Tag Archives: twitter

Context will be King – A Prediction for 2013

This article originally appeared in Greenlight Digital’s monthly magazine [PDF]. The theme of the magazine was predictions for digital marketing in 2013

Context will be King

Take a look back through one of your old, printed, photo albums. Even with the benefit of the scrawled notes on the back of each, how much about the photos from, say, 20 years ago do you clearly recall? Take away the notes on the back of each photo and you’re left with how I currently believe social media monitoring tools see the social web; small snapshots of the past with no context other than who is in each photo.

When people directly mention a brand or a key term, we can measure it and understand it, but in the context only of itself. A single photo in your family album for example. That however is only part of the problem, what we more specifically can’t do is measure and understand what’s being said about your brand when the brand term is not specifically mentioned. A photo in your album that you may have taken but are not actually in, for example. This is the focus of my prediction for social media in 2013.

Working on a daily basis with social monitoring tools has afforded me the opportunity to explore how they viewed me in an area I would consider myself relatively influential (at the very least, passionate!); football. For my sins, I’m a big West Ham fan and since 2006 I’ve been talking about my team regularly on my blog which was fairly successful, and on Twitter since about 2007. So, from a social marketing perspective, there’s an argument for someone like me being either an influencer or someone who would be ideal to seed related campaigns. However, I’ll never be considered so, the reason why: I don’t regularly mention the exact term, “West Ham” and will therefore be very unlikely to appear in influencer search results.

Of course, I regularly talk about West Ham in the context of “the game on Saturday”, “great performance” or (more commonly!), “we’re going to struggle this year”. Importantly Tweets like this regularly garner engagement and interest and often there’ll be some sort of discussion as a result of those Tweets. I’ve not needed to mention “West Ham” for my 1500 or so followers to know what I’m talking about because they’ve come to understand that, in all likelihood, if West Ham are playing at the time of my Tweet, it will most probably be about West Ham regardless of whether I directly mention them. Analysing only volume of mentions also opens the door for false positives or, more precisely, people who have a passionate disdain for West Ham and therefore mention them a lot in what they publish. To the undiscerning social monitoring tools, this person would actually be ranked as an influencer!

By way of example here are some Kred badges of accounts suggested by a well respected social media monitoring tool to be the accounts with the most mentions of ‘West Ham’. There is one caveat however, one of these badges is for my own account although it was not even close to featuring in the ‘most mentions’ influencer list. To add a little more context, one is a football news aggregator, another the official account for the Premier League and another is the official Sky Sports account (apologies for terrible formatting/alignment!):




If you’re not familiar with Kred, the two scores displayed broadly represent the following:


- Influence: “is the ability to inspire action. It is scored on a 1,000 point scale.”

- Outreach: “reflects generosity in engaging with others and helping them spread their message.”

Ideally you want a high score in both and I, personally, would consider that to be both greater than 700 and 6 respectively. So that rules out 3 from the above list (including the official Premier League account), but not mine. The aggregator and Sky Sports obviously remain (due to sheer wealth of content and therefore term mentions – which actually makes them unrealistic outreach opportunities). All in all from those seven, only two plus my own account (745/7) would be worthy individuals to actively pursue for a digital marketing campaign related to West Ham, yet my own account didn’t appear in the original ‘influencer’ list. It is my belief that contextual analysis of social media accounts will lead to far more accurate, reliable and relevant returns on influencer discovery.

My understanding of contextual analysis would be to study factors such as what events are happening at the time of the Tweet, are there any consistently referenced terms, what key terms are mentioned in Tweets as user replies to or retweets, what content do an individual’s followers  publish and an understanding of what content appears within links that a user shares.

Through Twitter’s Interest Graph and Facebook’s Timeline (why else would they surface entire timelines if it wasn’t to improve insight for marketers?!) it’s evident that not only is this data there in some capacity, but that Twitter are actually already using it. Take this Promoted Tweet for example:


It was delivered to me when viewing the stream for my West Ham account. On the surface it’s terribly targeted; I clearly don’t live in Southampton! However, it just so happened that West Ham had played Southampton just two days before this ad was delivered to me. It’s clearly an example of a bad use of Twitter’s Interest Graph as it doesn’t relate to me in the slightest, but the principle remains.

I’m confident that until social marketers have the tools to accurately translate and work with this data, we’ll always fall short in terms of reaching outreach potential. For 2013 I think we’ll see a rise of interpretations in the contextual analysis space, improvements in accurately targeted advertising and tools that bring this ability into the mainstream.