Book a Call

The Raw Social Media Data Hiccups You Need To Look Out For, And How To Solve Them

Sep 22, 2015 1:32:41 PM

Social Media Data Hiccups To Avoid

“Social insights isn’t just data, it’s knowing what that data means”
Karin Robinson: Associate Director, Social | OgilvyOne Worldwide

Social media data can lead to insights that present you with a clearer idea about what your audience are thinking, how they view your brand, how you can target them, and how to respond to them. But if you aren’t smart with how you’re gathering and implementing your data, it might not be as useful as you’d like it to be.

Karin Robinson: Associate Director, Social | OgilvyOne Worldwide Karin Robinson: Associate Director, Social | OgilvyOne Worldwide

Speaking at Social Media Week: London, Ogilvy’s Karin Robinson identified ways that raw data can tell the wrong story, as well some steps you can take to get the right one. An understanding of the potential flaws in some common measures will help you to extract a more accurate picture of your audience, and where your brand fits within their lives. We’ve highlighted two key points from her talk that got our attention, and also added our own thoughts in order to get you over these hurdles.

What Are The Potential Hiccups?

1. Demographics On Their Own

For decades they have formed the cornerstone of marketing strategies, basic facets of an individual such as their age, location, and gender. While they can be helpful to know, there’s now much more out there to see. There’s also a mix of things that may impact the precise accuracy of the data. As a result, it’s important to keep a critical eye on changes and be ready to look deeper if something looks odd.

      “In 2010, many Twitter users changed their location to Tehran to help confuse the Iranian government, who were arresting people who were protesting online. This then caused a lot of problems with social data, as suddenly there was a massive spike in users from Tehran appearing in brands’ social data. Meanwhile, targeting by age can be problematic too. Only a small percentage of Twitter users disclose how old they are, and do you really believe that everyone on Facebook has been honest about their age?”

Karin Robinson: Associate Director, Social | OgilvyOne Worldwide (Subsequent quotes are Karin's too)

2. Sentiment Analysis

For some brands, such as British Gas, sentiment analysis is one of the main things they want to measure on social media. But before you meticulously count every little green plus and red minus, it’s worth remembering what you’re actually trying to uncover with your social insights.

    “Sentiment analysis tools can be good for getting a very broad picture of a big audience. The problems can arise when you’re looking at a specific audience, or if you want a nuanced understanding of what people are saying. Sentiment tools only say positive or negative, what about worried, confused, or interested? Is the sentiment actually about your brand? What about sarcasm, slang or context?”

New Call-to-action

How Do We Solve These Problems?

Ignoring these issues isn’t an option, getting the highest quality of insights from your data is still a primary goal. To ensure that you meet that goal, here are some ideas from both Karin and SocialBro on how you can ensure that your data provides with with actionable information which can be trusted.

Look At Individual Users

When extracting data about your audience, it’s worth keeping in mind the individuals that the data is based on. They can then be used to check how accurate your data-derived hypotheses about your audience are, both in terms of who you are looking to target and to assess performance of previous campaigns. High quality segmentation tools will allow you to see the precise Twitter users who are forming your data within customisable parameters.

Consider The Context Behind The Data

Marketers need to retain an awareness of the context of their results, questions always need to be asked. If there’s something unexpected or unusual appearing in your social data, it’s always worth taking a closer look at what’s causing it before letting it influence your marketing strategy. This was one of the key points Karin bought up...

    “Consider the context behind the data, and how that could potentially skew it. Check that what you’re finding aligns with other data, if there are significant differences then research what could be causing them before acting on them.”

Look At Behaviours, Not Just Demographics

People generally don’t describe themselves as a ‘22-34 year-old professional female’. While it may be a useful bracket to start with, taking a deeper look will give you a much clearer idea on who your audience are, and give you a far more accurate persona for your targeting. Karin went into depth about why this should be something marketers take the time to do.

    “When looking at your data, go past the basic demographics and try to build a more fleshed-out, realistic image of the customer. You should routinely be layering in behavioural and interest-based targets that can have the added benefit of sweeping up customers with a proven purchase intent who would have been left out by a straightforward demographic sweep. After all, unless you’re marketing jock straps or tampons, you shouldn’t market your product only to one gender.”

Utilize Platform-Specific Data

Different people use different social networks in different ways. Your audience on Twitter may be unique in their own ways, which can lead to highly accurate campaigns tailored towards them, resulting in a level of performance unattainable from a one-size-fits all social campaign.

Experiment With Data From Other Platforms

As mentioned earlier, the social data gathered from Twitter users should be used to inform your Twitter activity. But if it’s performing well there, it’s worth experimenting with expanding facets of those Twitter-specific aspects of your campaign out to other areas of your marketing and communications strategy. Innocent Drinks did similar to this last year when they decided to expand one of their high-performing Tweets into a print advert (more details here).

Use Three Dimensional Data

No data is an island. A single statistic or metric won’t always tell the whole story, which is why it’s important to take data from a variety of sources and use it all together to inform your strategy. The importance of three dimensional data was something Karin was keen to emphasise.

    “One piece of data in isolation could be misleading. Using a mix of different types of data will assist you in building a better picture of your audience and brand. Pulling out sample posts to look at the types of language being used there, looking at your web traffic sources, and other data-gathering activities should be undertaken to help you gain true social insights from your data.”

As shown, these bumps in the road are certainly not a cause for abandoning your efforts to gain insights about your social audience. In fact, overcoming them will lead to a richer picture of your audience and where your brand fits into their lives.

Twitter Marketing - SocialBro

Subscribe to our Newsletter

Subscribe to our Newsletter

Comments (1)