Measuring social media engagement

Facebook engagement rates fall as your fan base gets higher. At first sight that seems odd. But of course it isn’t.

Most people tend to engage with brands infrequently on social media. Seduced by a particular post they may “Like” it on a whim, but then stay away from your page. They have little motivation to comment on or share other posts unless they are real fans. And of course unless you are spending serious money, chances are they won’t see those other posts.

So most Facebook “engagements” come from people who have recently become fans. The larger your fan base, the longer your “average” fan has been with you; and because of that, larger fan bases have lower engagement rates.

This is a problem for people using a common metric for evaluating social media effectiveness: the formula “Likes + Share + Comments / Fans” as it can make it look as though your campaigns are getting worse over time.

So how can you evaluate effectiveness on Facebook and other social media platforms? There are a couple of ways.

First there are some comparative measures. For instance you might want to compare engagement rate of different types of posts or different campaign themes. This won’t solve the problem of lower engagement rates with larger fan bases but it will help you focus on improving your campaigns.

Second you can measure engagement of recent fans. By dividing the number of engagement actions (shares, likes, comments) in any one period by the number of new fans acquired in the same period you will get a useful measure of true engagement that should be comparable across time periods.

Purists might argue that the engagement actions in any one period may be coming from people who became fans in an earlier period. That’s true; but I am not sure it matters much. For a start it is probable (at least, if you are using a period of 1 month rather than 1 day) that the number of actions from “old” fans will be low; but also if you compare different months you will be comparing like with like.

Ultimately though, “engagement” on social media is a measure of limited value. So what if people have engaged with you? What matters is their behaviour. Measuring that by using your web analytics to track visits generated from social media together with dwell time, page depth, loyalty and conversions is a far better way of measuring the effectiveness of your social media campaigns.


Should UK Government use social media?

 “Seriously though guys, which one of us hasn’t embezzled and cheated the taxpayer?? #FreeMariaMiller” (@DCMS)

It’s been a bad month for public sector use of social media.

Obviously things are going wrong.

The UK Government published a set of guidelines aimed at Civil Servants back in May 2012. The Guidelines, which are surprisingly short at 3.5 pages, encourage Civil Servants to use social media but only offer very general advice. And they make no real reference to the risks of using (or indeed not using) social media.

Without detailed guidance on why and how to use social media – and what to avoid – is it surprising that public servants are making mistakes?

Is it all a waste of time?

Why does the Government encourage civil servants to use social media? Well, we’ll assume that it isn’t simply because someone somewhere feels there is a need to jump on a bandwagon.

The 2012 Guidelines set out a number of reasons to use social media including:

  • Communicate with citizens in the places they already are
  • Consult and engage
  • Be more transparent and accountable
  • Be part of the conversation

I am really not sure about the first one. I don’t really want the Government to find me in places I frequent. Like many people I am sometimes to be found in the pub. I certainly wouldn’t want to be harangued by a civil servant on my way to the bar. Why should I find it acceptable on Facebook just because it is a big media opportunity?

The government also wants to “consult and engage”. Well fair enough: but are social media platform the right sort of environment for this? The danger is that any consultation will deliver very little, and only from a self-selecting segment of an already unrepresentative audience. And a lot of “engagement” may be from people with an axe to grind: useful feedback perhaps but not necessarily representative of the whole population.

Transparency and accountability I can agree with more. Social media provide a channel (so long as it is not the only channel) for government to explain what they are doing and why they are doing it. The main Government Twitter account @GOVUK isn’t doing a lot of that though, with only around 40 tweets in the first 4 months of the year. And while social media may appear to be free, the reality is that is costs a considerable amount of management time, and potentially media spend, to communicate effectively. It is perfectly possible that a lot of “transparency and accountability” would be better and more cost-effectively achieved on formal Government platforms.

Well, these guidelines were published nearly 2 years ago. And perhaps thinking has moved on since then with all Government Departments and local Councils having their own more detailed guidelines. You’d hope so at any rate.

But do these imaginary new guidelines address risk?

  • Are adequate social media password protocols employed in all Government Departments?
  • Do all civil servants understand what it is appropriate to do on social media and what isn’t?
  • Is content posted by civil servants monitored across all social media platforms?
  • Are civil servants protected against online bullying by colleagues and angry members of the public (and vice versa)?
  • Is the return on any investment in social media measured properly?

Given current events is seems unlikely that the answer to all these questions is “yes”.

Of course it is good that the UK Government is trialling social media use. But the opportunity has been around for a long time and a rather more professional approach might be expected by now.

Hats off by the way to HMRC who, as usual with things digital, seem to be doing a great job with social media, including an informative Twitter stream and a Facebook page aimed at graduate recruits.

Isn’t $16 billion just a teeny bit expensive for WhatsApp?

$16 billion. Not even Facebook can call that small change: by some estimates it is equivalent to 13% of the company’s value. So why did Facebook spend such a lot of money on WhatsApp?

Revenue? Well so far WhatsApp are claiming they won’t be putting advertising on the platform. It’s purely a subscription play.  At $1 per year per user, WhatsApp’s 450 million users have the potential to generate half a billion dollars a year. But that means Facebook paid roughly a 30x multiple on revenues.

And that’s only the potential. Users get WhatsApp free for a year. If conversion rates after the initial free period are low, or if churn after the second year is high, then that half a billion dollars may look optimistic.

Of course WhatsApp’s user base is growing – at around 1 million a day. But even at that rate it will be 3 years before they reach 1.5 billion users and the potential for a more reasonable 10x multiple can be achieved. And 3 years can be a long time in digital technology.

So if it’s not about the money, it must be about something else. There are a couple of possibilities: developing markets and teenagers.

The first is the growth potential in emerging markets. WhatsApp is growing strongly in markets like India. Crucially, it works well on the feature phones and simpler smart phones that are popular in emerging markets. At the moment WhatsApp is a very separate brand from Facebook. But who knows whether the two services won’t be more closely integrated in the future. If that happens then WhatsApp could be a springboard for Facebook (and its ad revenues) into markets in the developing world.

The second possibility is the teenage market. While Facebook is still extremely strong in this market there are signs that it is growing weaker – as mums and grannies join up and make it less fashionable. So if Facebook feels it is loosing the teenage market could WhatsApp be a way of retaining them?

Both of those theories depend on WhatsApp and Facebook becoming better integrated in the future. That might happen. But if that isn’t the plan, then surely the intention must be to strengthen revenues in some way, either by increasing the subscription rate or by selling advertising. Well, Google didn’t know what their business model was going to be when they started!

Valuing Facebook

Reports that Facebook has raised $500m from Goldman Sachs and a Russian investment company, in a deal that values the social networking company at $50 billion, take me back to the good old days of 1999 when Freeserve was valued at around $2 billion.

OK, it’s not quite as bad as that. Freeserve had around 1 million users, which meant that each one was valued at about $2000.

In contrast Facebook has well over 500 million Unique Visitors (according to Google), valuing each one at around $100.

At least Facebook has valid business model in terms of its advertising, even if it doesn’t yet necessarily justify such a robust valuation. According to the Guardian, Facebook is predicted by eMarketer to be heading for $2 billion ad revenues in 2011, making around $4 in ad revenues per user.

Given that it is hard to square $4 with $100, Goldman Sachs presumably feels that there is other value contained within the site. There might well be.

It is possible that the advertising could generate more value. At present the industry undervalues Facebook advertising because (in part at least) click through rates aren’t great. People are generally highly engaged when visiting Facebook and so less likely to escape to advertising.

Engagement doesn’t mean they don’t see the advertising, however. And highly engaged visitors may well be more highly swayed by advertising when they see it than they would be if they were less engaged. Certainly there is evidence to that effect from the world of TV.

And the data they collect means that Facebook can deliver some great ad targeting – by age and gender as well as by keyword association.

So if Facebook were successful in selling their site as a premium branding site rather than a commoditised cpc site, then they might generate a far higher yield from the advertising inventory.

It is hard to see where else they could generate revenue, though. A premium subscription service would surely be a very niche product.

And the data they collect may well be less robust, and less valuable, that they like to think (don’t tell me you have registered your real birthday with them, or that all your Facebook friends are people you engage with socially).

There are of course opportunities to make money out of e-commerce, but beyond that it’s hard to see where real opportunities lie.

So the $100 valuation per user does seem optimistic.

And of course it does assume that Facebook is going to be a leader in the social media space for a good while: remember MySpace, remember GeoCities…