By Jonathan Crossfield published March 13, 2016

Is Your Social Media Content as Popular as You Think?


Marketers use a variety of metrics to collate, crunch, and calculate how their content performs in social media. Metrics such as shares, retweets, and views are often the easiest and most obvious to gather, but they may be the most deceptive and unreliable when evaluating whether your content is genuinely making a difference.

The marketing industry has a terrible habit of devaluing perfectly serviceable words and phrases by turning them into vaguely defined whiffle-dust. These buzzwords might add impressive-sounding sparkle to a strategy, but they usually conceal the lack of any genuine substance capable of driving measurable business outcomes.

“Engagement” particularly irks me. Engagement used to mean something. It meant capturing and holding attention. It meant interaction. It meant getting the message across. All of these are important steps on the way to a conversion or bottom-line business goal.

But somewhere along the way, engagement became the goal instead of the journey. Along with the equally nebulous “awareness,” engagement often ranks high among the stated content marketing goals in each Content Marketing Institute research report. And too many marketers attempt to measure these goals with metrics that don’t necessarily indicate anything of the sort: Web traffic, clicks, “likes,” tweets, and opens.

According to The Fournaise Marketing Group, 76% of marketers use the wrong KPIs and metrics to assess the effectiveness of their strategies. The same research revealed that most marketers still consider marketing effectiveness to be about awareness (74%) and/or engagement (71%). Of those, 86% believed engagement was a form of conversion.

… these marketers believe that their Engagement KPIs actually prove they generated more business for their organization, even though they can’t really (and unequivocally) link these Engagement KPIs to actual business and P&L-related results. – The Fournaise Marketing Group

OK, so engagement metrics don’t prove a business outcome. “Likes” don’t necessarily correlate to a sale. Thousands of views don’t automatically equal a positive ROI. But these metrics still provide valuable feedback about our content, helping us improve and optimize … right?

Well, only if those numbers can be trusted.

Shares do not equal reads

Twitter has removed share counts from its widgets, buttons, and API (meaning your other tools won’t be able to access Twitter-share counts either). Twitter’s announcement in September prompted plenty of discussion about the implications for marketers. After all, a popular (and easy) metric would disappear overnight. Yes, Twitter has its own analytics platform, where you can log in and view various metrics related to your own Twitter activity, and people may have devised other workarounds by the time you read this.

But does this really matter? How useful was that share count to us anyway?

According to Twitter, not very useful at all.

As Michael Ducker, group product manager at Twitter explains: “The tweet button counts the number of tweets that have been tweeted with the exact URL specified in the button. This count does not reflect the impact on Twitter of conversation about your content – it doesn’t count replies, quote tweets, variants of your URLs, nor does it reflect the fact that some people tweeting these URLs might have many more followers than others.”

So the share count was never an accurate measure of social engagement with our content.

You might think an inaccurate measure is still better than no measure at all. If one post has twice the shares of another, surely we can assume it performed better. Right?

Maybe not. Chartbeat handles real-time traffic measurement for sites such as Upworthy. In 2014, Chartbeat did extensive research into sharing behaviors and content effectiveness. CEO Tony Haile discussed the findings in an article for

We looked at 10,000 socially shared articles and found that there is no relationship whatsoever between the amount a piece of content is shared and the amount of attention an average reader will give that content.

It’s worth reading the article, if only for the skewering of so many social media marketing myths with what the data really shows us.

Numbers don’t lie; they just don’t tell the truth

Here’s another dodgy social media engagement metric. Unless you’ve prevented it in your settings as I have, videos automatically and silently start to play as you scroll through your Facebook newsfeed. The video may only be on your screen for a few seconds as you slowly scroll past or pause to read the update underneath, but three seconds of silent streaming is enough for Facebook to count it as a view.

Alternatively, YouTube only counts a view after approximately 30 seconds, thereby reducing the risk of the numbers becoming contaminated with accidental views or bounces. So now we have different networks measuring the same content in wildly different ways.

Already any reliance on these numbers seems flawed. But even 30 seconds doesn’t necessarily indicate whether a video is successful. Your five-minute tutorial might seem a runaway success with thousands of views, but do you know how many of those stuck it out past one minute? Two? What percentage of your viewers made it to the end? Where in the video did most people lose interest? For these numbers, you need to click on the analytics button located beneath your video on the YouTube site.

Facebook also provides more detailed video metrics for page admins who want to crunch those numbers. Yet, I wonder how often that headline number – views – still makes it into reports sent to bosses and clients to justify so-called engagement (invoice attached)?

While hundreds of shares or views are definitely preferable to tens of shares or views, these counts tell us nothing about how many people genuinely paid attention right up to the last line or final frame of your content. These numbers certainly don’t tell us how many understood and were persuaded by the message, let alone acted on it.

Other than the split-second action of hitting a button, these counts don’t measure any action relevant to the business outcomes your CFO cares about.

What are you really measuring?

Social shares and view numbers are similar to email open rates or search rankings; they reflect your ability to distribute your content, but not whether the content itself achieved its purpose.

Upworthy measures content engagement as “attention minutes,” and has even released sample code for marketers to adapt and integrate into their own platforms. Meanwhile, Medium’s key metric is the similar “total time reading,” or TTR, derived from a number of data points including scroll positions. These are measures related to the activity of reading the content and paying attention, not sharing. Medium and Upworthy care less about whether you shared their content and more about whether you read it!

We should take Twitter’s cue to become less reliant on engagement metrics such as share counts. Let’s stop hiding behind numbers that we know are flawed, inaccurate, inconsistent, open to abuse, and ultimately meaningless in assessing the quality or effectiveness of our content and social media activities. Let’s all agree to stop taking these various social media metrics out of context, attempting to assess one thing by measuring another. Of course, we must keep sharing, but we should never confuse social media distribution with content effectiveness.

This article originally appeared in the February issue of Chief Content Officer. Sign up to receive your free subscription to our bimonthly, print magazine. Find more best practices and rules of engagement for working with today’s top social media platforms. Read our Content Marketer’s Guide to Social Media Survival: 50+ Tips.

Cover image by Joseph Kalinowski/Chief Content Officer

Author: Jonathan Crossfield

If it involves putting words in a row with the occasional punctuation, then Jonathan has most likely given it a bash; from copy writing to screenwriting, blogging to journalism. He has won awards for his articles on digital marketing and his over-opinionated blog, Atomik Soapbox. Follow Jonathan on Twitter @Kimota.

Other posts by Jonathan Crossfield

  • Soyracha

    Greetings – I’ve Found Facebook Insights A Most Reliable Gauge Of Engagement – Would Love To Discuss It With You

    • Kimota

      Not sure I’d agree with ‘reliable’, cause I’m a cynical ol’ bugger who will often doubt and question everything. 😉

      As the research I cite in the article suggests, there is little correlation between shares, retweets, Likes, etc. and whether people actually read/viewed the content in a meaningful way. Remember, hitting ‘Like’ is a split-second process with very little commitment and potentially even less brand recall later.

      You’ve definitely achieved a lot more activity on Facebook than your competitors (which is still something), but I suggest digging deeper or finding alternative metrics to understand how your social media content drives actions or conversions that impact your business goals. Is there a correlation between Facebook engagement and Facebook traffic to your website, for example? And does that traffic from Facebook convert into sales at a comparable rate to other traffic sources? Which updates or content are more successful at driving those actions than others?

  • Philippe Ingels

    I think all the measurements has some relevance but in their proper context. For someone to click on a share button although they’ve not bothered to read the article as oppose to not sharing or reading it, or reading but not sharing it, all means something. I agree those metrics are easily abused but that does not mean they should be discarded. Attention minutes is a good metric but also has shortcomings. E.g. it does not measure user satisfaction. And some content are not meant to hold the user’s attention for long. It is only when combining all the measurements into a system that can make sense of the various dynamics that we will have a end-all, be-all way of measurement.

  • Stephanie Manley

    Is it me, but, I don’t think Twitter is sharing the API any longer, so it is hard to count retweets. I definitely think it is hard to effectively measure the impact of social shares. Viewing the stats as a whole is a good gauge. You can measure bounce rate, time on page, and then see where your visitor exited.

  • Steven Moher

    Interesting. So what is the most accurate means to determine your ROI?

    • Kimota

      How long is a piece of string? It’s going to depend on your bottom line business goals and how each strand of your content/social strategy contributes to that.

      I suggest setting up relevant goals in Google Analytics to track conversions and assign a monetary value to each, and then building the funnel to include the various stages and content pages within your website you want to measure. But that’s a whole other topic for someone far more expert with Google Analytics than I. 😉

  • Cris Antonio

    I’d agree – only because I’ve seen some people go on massive share and RT campaigns without actually going through the content (kudos to those who scanned, at the very least).

    I’m occasionally guilty of this myself; so it only means that content marketers should align KPIs with goals. Hey, if they want to stop with brand awareness, that’s fine and dandy. But in the long run, businesses are here to make money.

    So yes, if those shares don’t create revenue, perhaps it’s time to rethink your strategy.

  • Alice Thwaite

    Hi Jonathan, interesting article. I run the commercial department of a start up – and for us we equate traditional ‘engagement’ metrics on twitter with brand awareness. If you have engaged with our content, then we know that you as an individual have engaged with our brand, which means we have reason to think that if we reach out to you independently, then you’ll be more likely to engage with us in person.

    We’re trying to use social media to create human to human conversations – which goes against the trend of automating everything, but it does give us results.

    • Kimota

      “If you have engaged with our content, then we know you as an individual have engaged with our brand,”

      Do you really, though? That’s my point. Most of those social media metrics prove nothing of the sort and the research turned out to be even more damning than my natural skeptical nature ever suspected.

      Firstly, what is brand awareness? You can’t effectively measure that because most businesses lack the resources to actually measure genuine brand awareness (audience samples who remember our brand and know what we do after a given period of time, usually through independent surveys and focus groups). We can’t decide to measure it wih a completely different set of metrics instead.

      Plus, liking something in your FB feed doesn’t mean you even noticed, never mind remember, the brand page that shared it. I share a lot of marketing articles each week, but I’d struggle to name even half of the brands behind them as I discovered most as links shared by others within my Twitter feed.

      Secondly, brand awareness still has to lead somewhere. A new subscriber, lead, enquirer, whatever. Those metrics are far more important, so it always dismays me how brand awareness often ranks higher than these actual conversion goals in marketing survey after survey.

      The wild inaccuracy of share metrics, as mentioned in the article, means you’re attempting to measure a goal with not only the wrong metrics but also metrics that can’t even be trusted.

      Conversations are something different. If you’ve used social media content as a deliberate strategy to spark actual conversations with target customer types (and a share or like isn’t a conversation- I’m talking a minimum of comment and reply on a relevant topic), then I’d measure those conversations to assess the effectiveness of your content.

      I’m just hoping that by “results” you mean measurable actions directly related to your start-ups business model and not hoping lots of shares and likes will somehow, someday, some way equate to a sale.

  • Birkhoff Research Group

    Yes, I’ve long suspected that many people have liked the content on the sites that I’ve worked for without actually reading it. It’s also been a complaint of others I know that many people like their posts on Facebook or Twitter but never actually say anything insightful about them, perhaps because they’re reading one or two sentences at most, or not even visiting the external link at all. 🙁

  • Bess Obarotimi

    So the only way to measure effectiveness is revenue. Google analytics is very important

    • Kimota

      Not necessarily. Depends on what your overall strategic goal is. Could be to reduce customer support costs, reduce churn, increase customer value, reduce cost per acquisition or any number of goals, depending on what the business needs to achieve. But there needs to be a link to at least one positive and measurable business outcome—one the C-suite cares about. And while engagement is still necessary as one step in the journey to achieve many of those possible goals, it isn’t a strategic goal or business outcome in itself—particularly when it is measured so inaccurately and inconsistently.

      However, I agree that Google Analytics abso-blooming-lutely should be at the heart of any effectiveness reporting, as it allows you to set and more accurately measure content journeys to worthwhile goals.


    Hi, great article, thanks for sharing your experience! I need to try some of them.