The social media buzz chart (like the one below, we’ve probably all seen them before). This chart provides a snapshot of the volume of buzz for a particular brand, topic, event or question – but what business or marketing intelligence does the typical buzz chart provide you with?
Buzz measurement has remained as one of the key social media measurement metrics throughout the years (see here and here for examples). I’d argue with you that you get very little from analysing buzz alone, so why do you rely on it?
Let’s dispel a few myths and look at the reasons why measuring social media buzz sucks. Not to leave you hanging, I’m also giving you three new analysis approaches to extract actionable business and marketing intelligence from the thousands of social media conversations happening every minute of every day.
#1 Focusing On Conversation Peaks Doesn’t Tell You Anything New
If you were presented with a buzz chart and challenged to analyse what drove social media conversations where would you start? I regularly ask this question in presentations and set analysis challenges to see who can be most creative with their analysis. So, where would you start?
The chances are that you started with the peaks, you wanted to see what was driving more conversation on the days with the peaks.
This is a natural tendency, so don’t worry if that is where you started. However, going forward, you shouldn’t focus your attention here. Don’t you think that the brand already knows what drove those conversations, be it marketing, PR, a crisis or other effort to engage an audience. The analysis of conversation peaks will tell you very little that was not already known. Yes, it can confirm your assumptions but the more interesting insight will come from analysing what is discussed day in and day out, not what drives short term conversations.
So, how do you get from analysing peaks to understanding what drives daily conversations?
The best way to find out what is driving social media conversations is to segment the data. I’m not talking about the word clouds that are integrated into so many of the social media tools on the market, I’m talking about the segments that are important to your brand.
I would recommend a two step approach to this type of analysis. The first is setting the segments that are important to you, these may include your products, services, values, store locations, to name a few. I would then recommend using thematic analysis to uncover key topics that are driving conversations that were previously unknown to you.
By pursing this approach, you can uncover insights into what drives most of your brand conversations in relation to key business areas instead of activities that drive short-term increases in conversations. Thematic analysis will also provide you with rich insights into other areas that are of interest to your customers – the areas where you need to spend more time.
Buzz peaks will not provide you with this insight, dig deeper into the data, don’t settle for surface level automated outputs.
#2 Numbers Doesn’t Measure Context
The obsession with measuring numbers in social media has to stop. I’ll say that again, the obsession with measuring numbers in social media has to stop. Social media conversations are made up of thoughts, feelings, opinions, intents, behaviours and actions, by analysing buzz alone you miss the context of these conversations – the part of the conversation that holds business value!
Yes, people may talk about your brand and you want to know how many conversations there are but what exactly are they saying, and what impact is that having on your wider business objectives? Can you really go into a senior management or board meeting with a buzz chart?
Even when you measure sentiment with social media buzz you don’t get the insight you are looking for. In one study Coca-Cola found that around 21% of the time automated sentiment analysis will find negative sentiment when the same data analysed by humans will say that it is positive – you can’t rely on these automated metrics to gain insight from social media conversations. It is a myth that you can click a button and gather real business and marketing intelligence from social media conversations, don’t listen to the hype from social media tools, you need to put in analytical effort.
So, how do you get context from social media conversations?
From properly segmenting your data (the approach as outlined in number one, above) we get a smaller amount of data to analyse all within a specific area of your business. It is then a lot easier to dig deeper into these conversations and look to see what is driving the conversations, is it something you are doing well? Is it something you are doing bad? Does it relate to specific marketing campaigns, PR campaigns or members of staff?
We can find out exactly what the context of these conversations are – and moving beyond sentiment because sentiment sucks. For example, in January I analysed social media conversations around New Year’s Resolutions and found that having financial freedom drove the highest level of conversations with people indenting to save more, get out of debt and be better with budgeting. This contextual information is more valuable than analysing the overall buzz of New Year’s Resolutions.
Contextual information is more valuable than knowing the volume alone, you can begin to tailor communications, solve problems and be more competitive with real customer insights and marketing intelligence.
#3 Buzz Doesn’t Directly Link to Sales
It could be easy to assume that the more a brand, product or service is discussed in social media, the higher the interest and purchase probability. Unfortunately, this is not the case – buzz does not necessarily translate into sales.
Brandwatch have published several studies that highlight that the most buzzed about product was not the product that had the highest rate of purchase. For example, one study highlighted that higher levels of buzz did not translate into computer game sales. Game one had 70,143 buzz conversations and game two had 30,456 buzz conversations but when sales transactions were analysed, game two sold more than double the units than game one.
So, what is the best way to predict purchase?
Instead of relying on buzz metrics alone you can measure the intent to purchase within the set social media conversations. Intent to purchase is a better predictor of sales success than measuring buzz. Going back to Brandwatch’s computer game study, when intent to purchase was measured, game one was found to have 1,295 intent to purchase mentions and game two had 4,012 intent to purchase mentions. Measuring intent to purchase is a better indicator of sales success than buzz – stop measuring buzz alone!
In the retail sector, H&M found that buzz did not drive sales of celebrity endorsed campaigns. Four celebrity campaigns from David Beckham, Beyoncé, Lana Del Ray and Vanessa Paradis were analysed and again buzz was not a strong predictor of sales intent. While David Beckham drove highest buzz levels, it was Beyoncé who had more business value through purchase intent – don’t be fooled by your own buzz, dig deeper!
Measuring on buzz is a risky business, there is no segmentation, no context and no intent. If you keep on using buzz metrics to track success you will never be able to link social media to key business performance or use the customer intelligence from social media conversations to make your brand stronger and more relevant. Dig deeper into the data to unlock the real value from social media conversations.