4 reasons why Unstructured recommends YouScan
Learn why Unstructured chooses YouScan over other tools to deliver meaningful market insights for our customers.
We here at Unstructured have worked with numerous social monitoring, listening and intelligence tools. They all have their role to play in an organisation, depending on the organisation’s maturity when it comes to effectively using social media data.
There are three stages of using social media data, and Daniel Gardner from GSK recently did an excellent post on LinkedIn, easily explaining the differences between these 3 core categories of using social media data.
Many organisations believe that they are doing social listening, which is understanding the meaning behind the words people use, but they are really just monitoring metrics and not listening to conversations to extract meaning. Even fewer organisations are really doing social intelligence, which is the ultimate goal and where the biggest value lies. When you start blending these conversations with your internal data, you can easily start identifying the ‘why’ behind the ‘what’.
What exacerbates the problem even further is that most social media tools can’t get you to relevant and accurate data, and that’s why most executives don’t consider the value of social media data. A recent client of ours was tracking five of their brands, yet 70%+ of the data they were capturing was irrelevant to their brands.
This resulted in them being unable to truly understand what was driving customer demands and satisfaction, identify areas of improvement or even find quick wins in regards to operational and product optimisation.
It also didn’t help that 90% of their data sentiment was rated as ‘positive’, when in reality, only about 10% of the conversations were really positive. This prevented them from identifying issues and brand reputational risks.
Even after we restructured their setup, created themes and topics and added sentiment detection, the limitations of the tool they had implemented within their organisation could, at best, only reach sentiment accuracy of around 40%.
So why even bother? Well the reality is that this is not a limitation of the data itself, but rather a limitation of the software they are using. We have extensively tried and tested multiple providers in real world scenarios, and for the past four years we have a firm favourite: YouScan.
YouScan’s technology automatically detects ‘spam’ content, and their advanced setup optimisation, which includes unlimited rule creation, allows users to reach 90%* accuracy for relevance.
Why is relevance important?
A key metric we use at Unstructured is Excess Share of Voice (eSOV). To use this effectively, it’s imperative for an organisation to accurately collect data that is relevant to their brand. This enables you to understand your true levels of brand awareness. Knowing this allows you to optimise your awareness marketing spend to drive the additional conversations you need to achieve the excess share of voice and, ultimately, increase your market share.
2. Sentiment Accuracy
Many organisations are aware of and are trying to use sentiment analytics. The difficulty comes in the sheer volume of work that is required to correct data, especially in large organisations with thousands of conversations.
YouScan’s sentiment accuracy hovers around 60%* before any optimisation.
*Accuracy estimates are based on Unstructured’s four years of experience working with the tool and observing and estimating data accuracy for our clients.
With expert intervention this accuracy can increase to 80% and above.
YouScan also has a data science team who is constantly working on improving their AI and algorithms to become more accurate. We at Unstructured have witnessed these improvements first-hand.
Human conversations are hard to interpret and the understanding of customers and prospects can never be underestimated nor undervalued.
Unstructured has experienced senior decisions being made based on a sample size of one. So having the voice of your customers’ sentiment, and thousands of unsolicited feedback points to accuracies upward of 70%, is extremely powerful to drive data-driven decision-making.
3. Aspects and Subjects
In our opinion, this is YouScan’s second-best feature (we’ll save the best for last!). Using aspect-based sentiment analysis, they automatically categorise the topics of conversations into accurate buckets, so you can easily determine where the shift in sentiment has happened. Plus you can track this over time.
What is a good use case for this?
After a product launch, you can accurately track not only whether people liked your new product, but also what they like about it, how they feel about pricing, whether you have any stock issues in certain stores or if your staff or resellers have insufficient knowledge about
your new product.
Other than just the aspects of conversations, they also detected additional ‘subjects’, like automatically tagging customer service responses and word-of-mouth advertisement of your products, along with plenty others. This provides a skilled analyst with hundreds of ways to segment and filter data to unearth real world insights. With YouScan’s advanced filtering
options, it becomes really easy to uncover these so that you can make accurate decisions, fast.
4. Image Recognition
They’ve won numerous awards for their image recognition technology and we know why. We, at Unstructured, are constantly amazed by the details that they detect in images, some of which are almost invisible to the human eye.
For one of our clients, this additional technology increased monthly conversation volumes by 30%, and in other cases, up to 80%! This means that with only text recognition, you’re missing out on at least one third of insights, where people are using your product ‘subliminally’.
Why should you care?
You get a glimpse of how customers are interacting, using and positioning themselves around your product, without their direct intention of telling you they are doing it. You get to see activities people are performing while wearing your clothes, or which other product types they use in conjunction with your brand.
How can I use this in a practical way?
Let’s say that you manufacture and sell a nutritional supplement. Using YouScan’s image recognition you see that 60% of the time, your product isn’t used how you intended (to mix with water/milk as a shake) it to be used. Instead, consumers are making smoothies, pancakes, acai bowls and anything you can imagine, with your product. Consumers aren’t referencing your brands in their ‘recipes’, but image detection will find this for you. Does your protein powder pair better with strawberries or chocolates? How well does it mix with yoghurt? Image recognition data can tell you.
Whenever we use YouScan’s platform, we get excited and inspired by the possibilities of social media data and the intelligence you can draw from it. The data is powerful, yes, but you are limited by the tools you have to extract that value. And here at Unstructured, we couldn’t be happier that we found YouScan 4 years ago. Not only has it saved us numerous hours, fighting with incompetent tools and correcting incorrect data, but it has unlocked a number of business opportunities for us. Being able to inspire our clients with data insights is something YouScan is incredibly capable of doing.
In fact, one of our global clients has completely scrapped their NPS program and is solely using YouScan’s sentiment and aspects data to track customer experience and satisfaction. They also use it to do research on sustainability topics and to identify gaps in the market from a product usage perspective.
If you are excited about what YouScan can do for your organisation, or if you just want to see a demo of this crazy good image recognition technology and how it can help your brand, get in touch!