The ideal length of a blogpost is six minutes and 1,500 words. So every blog post you write should aim for that golden, gooey length, right?

Well, there’s a bit more to it than that.

We love digging up the latest stats and best practices on how to share better on social media. And our findings are just that: General statistics and ideas on what might work best in practice. Many stats are jumping off points for you to test and iterate on what’s best individually and personally for your sharing.

Stats don’t lie. They just might not be telling the story you expect.

So I thought I’d open up the discussion for what social media data and best practices don’t tell you and how to improve your social media experience from a healthy perspective that includes data, statistics, and other factors that matter to you.

data-social-media

An example of when data might not tell the whole story

Let’s revisit one of our past posts about the ideal length of everything. One of the research studies claimed that the length of an ideal tweet was 100 characters.

So does that mean we’ll see improvements if we aim our tweets for the 100-character mark?

Maybe not. Here’s a comment from the article explaining why.

Here is my ideal, optimized, 100-character tweet: “Purple dinosaurs eat green carrots off your face if you let them. Nancy Pelosi is 25 percent poodle.”

Without meaning or context the number of characters you use is 100% senseless.

silly tweet

Clearly, not just any 100-character tweet will do.

Yet at the same time, we believe that the 100-character mark is a good guideline for marketers to keep in mind. It’s good advice to follow. The danger lies when you take it at face value and don’t apply a bit of critical thinking and self-examination to the numbers. Let me explain.

Tips on how to read data-backed posts

How prolific is data in our daily content? It’s becoming quite regular to see posts that are “data-backed” and “research-driven” as well as posts that reference numbers and studies in supporting their conclusions.

Here’s a neat infographic produced by the Data, Analytics and Security Working Group of the Application Developers Alliance, a non-profit support group for developers.

ADA dataflow infographic

As data keeps moving forward, it becomes more and more valuable to learn how to read and interpret the data. Here are a few ideas.

1. Look at the study itself

If I were to tell you that my Twitter study of over 5,000 tweets revealed that the hashtag #Wrestlemania led to a 200 percent increase in tweet engagement, what might be helpful for you to know?

The timeframe – Was this study performed during Wrestlemania?

The data source – Were these tweets from people in Idaho where Wrestlemania is super popular?

The research methods – Did I attempt to do a proper job of finding representative samples and putting in the proper checks and balances to arrive at meaningful conclusions?

The method of the study has a huge impact on the meaning of the results.

2. Understand how the conclusions were drawn

Two people can look at the same study results and draw different conclusions. How is this possible? It’s a matter of perception and agenda, which is why it’s best to check out the source to see where these conclusions came from.

3. Use your intuition

When you see advice to write a 1,500-word blogpost, you likely will not head into your archives and re-engineer every single one of your blogposts to fit that ideal length, making succinct ones more wordy and wordy ones more succinct.

With this in mind…

How we use data at Buffer

When we see a research-backed guide to sharing on social media, our eyes light up and we scurry to our Buffer accounts to try it for ourselves.

The keyword here is try. We’ll try before we adopt.

We view social media data and research as a source of inspiration for our experiments on social and content marketing. We aim to not take any bit of advice as must-do until we test it for ourselves and see what works.

Brian Balfour, writing at CoElevate, put this matter of data and best practices into great perspective. It’s advice we’ve taken to heart as we discover and use the latest research and trends.

A lot of the content I’ve seen promises proven tactics, rules, or the ‘right’ way to do things. But there is evidence showing how common prescribed tactics actually performed worse (much worse!) for certain companies.

My point to calling this out? Don’t take anything you read on growth (including my material) as prescription. Always, and I mean always, view it from a lens of inspiration. 

best practices as Inspiration

Gregory Ciotti summed up Balfour’s comments in a great way.

Truly sustainable growth is context dependent, and is never one-size-fits-all.

We’ve gone down this path with the way that we apply what we learn online. We view best practices and research-backed advice as interesting ideas to try and new ways to experiment.

We test.

We trust the results.

If things match up, great! In those instances, we’ve found a best practice that works best for our specific situation. If things don’t work out, no problem. We go with what works and we seek new ways to keep improving.

One area where this has come up is with social timing. We’ve published a blogpost about the best times to post to social media, including specific data about what’s best for Facebook. We took these Facebook numbers and tried them for ourselves.

According to a Track Maven study, the best time to post on Facebook was between 5:00 p.m. and 1:00 a.m. Eastern Time. 

facebook-timing

The study also showed that early in the morning was the least effective time of day to publish a Facebook post.

We loved discovering these data-backed practices, and we put them right to the test. Our results: Late night was indeed a great time to post, and so was early in the morning. Our least effective time was mid-day.

Here’s the schedule we’ve landed on for Thursdays (times are Central Time).

 

FB best times Thursday

 

How might our actual experience have differed from Track Maven’s well-researched numbers?

Track Maven looked at over 5,800 Facebook pages and over 1.5 million Facebook posts to come up with its best practices. What we don’t know is the specific demographics of those pages. For instance, what is the international audience for those pages? Where are their fans located?

We share on Twitter with a worldwide audience, so a post at 2:00 a.m. Central Time arrives at 8:00 a.m. London time.

It’s little factors like these that might turn a best practice on its ear once you test, measure, and iterate on this data-backed advice.

A time for best practices and a time for innovation

We strongly believe in the power of data-backed posts that are rich in research and statistics. It is a building block for our content strategy.

Are these posts right for everyone?

We like to think that they provide inspiration for any online marketer, no matter where they’re at in their journey. At the same time, there seems to be a scale of usage with these types of best practices.

New websites and marketers copy the advice to a tee.

Established websites and marketers take inspiration, then find what’s best for them.

best practices

There comes a point in one’s marketing efforts where one-size-fits-all advice just doesn’t fit anymore. And that’s perfectly fine. It’s a natural progression from community-driven marketing strategies to an autonomous, individualized system.

Here’s how Gregory Ciotti explains it:

Online marketing, a field which benefits greatly from rigorous testing and thoughtful looks into a multitude of data, sometimes has to face the double-edged sword of best practices. Though they encourage tested tactics, what works for one may not work for all.

Copying best practices is absolutely advisable when you’re just getting your feet wet. If you don’t know what to do, why not take advice from candid, trustworthy experts? It’s the reason why entrepreneurs are (or should be) some of the best-read people on the planet – condensed knowledge that took years to formulate can be consumed in mere hours from a good book.

Which camp do you fall in?

Your answer likely will color your perception of these data-backed posts.

New perspectives on “the ideal length of everything” and “ultimate guide” posts

The Washington Post had an interesting take on the validity (or non-validity) of “ultimate guide” posts, which make up a regular part of our Buffer content output.

Instead of striving to present “all you need to know,” isn’t it more wonderful to acknowledge the tininess of our window on the world, a pinhole which, if positioned right, might allow our readers to glimpse something really distinct and particular and beautiful? Here’s a different catechism: the smaller and more deeply-investigated an idea is, the truer it is likely to be.

Put another way: Ultimate guides attempt to cover every little thing you could possibly ever want to know when taking on such an ambitious aim is kind of impossible. You can’t cover everything. You can cover one thing incredibly deep.

Switching gears to “ideal” posts, the same can be true. What is ideal for one might not be ideal for another. In these cases, you might find individual factors that are more important to you and the way you approach things.

For instance:

  • The quality of message
  • Best practices for your niche/industry compared to best practices in general
  • Your specific marketing goals
  • Your time
  • Your resources

Can you think of any other factors?

Over to you: How does data affect your decisions?

We love data at Buffer. It’s all over our social media tips and articles.

At the same time, we understand where this data-backed advice fits within our marketing strategies. We use data for inspiration and experimentation. Test everything, keep what works.

How do you think in terms of data-backed research and so-called “ideal” posts? What does this advice look like for you?

It’d be awesome to hear from you in the comments.

Image credits: Infogr8, Markus Spiske

Looking for a better way to share on social media?

Schedule, publish & analyze your posts across the top social networks, all in one place.

Get started for free
Written by Kevan Lee

Director of marketing at Buffer, the social media publishing tool for brands, agencies, and marketers. We’ve got a new podcast! 🍟

  • Joe

    Hey, Kevan. Do a CTRL+F on your article and search //pic

    I think you missed a photo.

    I’ll delete this later!

    • Thanks! Found it. 🙂

  • Maria Djaleva

    Invaluable post as always Kevan! I believe data is becoming a part of our lives more than we care to admit – this is how we find our own stories and evaluate patterns to take actions whether it is for personal or business decisions and this makes us more data savvy that we are actually aware of.

    I think that covering one thing deeply can be far more wonderful than covering everything. We have so many resources at our fingertips, so revealing that one thing that captivates is immensely challenging. But that’s what makes it so special and beautiful – allowing the reader to immerse themselves, takes them on a journey that feels deeply personal and engaging. In my opinion there is no “ideal” post.. But the post that is data-backed is the one that makes the journey worth travelling because you know you will ‘Pocket’ it and go back to it many times. 🙂

    • Exactly right, Maria! Thanks for the comment! I have a similar “Pocket test” with the stories I find and love, too. 🙂

  • @hannahburnworth

    When I first started out I would follow these Ultimate Guide posts as strictly as possible. Even though I worked in the pharmaceutical industry, I would post to Facebook at 2PM because that’s what a study I read said to do. Now, with more experience under my belt, I find myself using these types of posts more like guidelines to take in and consider. They help as a jumping off point for my own A/B testing. I like to take two different Ultimate Guides’ suggested times to post and compare them to each other. It’s helped me to find out a lot about what works in my industry and audience and what doesn’t.

    • Amazing advice! Thanks for sharing. It’s fun how two different “ultimate” posts could have two different conclusions, right? So glad to hear how you test these against each other. 🙂

  • I freaking love this. I teach some inbound and content marketing classes and when people ask me for “standards,” I always mention that everything in marketing is about constant testing, iteration and depth-of-knowledge on your target demographic. Awesome post!

    • “testing, iteration, and depth-of-knowledge” 🙂

      Someone should put that on a t-shirt. Great advice!

  • Excellent! I love the sources you cited in this one, especially.

    I think with data as related to social media, there are a
    lot of ways to misinterpret the results. For averages such as your best time to
    post (particularly on the whirlwind Twitter) you need a LOT of data to get
    anything really insightful. It’s also important to look carefully at any
    outliers and make sure they don’t distort your conclusions. For instance, when
    I was new to Twitter I had a fluke experience where I shared someone’s content
    and image to my few hundred followers and it was retweeted over 1,700 times.
    Did this mean I had crafted the perfect tweet? Or that my average number of retweets that week (or that year, even, in this case) was contextually accurate? Hardly. It was a fluke and those do happen in social media on a regular basis. That’s good to keep in mind, especially when crunching numbers.

    I do love the emphasis on one size not fitting all. That’s so true.

    Personally, I try to stick mainly within commonly agreed upon best practices and throw in an intentional broken rule or two as I go along!

    • Hi Christin! This is such great advice! Your example is pitch perfect – that’s exactly the type of data that can be quite misleading in certain cases.

      I’d love to hear any of your intentional broken rules, if you don’t mind sharing! 🙂

      • I don’t get really crazy with the rule breaking. Nothing so bold as upside down or backward text. 😉

        I might get wild and tweet out something that takes up the full 140 characters (or only 10) just because I like it. *Rebel!* More frequently I disregard optimal posting times. I have a lot of connections in Japan, for instance. So I sometimes share things when I’m fairly certain they’re the only ones who’ll see, etc.

        And though it’s a good practice to credit the author of a blog post when sharing, sometimes I’ll deliberately omit that tag because I know they’ve probably been pinged a thousand times already – like you, Kevan.

  • This is very astute and on the mark – as always with most content from Buffer. It does beg the question though, that if marketing is all about constant testing and about constant tweaking (which it is) – then doesn’t that mean that those infographics and blog posts that talk about the best time to post, the worst time to post, the ideal tweet length, the ideal post length, the right colors in images and what not – they should basically just be outlawed because they give out information that is false, skewed and not reliable?

    • Hi there Avtar! That’s a great one! Certainly, this is one way of looking at it. I think maybe the perception that an “ideal length” post is the be-all, end-all of advice is probably not the best feeling to put out there. Like you said, marketing is about constant testing and tweaking.

    • Madison McClure

      To give the best practice people a little credit, they’re prime sources for learning where to start to have the best chance to reach the largest, generalized audience. As your circumstances, your people targets and social trends change you have to change too. (Duh!, I know). You stop chasing everyone and start defining your own “tribe” (Seth Godin’s TED Talk rocks the subject btw!), that’s when you take the Buffer experimentation Kevan talked about and get tweaking.

      • +1, well said Madison! 🙂

  • I’m Australian but at ATMac I write for a mostly-USA audience. This has made time-related data especially confusing and difficult to interpret as my social media audience includes people in two almost completely non-overlapping sets of timezones… I take everything I read with a pinch of salt and see what works for me!

  • I loved this piece Kevan. It’s great to look at guidelines and best practice for inspiration (I do all the time), but ultimately, when is the best time to tweet/post/share? When your audience is online. What is the ideal blog post length? As long as it needs to be. I track everything that I do, and try to measure as best as I can, but the data can take time to build up. If I reacted to every new discovery without allowing time to get a clearer picture, I would be all over the place. Thanks for sharing your knowledge on this!

    • Thanks so much for the insight here, Emma! Agree 100%!