Our Solution-Based Approach to Social Media Intelligence

Don't forget to share this post!

blog-post-example-media

A couple of weeks ago, we launched our brand new website. More than a complete redesign, the new linkfluence.com articulates the solution-based strategy we’ve developed with the team and with our top customers after I took over as CEO a few months ago. I wanted to come back on why we made this change and what this approach means for our customers.

The social listening industry made a broad promise but didn’t articulate the concrete actionable use cases

As a relative newcomer in the social listening space, I’ve been surprised by the way the social listening industry presented itself. The broad promise of knowing your customers and your market better is noble, bold and ambitious. I subscribe to it. But beyond this, much of the talk has been about the what (social content / social data) and the how (big data / AI) rather than detailing the why beyond that initial promise.

By discussing with many brands who were using various social listening tools, I heard that a piece was missing: the actionable use cases of social listening. Sure, a lot of marketers are now using social listening. They collect data, they have metrics they can analyze or report. In many ways, they have a much better understanding. But many of them also asked: “so what?

What are the concrete actions they can make with social data?

Up to the point where analysts such as Forrester’s Jessica Liu and Arleen Chien observe that

in our 2016 Forrester Wave ™ evaluation of social listening platforms, we acknowledged that the technology had great potential, but its heyday was still to come. We’re still holding our breath in 2018.

And in that same article, Barry Levine notes that

beyond detecting anomalies, the platforms don’t generally offer many options for acting on the data.

For many brands, the social listening industry has struggled to show what exactly the use cases were.

The change in data access drove the need to revisit unrealistic expectations

In the early days of social listening, regulations were few and far between. As a result, a number of players set the wrong expectations. Social listening was about becoming omniscient about each and every consumer. From working with a few hundreds or thousands of sampled data from market surveys, brands were promised a comprehensive data set of everything.

It shouldn’t happen. It didn’t happen. And more importantly, it needn’t happen… for social listening to bring tremendous value. Social media intelligence isn’t about "big brothering" your customers.

Now, with the Instagram API changes that followed Facebook’s Cambridge Analytica scandal as well as the GDPR implementation last year, there’s been a lot of changes in what social data could be accessed by social listening platforms.

The big brother promise of universally knowing all of your customers without any limitation is gone.

Instead, there’s a much more viable, sustainable and regulated access to data that is still much richer than what the industry can actually leverage.

That includes public content by influencers that is specifically meant to be widespread, businesses of all sizes or average consumers vocal and passionate enough to exercise what is now called micro-influence, as well as posts from John Does retrieved on an anonymous basis.

Even after social media platforms limited access to social data and shocked the whole industry, social media remains the biggest focus group for valuable consumer insights. Simply put, we don’t need access to every single post to detect consumer preferences and emerging trends.

The important point is that social listening is here to stay. Not to broadly spy on people with a vague hope of digging gold but to enable certain specific use cases that the social listening industry needs to define and that drive marketers to take concrete and valuable actions.

AI is a means to an end

Another interesting way to observe the disconnect between the promise and the reality of social listening is to look at how technology - particularly AI - is used.

AI is every tech company’s favorite buzzword. Compared to many product categories that are just introducing AI or others that are barely pretending to use AI, there’s some really cool use of AI in social listening.

AI learns by crunching data. With social listening, data is abundant, giving machine learning and deep learning systems the fuel they need. I would even venture to say that within the MarTech space, social listening is probably one of the areas where AI is used the most.

Social listening platforms can detect logos or people in image, scenes such as nightclub parties or car races, any of the basic seven human emotions, underlying topics of conversations, etc…. And they can do so at scale and some reasonable statistical accuracy which has a lot of value when you deal with hundreds of millions of pieces of new content on a daily basis. More details in this overview article from our AI team.

But most of the AI development has been driven by engineers based on what they could do, rather than what the market needed.

At Linkfluence, AI is not an end, but a means to answer our clients' questions.

Let me explain.

Linkfluence’s hybrid model gives us the unique opportunity to operate our own technology for our clients and identify the most valuable and actionable use cases of social data

Working with global brands has given us an edge. These brands see potential in social data way beyond digital marketing. To them, social listening is more than just answering to every tweet that mentions them. It opens a window into consumer preferences and emerging trends.

As any software vendor, we can learn by observing how they use our platform. But unlike other players in our market, by providing them with research services delivered from our team of 100+ local analysts and data scientist, we’ve seen that social data could be much more impacting and - more importantly - actionable.

Having worked on hundreds of briefs from various brands over the past years, we’ve been able to explore advanced usages of social data to answer key questions for our clients. Whether they had to make decisions around their marketing strategy, their advertising budgets, their choices of influencers, their media mix, their brand positioning or product innovation, social data is leveraged as the voice of the consumer.

Every time, we bring a research team to work on a specific client problem and we assess whether social data can help and how it can help. If it does, we then produce a research report using Radarly to answer our clients’ questions.

By design, this approach drives actionable results. Here are some concrete examples:

  • Analyze consumption habits
    We’ve helped a major fashion brand discover they were a go-to gift Japanese students spent their first salary on to reward themselves - something no traditional market research approach could have uncovered and that can be amplified through specific advertising campaigns and messaging.
  • Identify target tribes
    We’ve enabled a wine & spirit group to identify seven communities within their core target audience of bartenders and identify their needs - something they’re now able to use to create specific content, training programs and specific messaging for.

  • Detect emerging trends
    For one of the major food companies in the world, we’ve identified the popular ingredients of tomorrow - unlocking new product development potential and optimizing promotional budget on the products containing these ingredients.

This organizational model makes us much closer to the market needs. It allows us to identify patterns in client needs: repeatable, high-value questions, and drive our technology roadmap accordingly.

Introducing 6 social intelligence solutions that deliver actionable and valuable insights to global brands

So to capitalize on this unique model, we’ve decided to take it a step further and create the first offer of hybrid social intelligence solutions combining our social listening technology and the know-how of our 100+ social data analysts and data scientists.

This strategic change is made visible by our new website which articulates our offer around an initial set of 6 solutions:

And we’re not just doing this for marketing purposes. We’re also changing the way we work with our clients:

  • Of course, we’re still providing our SaaS platform which includes Radarly, Linkfluence Search and Linkfluence API as a standalone and self-service offer. Our goal is to empower our clients to make the most of social data so our platform comes with set-up, training and professional services that ensure rapid adoption and increased maturity with social data.

  • We’re also continuing to provide ad hoc research based on specific client briefs. This is also important for our clients that have specific needs and as a way to keep innovating and prepare future solutions to standardize.

  • However, we’re now also investing some of our resources to develop these solutions over time from initial PoC’s to further evolutions. We spend time defining and refining the best methodologies for these solutions and training our team accordingly. Or by structuring the data specifically for them across various markets and industries.

The benefits of this solution-based approach for our clients are multiple as our social intelligence solutions are:

  1. Actionable: these solutions are designed with a clear purpose. Whether it’s optimizing a mix of influencers to work with, allocating a marketing budget, designing a new campaign or creating specific messaging, Linkfluence’s social intelligence solutions correspond to specific actions across the marketing cycle.

  2. Predictable: while it’s tempting to think social data will answer any question, our social intelligence solutions set clear expectations on what they address and how they address it.

  3. Turnkey: from acquiring the right type and amount of social data (and yes, they work with current social network API’s), to leveraging the right type of AI models and combining the right type of human expertise, everything is designed upfront and included. This eliminates the risk of slower adoption or missing a piece of the equation and not getting ROI from social listening technology.

Social intelligence has been a promise of social listening for a long time. By combining AI technology and human expertise in a hybrid model, Linkfluence has had a unique first-hand experience of global brands’ needs with social data. This enabled us to identify the most valuable use cases of social listening and build a solution-based approach to address them.

And we felt it deserved a new website.

What do you think?

Don't forget to share this post!