Our Crowd consistently and significantly outperforms algorithm-reliant methodologies.
Famous algorithm-based tool
The BrandsEye Crowd
- AboutSearch engine
Here’s a quick look at how we mine opinions
First we drillWe continuously seek, gather and store mentions based on predefined criteria across online, social, print and broadcast media.
Then the algorithms sift and selectOur sophisticated machine learning algorithms evaluate each mention to ensure relevance and the discovery of useful metadata.
Then the Crowd verifies sentiment and topicsA statistically random selection of opinions, enables a 99% confidence level with a 1.75% error margin, is fed to the Crowd. The Crowd will accurately verify and rate both the sentiments contained in the opinions and the topics driving that sentiment.
Finally we visualise the dataData is brought to life in a highly customisable interface with a wide range of analysis tools giving you accurate, reliable data.
Why crowd verified data is important
Our integrated Crowd verified approach gives clients unrivalled data accuracy of up to up to 97%. Find out how we do it.
Only relevant data is kept
Our Crowd checks to see if the mention is relevant to you or your research topic. For organisations and products with generic names, or international alternatives, this is the first hurdle towards accurate data.
Nuanced sentiment is understood
Algorithm-based systems often get sentiment wrong because of the complex and emotive nature of social conversations. Our Crowd examines the nuance and meaning behind these opinions, ensuring data is correctly categorised.
Topics are inferred and discovered
The Crowd organically uncovers the themes and key topics driving people’s opinions, even when they don’t spell them out in as many words.
Rock-solid features come standard
Trained, skilled contributorsAll Crowd contributors are trained and continuously monitored by the BrandsEye proprietary training system and in-house “Crowd Wranglers”.
Multiple contributorsMultiple contributors evaluate each data point to eliminate user error.
Detailed evaluationEach data point sent to the Crowd is evaluated for relevance and sentiment, as well as conversation topics.
Versatile and international capacityThe Crowd is readily scalable in proportion to data density. 12 Global language Crowds are currently operational with more planned.
Fast and cost effective evaluationsEach data point is evaluated in just a few minutes at a considerably lower cost than in-house data cleaning.
Local crowds and languagesLocally relevant crowds provide regional understanding of context when evaluating opinion data.