Like you and your brand, we at Synthesio are unsure about what to say and when to say it. Together, we’ve all come to understand the difference between essential and non-essential work. Running ads and pushing products at a time like this feels tone-deaf. So why data quality?

To steal a word, if there is an ‘essential’ task that marketing and consumer insight teams can perform right now, it’s this:  listening and trying to understand one’s consumers. In today’s new world, to stay relevant, you must be thoughtful. To stay thoughtful, you, of course, need to listen to what people are saying, doing, and thinking.  And to listen effectively, you must filter out the needless noise.  You need to work with quality, representative data. 

And so, while we tread lightly on pushing products of our own, we did feel it was worth sharing a new enhancement, one that should make it easier for you to home in on how your consumers are feeling and responding to this crisis.

Noise Reducer: Improve data quality. Make better decisions.

Our newest enhancement is called ‘Noise Reducer.’ As its name suggests, Noise Reducer removes irrelevant mentions from your data-set. Noisy mentions – those that merely include keywords as opposed to capturing genuine consumer expression – expose brands to risk and wasted time.  Noise Reducer removes this proverbial noise, directing brands to genuine consumer expressions, and enabling better, more confident decision-making.  If social intelligence offers the ability to be quick, Noise Reducer adds the ability to be sure and quick at the same time.

As we’ve discussed above, during a crisis like this one, it’s critical that organizations spend more time listening to consumers.  Noisy mentions distract from this goal.  Looking at Twitter alone, it has been estimated that 15 percent of active Users are in fact bots. But not only do noisy mentions distract, they also carry a substantial economic cost. IBM estimates that poor quality data cost the US economy $3.1T in 2016. Hard to fathom. As social media usage continues to rise – Facebook lays claim to 1.66 Billion Active Users – data quality becomes essential to organizations looking to unearth social listening insights. It’s under this backdrop that, again, Noise Reducer seeks to save you time, reduce cost and risk, and guide you to confident, consumer-centric decisions. 

Consumer Insights teams need quality social data so they can fully understand the consumer’s viewpoints…Social data can be tricky because it may include noise that may not be helpful in answering key business questions…Noise Reducer will help our analysts be more efficient and help us to successfully answer the client’s business questions.

Kathleen Volker, Senior Account Manager, Social Intelligence Analytics at Ipsos

Aside from traditional SPAM, Noise Reducer automatically removes mentions related to job postings, online games, advertisements, hate speech, vulgarity, and more. You can remove these categories of mentions before and after digesting your initial data-set. Powering Noise Reducer is Google’s BERT technology. BERT, different from traditional natural language techniques, looks at the full context of a social mention. Our use of Google BERT ensures that Sythensio is classifying mentions as accurately as possible. 

Noise reducer improves data quality by automatically removing irrelevant mentions.

                 Automatically remove noisy categories before and after data ingestion. 

Review, edit and bucket noisy mentions across your data stream.

Again, like you, this crisis has refocused us on our purpose. Our purpose is to help you understand your consumers better. We know what we do is not the end-all-be-all, but we hope and believe it can help. Thank you!