Before Christmas, Neil blogged about how we’d introduced social sharing buttons as an experiment to four of our content types.
Now we’ve had the chance to look at the first 10 weeks of data, we wanted to report back.
Why we added social sharing buttons to GOV.UK
Many departments' former sites had buttons that let users easily share content to Twitter or Facebook or other social sharing sites. Several mourned their loss in the transition to GOV.UK.
Based on our previous experience with Directgov and other government sites, we did have doubts about how much these buttons would be used in practice (there are some interesting comments about this on our previous post). Also, it wasn’t a feature that had been particularly requested by end users.
Nevertheless, it was one of the things that we and colleagues from the Government Communications Service wanted to try out and evaluate.
Our findings
We’ve recently looked at data covering the period 3 December 2013 to 17 February 2014, roughly the first 10 weeks of the social sharing buttons being live on GOV.UK.
From the figures, it’s fair to say that the introduction of sharing buttons didn’t exactly set the world alight. Overall sharing rates using the buttons are pretty low, way less than 1 percent.
The introduction of sharing buttons on GOV.UK didn’t exactly set the world alight
However, there is evidence that social sharing buttons can be substantially more useful for people using mobile phones than on desktops; and - as you might guess - that some types of content are more popular than others for sharing.
Social sharing figures across GOV.UK formats
During the time period we analysed, GOV.UK URLs were shared a total of 14,078 times to Facebook and Twitter using our sharing buttons - that’s 0.2% of the total of 6.8 million pageviews.
Overall, our social sharing rate was 0.2%
News articles were by far our most viewed content format - for these, the average sharing rate was also 0.2%.
Consultations were the least shared format overall, with only 0.06% of total users using our sharing buttons.
|
Section |
Pageviews |
Social media shares |
Shares as percentage of page views |
|
News |
4,841,716 |
9,994 |
0.21% |
|
World location pages |
330,112 |
1,904 |
0.58% |
|
Speeches |
636,689 |
1,531 |
0.24% |
|
Consultations |
1,041,264 |
649 |
0.06% |
|
Total |
6,849,781 |
14,078 |
0.21% |
Subject matter influences news sharing
Within these broad content types, the extent to which a particular page was shared was largely dependent on its subject matter. There wasn’t always a direct relationship between how much a page was viewed, and how often it was shared.
This can be seen by looking at sharing rates for the five most popular news stories in the period, sharing rates for which vary between 0.18% and 1.6%.
|
Page |
Social media shares |
Pageviews |
Shares as percentage of page views |
|
356 |
97,867 |
0.36% |
|
|
UK electronic visa waiver introduced for Oman, Qatar and UAE |
287 |
157,243 |
0.18% |
|
163 |
10,007 |
1.63% |
|
|
129 |
43,138 |
0.30% |
|
|
Motorists warned about driving licence reminders received by text or email |
124 |
34,740 |
0.36% |
It’s interesting to see, by the way, that these most popular stories are straightforward, factual, explanations of changes to government schemes, or report significant data releases.
For speeches and consultations, sharing rates for the most popular content items were roughly comparable:
-
sharing rates for the most-shared speech - Nick Clegg’s speech on mental health - were 1.3% of its 3,538 pageviews
-
most-shared consultation was the one on public libraries, for which sharing rates were 1.8% of its 4033 pageviews.
Mobile users can be twice as likely to share - if the content's right
Broadly, we found mobile phone users and desktop users equally likely to use the social sharing buttons, as demonstrated by our breakdown of people sharing news stories:
| Device Category | Total social media shares | Total visits | Percentage sharing by device |
| Desktop | 6,366 | 1,694,874 | 0.38% |
| Mobile | 2,560 | 649,337 | 0.39% |
| Tablet | 1,068 | 323,487 | 0.33% |
However, for some individual news items, mobile users proved much more likely to share.
For example, on the Direct debit and abolition of the tax disc page, twice as many mobile users clicked on our social sharing buttons than desktop users (0.75% to 0.33%)
|
Pageviews of tax disc news page |
Social media shares |
Shares as a percentage of page views |
|
|
Desktop |
30,245 |
101 |
0.33% |
|
Mobile |
15,084 |
113 |
0.75% |
|
Tablet |
8,354 |
34 |
0.41% |
Similarly, on the news article on the death on Mandela, 2.2 percent of mobile users used our social sharing buttons, compared to 1.3 percent of desktop users.
There are a number of possible reasons for this - for example, the time of day that news stories are posted; or that an intrinsic 'shareability' of particular stories being suited more to mobile audiences. We'll need to do a bit more digging to find out.
Facebook and Twitter
Facebook was generally more popular than Twitter as a place to share GOV.UK content:
|
Section |
Pageviews |
Twitter shares |
Facebook shares |
|
News |
4,841,716 |
4,689 |
5,305 |
|
World location pages |
330,112 |
534 |
1,370 |
|
Speeches |
636,689 |
862 |
669 |
|
Consultations |
1,041,264 |
338 |
311 |
|
Total |
6,849,781 |
6,423 |
7,655 |
The news about tax disc abolition was shared by 10 times as many people on Facebook than on Twitter
In fact, for our most popular news story, the abolition of tax discs, 10 times as many people chose to share on Facebook than Twitter using our sharing buttons.
What next
There’s patently more data we could be looking at here.
For example, we need to compare the rates for numbers of people who choose to use GOV.UK sharing buttons with the numbers of people sharing content items in general.
We also need to benchmark our social sharing rates against similar figures from other sites (does anyone have any comparative data they can share?); and do some A/B testing on the positioning of the buttons.
But notwithstanding that, from what we’ve seen so far, our users aren’t exactly demonstrating an overwhelming case for us retaining social sharing buttons - at least on desktops.
[Image by mkhmarketing, used under creative commons]
Keep in touch. Sign up to email updates from this blog, or follow Graham on Twitter
57 comments
Comment by Matt posted on
Interesting figures.
You don't mention the main reason you'd want to have social sharing buttons, though I suspect it was a factor in your decision to try them - the inexorable rise of the social web.
Here are some Twitter numbers, for example - http://www.statisticbrain.com/twitter-statistics/
Obviously GOV.UK is very different from clickbait such as the Mail site, but the growth in users would suggest that social sharing isn't going away soon.
Do you have the breakdown of mobile users vs desktop users sharing to Twitter vs Facebook?
I have a theory that mobile users prefer Twitter cos it's a more natural fit. If it's true could help explain why mobile users can be 'twice as likely to share' some stories - cos Twitter's more news-y.
Comment by Orde Saunders posted on
Wondering what the breakdown for mobile operating systems is? Does the native sharing functionality between Android and iOS make a difference.
Comment by Andy Davies posted on
As Orde Saunders asks I'd be really interested in how much traffic is generated via sharing buttons vs people sharing the link natively.
I've a theory that native sharing on mobile reduces the need for buttons but it's only a theory
Comment by Pete Duncanson posted on
I've found there is often the demand from management that there has to be a "Facebook button and we need to be on 'the Twitter'..." which means all the usual sharing buttons not just Facebook. Wonder if your test was a caving into management thoughts of what the site simply had to have and a test to give you some numbers to use against them. We've seen similar on sites we've put sharing buttons on but regardless of how little they are used management always respond with "yes but some people are finding them useful so lets keep them!"...
Interested by the mobile stats though. I always find sharing on mobile clunky as you get taken to another site, have to login (as I normally access Twitter and Facebook via their apps not their sites) all to share a picture of a cat so end up not bothering one the "please log in" box appears. Are the stats based on completed shares, ones that have actually been posted or people who simply clicked on the button I wonder?
Either way, thanks for sharing this great info.
Pete
Comment by Graham Francis posted on
Cheers for your comment Peter. The stats are just based on people who clicked the 'social share' button, not completed shares.
Comment by Phill Price posted on
You're missing the other end of the story though. I'd love to know How many people arrived from those shared links?
Comment by Dawn posted on
Out of interest, did you track how many people came to your site by clicking on a shared link that was created via these buttons? Or if you got more traffic generally from these sites after adding the buttons? The main reason for sharing buttons existing is to provide a "nudge", which has a much bigger value for the provider than the consumer.
For now, the flow of sharing on mobile is much easier using the phone's inbuilt Share features (assuming a smart-ish phone) that link up with the twitter/facebook/pinterest app on the phone rather than the web version of those sites (assuming you have them installed).
Comment by Graham Francis posted on
Hi Dawn. No, in this we just looked at people who clicked on the social sharing buttons on the site. The buttons themselves don't generate a unique link - they're just really simple things that generate a pre-populated field in Twitter or Facebook. The advantage with this method is that they don't ask you to give away your privacy to a third party in the process; the disadvantage that they don't generate unique links for us to track. Good point about phone inbuilt sharing features - may explain why we saw surprisingly similar desktop and mobile sharing rates.
Comment by Lewis Harper posted on
I think that the content being shared in itself justifies the share-buttons' presence. By this I mean that the content being shared isn't the usual 'Top 10 Travel Tips' type content with the sole intention of SEO and traffic increase. Rather, the content is often truly valuable and of benefit.
As an example, the UK Electronic visa waiver article that was 'only' shared 287 times would have real benefit to the people that saw it-it's a niche 'market', esoteric article that can't be found elsewhere.
Comment by Benjamin Rusholme posted on
People may have shared these articles regardless.
Comment by jon walmsley posted on
Great article. Glad you posted a followup to the previous one. You never know, they could've made a massive impact!
Were the sharing figures exclusive to users pressing those share buttons, or does it include people sharing more organically (i.e. copy and paste the URL to twitter)?
Is there any comparison between cut-and-paste sharing compared with the sharing button usage?
Comment by Ross Ferguson posted on
+1 on this request.
You saw in user interviews that people were looking to copy-and-paste URLs rather than clicking on share buttons. Could you look at the 'organic' share rate for the URLs in your study?
Good bit of data detective work.
Does this mean you finally get 'that' signature?
Comment by Graham Francis posted on
Hey Ross. As you mention, in user testing on desktops we definitely saw a tendency for people to cut and paste rather than use sharing buttons. So we were expecting to see much higher sharing rates from mobiles than desktops. Weirdly, that wasn't actually the case across the board - mobile and desktop rates were within 1% of each other.
Organic shares would definitely be a good thing to benchmark against. That's our next thing to look at.
Comment by Neil Wholey posted on
Maybe the evaluation shows your content isn't interesting or useful enough to be worth sharing? It may not be the buttons that are failing to set the world alight.
Comment by Graham Francis posted on
Yeah, that's definitely a factor. Certainly, we found sharing rates differing dramatically for different content pages. Some types of content are by their nature more shareable - others are written in a way that makes them more enticing to share.
Comment by Alexis Bailey posted on
The sharing buttons are easy to miss because they're at the bottom of the article and they're not in their usual familar colours. They'd probably get used more if you made them noticable.
Comment by Graham Francis posted on
True, there are lots of variables here. A/B testing would be a great way of seeing if changing the colour or positioning of the links makes any difference - we'll try to get to this shortly.
Comment by Ed Ridout posted on
Even if the share buttons weren't used much, the reach of the content improved, potentially by a lot depending on who's doing the sharing. I understand the user need argument against them but I don't see what harm they're doing. I'm going to share this article on twitter just to annoy you 😉
Comment by Graham Francis posted on
Yeah, I know what you mean. But everything we put on a page which isn't really needed by users just adds clutter and distraction, so we do need to work hard to make sure every feature we add earns its keep.
Comment by Benjamin Rusholme posted on
I agree, share buttons are a classic example of website clutter.
Comment by Graham Spicer posted on
I agree in part, as I instinctively don't like website clutter. But does website clutter matter in itself? Maybe from a design purist point of view but I would say the main thing to think of is does their presence hinder the user completing their task. If they don't, and some people are using them, then surely you should keep them, even if they look a bit messy. I guess as Graham said, I think some A/B testing would be useful along with some industry benchmarks (although have to be a bit wary with this as comparing government/informational material with more commercial content is obviously not like-for-like).
Comment by Robin Riley posted on
Nice analysis Graham.
I think the next step would be to cross-correlate the pages with the most social shares with the pages that got the most social traffic (vs other traffic). It's not a perfect method of showing the relationship between the two (for that, you need some kind of tracking or unique URL etc.) but it's pretty decent evidence. With the proviso, of course, that a persistent correlation between the two might just mean that some pages are inherently more "social" than others.
Comment by Glynn Phillips posted on
I would agree with Robin here it would be great to at least cross reference social traffic with shares and normal traffic as it makes sense you wouldn't want to add unique urls when you have to ask for those permissions. Saying this i think this is important info for deciding if share buttons are important because because what you are really interested in here is the reach of that share. Maybe once its on a social platform people are retweeting or sharing a status and these are not accounted for in the number of button clicks but will impact the number of visits.
Comment by Graham Francis posted on
Thanks both. That seems like a good next step. We're working on something now.
Comment by Tom Szekeres posted on
Might be worth adding ?UTM_medium=twitter_share_button to the end of the link being shared via the button? Quite a simple way of measuring the traffic you get without impacting user privacy?
Comment by Mark Churchill posted on
Great post—I like your approach to the analytics, to presenting your findings, and to sharing this on the blog.
I also tweeted this using the sharing button btw.
Comment by James Crothall posted on
I agree with Mark - excellent findings.
I'd also suggest that the News, World location pages, Speeches and Consultations sections would benefit from Open Graph Protocol (ogp.me) and Twitter Card metadata (dev.twitter.com/docs/cards).
Once shared, this would encourage more clicks from Facebook and Twitter, by way of a more attractive post.
Comment by Quammi posted on
The problem I am having with social buttons is that they are usually right ontop of the article, while everyone's attention is actually occupied with reading. Or they are in the ending, when users are too busy moving on.
Comment by Chris posted on
I suspect it's also how those share buttons are presented. Lining up the buttons at the bottom of the article doesn't give anyone an incentive to share the story with friends. It's just sitting there, and very easy to miss.
Comment by shane dillon posted on
Agree with Chris, sharing buttons need to be much higher up the page rather than burying them at the bottom.
Comment by Graham Francis posted on
Thanks both - this would be something for A/B testing. But still, it's worth noting that for some types of (more shareable) content, people do seem to be finding and using these buttons where they currently are.
Comment by Joseph Holmes posted on
Well, I'm not quite sure about this. If you read an article and you like it and want to share it, I suppose the share buttons have to be after the article...Otherwise, why should I bother go on top of the page to use the button, while simply copy and paste the URL instead? - It's faster, you know...
Comment by Patrick posted on
I actually thought your share numbers were pretty good, considering the button placement and content - what level would have made you feel like they were worthwhile?
Comment by Graham Francis posted on
That's a question I'd like to throw out there. Does anyone have any comparable social sharing rates they'd be happy to give? Or experiences with A/B testing of social sharing button positioning? Would consider opening up a guest blog on this, if so...
Comment by Barbara Bermes posted on
Thanks for sharing this case study.
No word on performance so I'm wondering if you had a chance to do some performance analysis and A/B testing as well since you've included 3rd party content that could potentially uselessly bloat your site.
Can you comment on the conversion rate in regards to page speed and any other worth mentioning performance aspects (load time, page size)
Thanks
Comment by Graham Francis posted on
Hey Barbara. Just a quick note to say that our social sharing buttons *aren't* 3rd party content - they're built by us, and really do the bare minimum to enable social sharing. They're pretty much just basic links out to Twitter etc - and they don't do anything like ask you to sign away rights everything you post to 3rd parties…
Those other aspects - page load time etc - seem interesting, but my gut feeling is it's still something much more to do with whether a page is *worth* sharing.
Comment by Barbara Bermes posted on
Thanks for your reply, Graham. It's very interesting that you've decided to not use common externally hosted 3rd party scripts to include social media.
Was there a specific reason why you decided to use the simplest implementation of Twitter and Facebook share functionality, hosted on your end? Was it because of performance?
And more so, do you think the lower interest in using those buttons to share had something to do that people felt more comfortable using those social buttons via 3rd party providers, e.g Gigya etc?
Comment by Chris Condron posted on
Hello
I love the openess of this post, how it poses as many questions as it answers, and the quality of conversation below the line enabled by that openess. Big respect.
All best,
Comment by GM posted on
Is the intention for the information in any digital blog post to be read and subsequently shared? If the information is not to be shared then why post anything to the internet in the first place.
I personally expect all gov.uk pages to have a sharing button. However, that doesn't mean I will share everything I read.
As it is an interesting post, and it may be interesting to others I know that may not have seen then I will be sharing using the tweet button at least.
NB, This post came to my attention as it was shared digitally by A.N. Other.
Comment by Graham Francis posted on
Cheers GM. Yeah, we like sharing. It's just we're still not convinced that sharing buttons are having that much of an impact - lots of people seem perfectly capable of using native sharing facilities on mobile phones, or cut and pasting URLS on desktops - rather than us needing to duplicate that functionality on GOV.UK.
Comment by Rosemary posted on
Hi,
It would be interesting to see if the page views increased as a result of the introduction of share buttons. What is the percentage of traffic coming to the pages from social media sites and has it increased? If it's arguable that people on desktops are copying and pasting links into sharing sites then perhaps cross reference with stats prior to the introduction of the share button, as well as mobile use where it's less common to copy and paste links.
Many thanks,
Comment by Corey posted on
The social buttons are definitely an improvement to the blog and like many others in this topic I will use them and share this post on tweeter. I will, however, suggest using different colours for the buttons to make them more noticeable.
Best Regards
Comment by Nick Halliday posted on
Fascinating discussion.
It might also be interesting to look at trends instead of just total figures. for example is it possible to see a trend upwards or downwards with sharing. We have tended to see an increasing trend upwards as more people get used to the idea of sharing.
Personally I am happy with any sharing as one of my objectives is to make our content more shareable.
Is there any data on demographics and the likelihood of sharing of content and is this linked to your content types and user tasks?
Is there any mileage is having something at the end of a transaction saying 'I just applied for my passport online - why not do the same' and that being shareable?
Nick
Comment by slpw posted on
Social buttons have their place but they should never be on transactional pages as it takes the focus away from the task for the customer which affects your conversion rate. However on content strong pages such as this blog social buttons are a good idea to syndicate content across the web with other users
Comment by Saurabh Hooda posted on
Your social sharing buttons are at the end of the articles. Web users are always in hurry and once they finish reading the article they are generally in hurry to move to next tab/page. I believe that if you display your sharing buttons prominently at the top of the page (in addition to the ones at the bottom) then sharing percentage will Increase for sure.
Comment by Stephen posted on
All modern mobile browsers have simple share buttons so I'd get rid of.
Comment by Amy Nash posted on
I hate the share buttons that are built in the browsers, especially on a mobile device. I don't really see the point of a button on a regular browser since it is way more natural to just copy and paste the link whenever you need it.
Comment by Tim posted on
I'm from Australia, found this article via Benedict Evans.
I have to commend the transparency and sharing with votes/public.
Excellent to see!
On sharing, I agree, it's a management imperative, and this is born out by the stats.
If someone really wants to share the page, they'll copy paste the URL somewhere.
Comment by Will Robertson posted on
I use the Email button all the time from websites when I am on my Work computer and want to share to a friend or family member. Social media sharing buttons are less useful than the email button
Comment by Daniel Lampard posted on
I use the social media buttons on a regular basis when reading content on my computer. However, when I browse through my mobile phone I very rarely use them. But still, in my opinion, the change is for good. As other users said, there is better not to place the share buttons in the middle of the article or even at the beginning as users sometimes will just close the window once they finish reading. And when the buttons are at the end it will be really simple to share what you have just read.
Comment by Paul O'Gorman posted on
Great article Graham. I agree with aspects of both sides of the argument with regards to social media buttons but essentially even if the social media buttons aren't used alot they need to be there now to improve the share reach of any article posted.
Comment by Podaraci za Svatba posted on
I don't think Social Media buttons should be placed on all pages of a website. Most websites have either About Us page or Contact Us page. There you can place all buttons and links. Otherwise, the focus of the inner pages is moved to the social buttons because they are very distinctive.
Comment by stephen posted on
I agree, share buttons area unit a classic example of web site model
Comment by Edith Brown posted on
Pretty difficult that sharing buttons are not in their usual colours. I am not so sire they are really usable because they are not noticeable. The risk to miss them is really possible.
Comment by Russ posted on
I don't use Share Buttons, but I do share quite a bit. I use plug-ins from the browser (or the Share in Win 8) or copy the URL. While I have Social accounts, when I try to use the buttons, some sites attach themselves to my social account and I have to disallow the web sites.
Every month I hold a "clean the social accounts" evening. My family goes through the accounts and remove apps and sites that have attached. At least the ones we don't use. My daughter had a month where she had over 100 apps and pages. Some she had no idea what they were.
Comment by Jonathan posted on
This research echoes a similar study I did (and for similar reasons) on "print" buttons about 7 years ago.
The feeling then was that on "informational" pages similar to that of gov.uk, "print" buttons should be provided. These typically did the same thing as using File/Print in the web browser. Our main finding was that these buttons had very low usage rates in all cases other than order confirmations.
What I find interesting about the gov.uk study is that it didn't cover "native" functionality. We also couldn't tell how many people were using native File/Print. In a qualitative study we did, only about half of participants expressed interest in printing pages, and those that did tended to use File/Print, with a minority looking for the print button on the page instead. Some participants thought that the absence of a print button meant you could not print the page (or that it would format badly if you tried with File/Print).
I note that debates about having "Print" buttons on pages have now died away given that (I assume) native printing it now fully understood. At least, gov.uk doesn't have them 🙂
Might the gov.uk study indicate the same might be happening for social sharing? Will we thing such buttons quaintly irrelevant in 5 years time?
Comment by Roland posted on
I added Sharing buttons to my site and i can see the major benefit of it. The traffic for my site has increased since i started boosting my social media.