Jon Thomas
Jon Thomas
Communications Director

Making Sense of Social-Media Timing

This post originally appeared in our October ’12 issue of “Live Report from the Future of Marketing,” our monthly Post-Advertising newsletter. Subscribe for free here.

Data can be maddening. Like Play-Doh, it can be molded into a variety of shapes and sizes and can resemble whatever the creator wants it to resemble. Let me explain.

Imagine you’re a salesman and you have only one hour to sell your product (and you can’t divide that hour) to a roomful of potential customers. You can choose any day of the week, any time of day or night. If you don’t have any further information, it’s pretty hard to decide when to do your selling. You’ll probably just follow all the other salesmen. 

But I’ve got some data to give you. I know that the room with your potential customers is fullest on Wednesday from 1 to 4 p.m. Does that help? 

What if I also told you that while Wednesday from 1 to 4 p.m. provides the most potential clients, Saturday from 10 a.m. to 2 p.m. provides the highest percentage of those who are willing and ready to buy? On Saturday the room is noticeably less full (of both salesmen and potential customers), but the likelihood of a sale is much greater.

When would you sell? Note: There are no right answers…or wrong ones. 

Data Deluge

The hypothetical circumstance I just laid out is not very far from the real-life situation that community managers face when developing content plans for brands. Marketing-automation platforms and link shorteners have allowed massive data sets to be compiled and compared to see exactly when audiences are using social media, when they’re engaging, what type of content they engage with most, how these elements vary by platform and more. But we’ve seen people come to different conclusions about social-media timing depending on where the data is coming from. 

This past summer Buddy Media conducted a Twitter engagement survey that analyzed user engagement of more than 320 Twitter handles for the world’s biggest brands. What it found was that tweeting during the busy hours of 8 a.m. to 7 p.m. generated 30 percent greater engagement (a combination of replies and re-tweets) than tweets published outside that frame. Twitter engagement rates for brands, however, are 17 percent higher on Saturday and Sunday as compared with weekdays. Brands in the publishing industry, especially, could take advantage of this, since their weekend engagement rate is 29 percent higher than average. 

Going further down the rabbit hole, link-shortening service Bit.ly also conducted a study of its link-click data, which spans a variety of social-media channels (wherever users want to paste their links). It concluded that with Twitter, posting in the afternoon from Monday to Wednesday is your best chance for achieving a high click count, with 1 to 3 p.m. EST on Wednesday being the sweet spot. Also, and quite contrary to Buddy Media’s recommendations, Bit.ly recommends not tweeting after 3 p.m. on Friday, since, far from being a gateway for driving traffic to your content, it appears that Twitter doesn’t work on weekends (according to Bit.ly). The chart below is a little confusing but day of the week is the Y axis with Monday at the top and hour of day is the X axis with 12am EST on the far left. (Click to enlarge) 

Dan Zarrella, Social Media Scientist at HubSpot, has conducted extensive studies on social-media data as well and found that Friday, Saturday and Sunday had the highest click-through rates, as did tweets made after 1 p.m. (If you’d like to learn more, go check out his Science of Timing webinar).

Is your mind spinning yet?

Let’s look at Facebook

We know that audiences and experiences differ from platform to platform, especially from Facebook to Twitter. Buddy Media also conducted a Facebook study, which found that Wednesday is the worst day for Facebook engagement (7.4 percent below average). Plus, as on Twitter, Wednesday is the most active publishing date for brands, which means that your brand will have lots of competition.

Another study, though slightly dated (2007–2010), by Virtrue, produced one similar conclusion that Wednesday around 3 p.m. is the absolute peak for activity, but also found that Facebook posts made in the morning were 39.7 percent more effective. 

Zarrella’s studies found that Facebook posts made later in the day (6 to 8 p.m.) get the most Likes and Shares (contrary to Virtrue’s findings), while Facebook posts made on the weekend have the highest Like percentage.

3 Questions with Dan Zarrella

For a marketer, consuming all this data still leaves questions, especially when it comes to reaching conclusions for varied industries. To get answers from someone who makes a living poring over this data, we contacted Dan Zarrella with some questions that may linger for brands. The bolded emphasis is our own.

1. While understanding that every brand is different and should test and learn, what’s a brand to do to sort out this confusion?

That’s really the key with any marketing data gleaned from large data sets. All my work, for instance, comes from big databases of thousands, millions and, in some cases, billions of rows of data across multiple industries, company and audience sizes, countries and time zones. Brands should use these kinds of recommendations not as set-in-stone laws but rather as jumping-off points for experimentation. If they find a study recommending a time different from the ones they’re currently using, they should experiment with it. If it works, great; if not, keep testing.

2. If your brand isn’t one of the 320 largest, should you pay less attention to Buddy Media’s findings and more to Bit.ly’s?

The data you should pay the most attention to is your own. Test the two recommendations on your own audience and content, and see which works best. It may also be that each time can work for your audience, depending on the type of content you’re promoting. Always be testing.

3. Is it worth it to change the standard approach and start actively posting on weekends and the tail end of the week? In other words, if your brand is doing the nine-to-five with social, how much should your approach change? 

If you’re not doing any weekend or late-week posting right now, you should make an effort to try some for a time, perhaps a few weeks or a couple of months. You may find that certain kinds of content work best on the weekends, or that there isn’t the big drop-off in response rates you expected on the weekends. And don’t be afraid to use scheduled posting to manage this. You don’t have to be glued to your computer to do this kind of testing.

*Make sure to visit Dan’s blog as well as HubSpot’s blog for in-depth information on social-data discoveries and inbound marketing, and pick up his book Zarrella’s Hierarchy of Contagiousness on Amazon. 

Setting Your Content Plan

A number of conclusions can be drawn from all this data, but the most important one may be that the data tells a variety of stories and that the truth lies somewhere in between them. Every audience is different, and your brand and/or industry may be drastically different from those featured in the study.

Always keep an open mind when viewing social-media studies. Don’t shy away from them, but realize that it’s impossible to define specific content-timing rules that will apply to every brand in every industry. Instead, use that data and test it. If a new study suggests that midnight is the optimal time to reach your niche audience, test it, and pay attention to the results. If it doesn’t work, it doesn’t work. No brand has gone under because it tested different theories regarding its social-media content. It’s possible, however, that a brand could suffer if it refused to change, or at least test, its content approach.

How do you decide when and where to post your content? Are you utilizing off-hours and/or the weekends?

  • bsimi

    Great post on my fave SM persona!

  • http://www.postadvertising.com Jon Thomas

    Thanks Brian!