After a series of articles on influencer marketing, we decided to talk about another topic close to our hearts – social media research. I have an idea to share, but let me first explain how we got here.
Social media data is rich, vast and insightful. Soon I will have been working in this field for 10 years and it has been rewarding to see how good market research can bring companies closer to their customers. Eventually bringing people closer to the products they love.
Real innovation comes by rarely though. AI and Machine Learning have had a visible effect on our industry, yet I find myself wondering if these new tools confuse us more than help us.
Swimming in Data
I believe there are three key challenges in the social analytics, insights and measurement industry today:
1. Innovation in Isolation
There are pockets of great work across the globe, no doubt. Like Social Intelligence Lab, Significance Systems, Sqreem and some others. The large companies in the industry seem to talk about how one can identify their logo on an Instagram post.
If there is one piece of advice I can give – build your own tech. If you think innovation happens only on a presentation slide or in the way you use SPSS, then you’re mistaken. It starts from asking the relevant questions from the engineer who designed the data architecture all the way through to the end of the insights process. The ultimate litmus test is still the action that the marketing team took as a result of the insight. Collaboration is key in this process, and no, there is no out of the box social listening tool out there (as of writing) that truly caters to a serious analyst. Push button insights are far from reality.
2. Industry Standards
This means if you’re a client like me, you need to make sure to spend an adequate amount of time training your agencies and establishing standards. Standards are crucial for making sure social media research spreads across the organisation and doesn’t become a function for the intern in the marketing department.
3. Data Pukes
Many corporate jobs have become increasingly analytical and tech infused over the years. And marketers love data and dashboards, even if they don’t provide much value.
Avinash Kaushik has written a fantastic post about Five Strategies for Slaying the Data Puking Dragon. It does a great job of explaining the world of data pukes.
Imagine you go into a room filled with 100 people and ask them to raise their hands if they have filled in a structured survey or participated in a focus group recently, you’ll likely get a few hands. Then you ask the same room, if they posted anything on social media recently – you’ll have a very different outcome. Social is a more natural venue for us to not just talk about our lives, but also the products and services we care about. We’re swimming in data, but often lack good analysts, methods and tools to unlock the real value.
By the way, by no means do I claim to have all the answers. This post is actually about an idea I wanted to share, so without focusing too much on the “Siim is too old, who cares”, let’s talk about what can be done.
Insight Led Marketing
One of the reasons why I’m excited about this industry is that I enjoy the work (yep, still!), but I also fundamentally believe social media research deserves a seat at the management table. If done right, it provides valuable insight into people’s needs and desires. It can also be faster and more cost efficient versus alternative market research methods. Ultimately it helps business to stay humble by understanding user criticism and develop products that matter. I believe every marketing effort should be insight led.
Here is a view at the social research process from start till insight:
We could talk about solving several parts of the process, but we’ve been tinkering with the biggest vacuum first. Manual data coding. If you don’t know what that is, you can start with this video. I believe it’s a fundamental skill for any serious social media analyst today.
This idea really started many years ago, but got more real once we started interviewing professional social media analysts earlier this year. Several interviewees analyse social media content 5 days a week, some of them manually coding more than 10K social conversations a year serving some of the biggest brands and agencies.
Some highlights from our interviews:
- Learning to become a social media analyst is difficult. There are few tools and credible resources.
- Unstructured data analysis is often not a linear process. Researchers often go back to their previously
- It is easy to make errors, especially when many coders and languages are involved.
- Spreadsheets are still preferred. Most analysts get quickly annoyed with slow-loading cookie cutter charts.
- Social media listening tools are all designed for super busy marketers and not their analysts. Ironically, the people who actually have the time to use them.
After many in-person and online interviews, and several prototypes – we’re excited to share the latest prototype.
WorkHero democratises unstructured data analysis for market research purposes and offers an extra income stream for amateur analysts using their mobile.
We believe this approach could dramatically improve several key factors that are important for analysts – speed, structure, overall quality of data and methodological rigour. Enabling analysts with better tools to craft better insights.
While we further develop WorkHero, we look forward to hearing your thoughts, concerns and comments about this idea, especially as an analyst or brand marketer. What are your frustrations with social media data? How do you get the best value for your brand?