Insights Spotting the signal in the noise

The Power and Potential of Focus™ When You Can’t See Ahead

We’re all constantly forming and refining our opinions by reading what others have to say online. This is true whether we are consumers making a purchase decision, or marketers trying to gauge the impact of a recent communication campaign. 

We might have a host of questions about consumers or the brand experience or the competitive landscape (and sometimes we might think we know ourselves better than we actually do). What are my customers interested in? What are their pain points? Are we spending in the right places? Is my campaign actually working? How do I compare with the competition?

More often than not, the answers to our pressing questions are already out there. But the numbers are daunting. For a given theme, the number of social conversations can range anywhere from 50,000 to over a million. What we need is a methodical process to identify and extract insights. And that’s where FocusTM comes in.

1. Gathering, and hunting

We begin with trillions of online conversations. Searching and scanning online posts, comments, pages, customer reviews, blogs — so in that sense, it’s a lot wider than just social media.

But that’s too much noise, too much social chatter. So we iteratively refine our search keywords to identify relevant conversations based on themes, geographies, demographics. 

Clients help us decide on the parameters for the scan based on their goals. We can then narrow down our search by timeframe, travel backwards and forwards in time, say for example, the last 12 to 36 months. Or we can narrow by geography and language, filtering out only the relevant conversations. Or by channels, say just Facebook or just one major review site. Our algorithmic sampling is based on metadata — keyword relevance, age or gender, timestamps, and content richness. 

Our analysis is iterative and we can easily backtrack and refine, which isn’t really possible with traditional research methodologies. We’re also finding interesting applications for our longitudinal studies. Say we’re analysing the spending habits of Gen Z or the long-term impact of COVID-19, it would be unrealistic to expect all the answers from a single research study. Instead, the findings need to evolve to keep track of recent developments. FocusTM makes it easy to go back or go forward and ask a slightly different set of questions. 

So we’re down to 500 to 1000 samples. But we still have some way to go.

2. More human than human

Why can’t machines do this next part of the job? Machines rely on having lots of data, and a fairly standardised set of instructions. Humans, on the other hand, can draw fairly accurate conclusions from smaller volumes of data, they can generate hyper-specific micro tags manually based on their deep subject expertise.

In this next step, our team of analysts sort and hand-tag conversations. The advantages of the human element are evident when you think about how much we can learn from context.  

Say for instance we’re studying Instagram posts for a lifestyle brand. We have a single post from a restaurant, but we’d like to know more. Why did they post in the restaurant?

Is it because of a famous chef? Is it because the food looks good? We can go back and study other posts from the same author to gain a depth of insight that would be impossible with a purely AI-based approach.

If we’re working across global markets, we draw from Weibo in China, from Twitter, from local forums like Naver in Korea. Here the process is carried out by local language speakers so we can capture all the little nuances that we might otherwise miss. We’ve already completed studies in Japan, China, Korea, Russia, Poland, UK, Hong Kong, Singapore, and Dubai.

Another advantage is working with otherwise inaccessible groups. Traditionally, for instance, it’s difficult to research tourists, because you can’t get tourists to a focus group, but with FocusTM it’s actually quite easy. One of our studies was specifically around Chinese travellers in Dubai. We looked at not only what they were posting in Dubai but the discussions they were having on Chinese forums planning their trip. This has interesting applications, we can figure out how consumer priorities have changed post-COVID, if we’re looking at Chinese domestic travel, we can discover what they’re most interested in now, and how retail and hospitality sectors can evolve their offering.

3. Final stop, and it’s not a 200-page report

Our final step is where we study and synthesise our tagged results. At this point we’re looking for trends and correlations, we like to end up with four or five practical recommendations that our clients can immediately act on. 

There’s a reason we chose not to present our conclusions in a 100-page powerpoint. This is sometimes a bit of a challenge for clients who’ve come to expect the comfort of really long report, because that’s what you might get from a traditional platform. Maybe there are nuggets of gold buried somewhere in there, but no one’s entirely sure, because no one actually gets to the very end. 

But we like to keep our narrative tight. The data is statistically significant, it’s also rich in depth. The findings may point to new areas that need further exploration. So we have the ability to follow-up with more in-depth persona research, or engage with a core group of customers over an extended period of time to keep abreast of trends. 

FocusTM really is a bit unusual in terms of its capabilities, and clients aren’t always sure where to put us on their shelf. We like to think we combine the best of both — AI-based social listening tools and conventional research, qualitative understanding with quantitative relevance. 

The Best Breakthroughs Come From Moments of Listening

One of the challenges is that everyone talks about using customer insights in the industry, but we know from experience that it’s not the case at all. Often products and campaigns are shaped by limiting mindsets, or an old school shoot from the hip attitude that relies on minimal research, pushing ideas and concepts that might appeal to a small group of decision makers, without any actual input from end users. It’s a luxury we can hardly afford. With FocusTM, we try to impose a rigour on ourselves that we actually, properly develop the insights first, so that when brands say they’re “putting customers at the heart of their experience,” they’re actually doing exactly what they say. 

It’s a bold approach, but we like to break it down to an executable set of recommendations. Transformation doesn’t have to come all at once, but with the first step, we demonstrate a readiness for action, we build the momentum for change.