believe this is far more impactful to define than your “brand voice,” which is barely noticeable in most cases to your target audience. Unless you have a truly unique style, the things you believe and the stances you take will travel much farther than choosing a “confident” vs “helpful” tone.