The last time we discussed influencer marketing on Link Humans we covered a number of topics, including the definition of an influencer, the various types of influencers you may encounter, how to work with them, and more.
There’s one important aspect that we left out: how to pick the best influencers. Whether you need them for a campaign, or you just want to expand your brand reach by partnering with influencers, or for whatever other reason – how do you know which influencers are the best to work with? How do you pick the best influencers to make your next campaign or endorsement a success?
It’s not uncommon for brands to pair up with influencers just because of their name and notoriety, but that is not always the best approach. For instance, a partnership between Adidas and David Beckham makes total sense, while a partnership between Burger King and David Beckham will make you raise an eyebrow.
There are 5 main things you need to look out for when choosing influencers to work with:
- Relevance: the influencer needs to be relevant to your brand and/or to your campaign. Now, they don’t have to be an expert in your industry or an expert in what you do, but they do need to be relevant to either your brand or your planned campaign. It makes sense for UK telco EE to partner up with Hollywood actor Kevin Bacon for their advertising, with the ads focusing on how fast you can stream movies and other media on EE’s 4G network, and having a famous actor tell the story. It made sense for Apple to partner up with U2 back in 2004 to advertise the iPods, thus focusing heavily on music. Sometimes a great campaign calls for a partnership with someone who the audience wouldn’t normally think of when thinking of your brand. Think, for example, of the time comedy duo David Mitchell and Robert Webb were recruited by Apple for an ad pitting Mac against PC. Comedians aren’t what comes to mind when you think of Apple or Microsoft, but the ad worked so well because the two comedians were relevant to the campaign idea: a humorous side by side comparison of Mac and PC, showing PC as a confused, scruffy Mitchell, and Mac as a cool, relaxed Webb. Campaigns or brand endorsements with relevant influencers (celebrities or otherwise) are very often the ones with the highest brand memorability, i.e. how likely you are to remember the ad in the long term, very often to higher acquisitions.
- Reputation: there are some influencers which you may not want to work with based on the negative perception around them. Working with an influencer won’t leave your brand sentiment unchanged, as the influencer’s perception can easily rub off on your brand: a partnership with a positively received influencer may work well in favour, while working with an influencer who’s being received quite negatively may cause a backlash against your brand. This includes people who are currently going through highly publicised legal troubles, or people who are notoriously disliked on social media. For example, what would your perception of Innocent be if they were to announce a partnership with the infamous Katie Hopkins? Would you still get your Caramel Frappuccino from Starbucks if they were suddenly to collaborate with Protein World, considering the “Twitter storm” they went through last year? Now, that’s not to say that such collaborations wouldn’t blow up online: sure, they may go viral, but most likely for all the wrong reasons.
- Brand perception: check if they’ve spoken positively or negatively about you. When looking at Twitter, most free tools will only give you either one week or one month worth of data, due to limitations in Twitter’s API. (Unfortunately, with the untimely death of Topsy, the only other free advanced tool that lets you search for any Tweets historically is Twitter’s own advanced search.) When looking at other sources outside of Twitter, a Google search can often help, especially since public Twitter posts and public Facebook posts are indexed by Google and other search engines anyway. When doing this search, search for the influencer’s name (including their social handles) and your brand name, products and services. So, if you’re putting this in your social listening tool, your Boolean query will look like this:
(\[influencer’s name OR influencer’s social handles]) AND (\[brand name OR brand products OR brand services])
Make sure you also include names of notable people in your company, so CEO or any influencers who may work for you. For instance, when looking at Microsoft you’d include CEO Satya Nadella, in the same way that you may include current CEO Tim Cook and late CEO Steve Jobs when looking at Apple. See if they’ve ever mentioned your brand, be it positively or negatively. If they’ve ever mentioned you negatively to their audience and now all of a sudden they’re sponsoring you to the same audience, they may be looked at as sellouts, the campaign won’t feel genuine at all, with the focus shifting to what favours (monetary or otherwise) your brand must have offered to the influencer to jump on board. This may easily result in a PR disasters if not handled properly.
Also, don’t forget to do the same exercise looking at your competitors: while the influencer may not have mentioned you in the past, they may be avid fans of your competitors – or perhaps they absolutely hate your competitors. Either way, you need this background information before you choose which influencer(s) to contact.
- Engagement: always check how engaged each influencer is online. This is essential information that you need when putting together the KPIs for your campaign. Don’t only check how engaged they are online, but also how much engagement they get as a result of what they put out, and the sentiment of that engagement. You do get some influencers who still get a considerable amount of engagement even when they don’t post anything on social media. By doing this research, you’ll know what kind of comments and feedback to expect from their audience and their extended reach.
- Potential reach: perhaps you know how many times you want your influencer to post about your campaign or brand (at minimum). However, what will the effect of that be? How will they amplify their posts? How many people are likely to see content when it’s posted from their account(s) vs. yours? This is very important to know especially if you’re planning on using paid reach for your campaign posts. If you have a target reach you want to achieve, you first need to know how much that influencer achieves on a regular basis before you can then calculate how much more you need to spend to get to your desired reach. For example, if you’re aiming for a total reach of 1 million people and this influencer has a potential reach of 2 million people on Twitter alone (organically), you may not need to spend much money to amplify your reach. You may even choose to simply rely on the combined organic reach from your own accounts and your influencer, although some investment is always recommended.
A Quick Note on Engagement and Reach
While we could’ve easily bundled reach under engagement (as they’re both social analytics metrics after all), it’s important to look at them differently. Reach is the equivalent of website traffic, while social engagement is the equivalent of on-site engagement (are visitors engaged on the site? How much are they reading?). Both metrics lead to each other: reach drives engagement and engagement drives reach. While these two metrics are related, it’s important to take an in depth look at both metrics individually. Remember this when putting together your campaign reports.
The Problem with Social Influence Scores
It’s quite common for brands and agencies alike to rely on social influence scores to judge whether they should partner with an influencer or not: they’ll pick a list of 5, 10 or more influencers, then choose the ones with the highest influence score, and then either send them all a request, or just pick the influencer with the highest score.
Most marketers are familiar with social influence scores: Kred, PeerIndex (which was acquired by Brandwatch in 2014), and the most famous one – Klout. Are these scores still worth looking at? You can definitely check them out, but they’re not really metrics to base decisions on. Here are two reasons why: relevance and offline influence.
We’ve already mentioned relevance as one of the first factors to look at when choosing an influencer. Relevance is one thing that influence scores rarely get right. For instance, my Klout score is currently at 57. If we’re solely relying on Klout scores, then that makes me as influential as:
- Will McInnes, Brandwatch CMO
- Ronan Dunne, O2/Telefónica UK CEO
- Michel Van Der Bel, Microsoft UK CEO
- Kevin Systrom, Instagram CEO
- Dong Nguyen, creator of Flappy Bird
— Dong Nguyen (@dongatory) December 22, 2015
As flattering as that sounds, I can assure you that I’m not as influential as these people.
Next we have another flaw: offline influence. My Klout is usually around 60 – between 60 and 65 when I’m in my chatty moments, or between 55 and 60 when I don’t tweet as often. Klout looks at your social engagement for the past 90 days, and as I’ve been fairly quiet on social media in the last few months, my Klout has gone down a fair bit, hitting a low of 55 in October, and now fairly stable at 57. My situation isn’t vastly different from others, including people who go on “social media breaks” for whatever reason: writers taking time off from social media to focus on their next book, artists taking a break from posting online to focus on their next opus etc. Social influence scores see that break in engagement as a decrease in influence. Does that really mean that that influencer is less influential now? Is Ed Sheeran less influential now that he’s taking a “social media sabbatical” to focus on his work? Definitely not, although Klout and similar scores don’t understand that. They don’t take offline influence into consideration, i.e. the influence that a person has regardless of how often they post on a social network.
Influence scores are nice indicators, but they’re not the best metrics to base business decisions on, including which influencer to attach your brand to for a campaign or for a long term cause. (You can find a more in depth analysis of social influence scores and their flaws here.)
Influence scores: good to use, bad to rely on.
Tools to Track Influencers
This checklist requires a fair bit of research from your part. If you already have social analytics and social listening tools available, then you’re ready to go. What if you don’t have any of these tools? Not to worry – you’ll still be able to go through this checklist with free (or fairly inexpensive) tools. Whether you used to rely on the now defunct Topsy, or you don’t know where to start, you can find some great suggestions here, looking at Topsy’s alternatives.
If you decide to go through any of those tools, I’m sure you’ll find yourself using two or more together. You could perhaps use a tool like Keyhole to run some advanced searches to find out whether an influencer has spoken about a number of topics (relevance); you could then use TalkWalker to find if they’ve ever mentioned your brand or any of your products and services, positively or negatively (brand perception); TalkWalker can also help you see how this influencer is perceived on social media (reputation); Twitonomy is a great tool that measures the social performance of influencers, looking at how engaged they are on Twitter and how much engagement they get in return (engagement), while giving you some insights into their reach too (potential reach). That’s 3 free* tools to cover the 5 points of our influencer checklist.
(* While TalkWalker is free to use, you’ll need to get on their paying plans to search for mentions older than one month. You can see what you can get for their paying plans here.)
What Else Do I Need to Know?
We’ve covered pretty much everything you need to know about influencer marketing here. If, however, you have any questions that aren’t covered in either post, feel free to let us know in the comments below.