Interview: Carnegie Mellon Professor Ari Lightman On How Students Are Empowered By Learning To Use Brandwatch Consumer Research
By Kara FinnertyJun 10
Published December 2nd 2014
I have to admit, the data geek inside me squealed a little when I found out we were launching minute-by-minute charting. So many opportunities for some super cool analysis and very few platforms allow you to get this granular.
As we said in the launch post, there are many many different ways of using this feature, along with its accompanying time picker options, but today we’re focusing all on ads and television programmes.
So, enough intro… let’s get into the data and explore what minute-by-minute charting can let you do.
Today, we’re going to take everyone’s favourite, the John Lewis Christmas ad – an event in itself these days – and look at the online response. Whilst this is a UK ad, the same analysis can be applied to any TV advert or show (to those from outside of the UK, here’s a bit of background on why the UK go so crazy for what is essentially an advert designed to sell us woolly socks and expensive appliances).
Of course, if you were John Lewis – or even a competitor – you would want to know what the response was like. In the past, you could have plotted this by hour, which would show you a big peak in the hour the video was posted online (just after 8am).
Interesting, but what happens when you get more granular? See what happens when you chart this minute-by-minute for the first few hours of it being online:
First of all, you see a spike here that wasn’t apparent in the hourly breakdown (because, when looking at it hourly, the total is an overall lower volume).
This spike is driven by a tweet about the ad by YouTube blogger Dan Howell being retweeted by many of his 1.6M followers – gaining much higher engagement than John Lewis’ own tweets.
It doesn’t look like John Lewis responded or engaged with him – possibly a missed opportunity there to capitalise on his huge fan base. Engaging with influencers can be a great way to build awareness and a strong, positive brand image among new audiences.
Charting by minute allows you to find these peaks and uncover the influencers helping to drive conversation about your content that you might otherwise have missed.
We can also see that conversation about the advert slowly built over the morning as people discovered it online since its announcement at 8.07am, hitting a peak for the morning at just before 10am.
Conversation volumes dropped after lunchtime and then remained at a fairly consistent – and not insignificant- volume for the rest of the day. Other than the occasional tweet thanking people for liking the ad, John Lewis don’t appear to have tweeted much else – perhaps another tweet in the afternoon linking to the ad could have pushed up engagement a little more for a secondary peak.
So, what about when the advert was actually shown on TV, like in the old days?
Well, John Lewis premiered their ad on Channel 4 the following day after announcing on social, during the evening showing of Gogglebox (a TV show about people watching TV, for those who don’t know) at 9pm.
Looking at the chatter about the advert, it’s pretty obvious the moment when it came on TV:
Unlike the Twitter reveal, we can see that the TV showing resulted in a very sudden, sharp spike in conversation – a much higher volume in the following few minutes than we saw for the Twitter reveal.
However, it also died down much quicker and the total volume was much lower – as shown by the hourly chart below – suggesting that an advert on TV is quickly forgotten and doesn’t have the same sharing/spreading nature. In fact, the day they launched on social had nearly 3 times the volume of conversation than the day of the TV premier, with conversation gradually dwindling since then, suggesting the social reveal dampened the excitement for the TV ad a little.
If we dive into this data a little bit more we come across an interesting nugget of information.
Many of those tweeting about the advert whilst it was on TV are actually retweeting the ‘unveiling’ tweet from John Lewis from the day before, rather than writing their own tweets.
In fact, if we compare tweets over the preceding few hours to the time when the advert was being aired, you can see how many people have sought out or found that tweet online and then retweeted when the ad was on TV – possibly because they wanted to share the ‘official’ tweet, or maybe due to laziness in writing their own!
Curiously, John Lewis didn’t tweet at all whilst the advert was airing or directly before/after, which seems again like a bit of a missed opportunity, although they have been doing a good job of engaging with those tweeting about enjoying the ad.
Had they tweeted another link to the advert while it was airing on TV and engaged more with consumers, they might have received yet more retweets by those watching the ad and more sustained online engagement past the actual TV slot, curtailing that big drop after the initial spike in conversation.
If we plot the #montythepenguin hashtag over the same period, we can see that it follows a very similar pattern of the overall conversation for both the social launch and the TV premier – suggesting that many of those tweeting about the ad also used the hashtag – a very wise move of John Lewis to include this in both the online and televised versions.
There is so much more we could analyse with these new features; it really does make us a little bit excited.
Being able to specify particular time periods on any component means we can analyse the topics and sentiment of conversation at the exact moments we care about, as well as links shared, hashtags, @mentions….
You can also chart many more metrics minute-by-minute, including your own categories and tags, page types, sentiment, earned and owned media and more to understand how conversation is changing over time in response to every little moment.
We’ll leave you to continue the journey of discovery yourselves – only your imagination can stop you!
We’ll also be posting on this blog over the following week with more examples of different use cases for the feature, so keep an eye out here on the blog or follow us on one of our social channels.
If you have any great ideas, amazing discoveries or feedback on the feature, then do let us know in the comments below. Most importantly, have fun!
If you’re a Brandwatch user and want to know more about where to find the features and how to use them, head over to the Help Center for some super simple user guides.
If you’re not a client but want to know more about this or any of our other features, get in touch.
Love this release or got a great idea for using it? Let us know in the comments or tweet us @brandwatch!