MutualMind gauges sentiment around social conversations but does so with the standard margin of error found in most social analytics products. The nature of human conversation makes it difficult for a machine to categorize sentiment with 100% accuracy. MutualMind gets sentiment right between 50% to 70% which is within acceptable industry averages for this type of product. With additional tuning (which MutualMind provides) the accuracy levels will increase.
By showing increases in engagement and share of voice (defined as the amount of content generated for our advertisers versus the amount of content generated for their competitors over time) we are able to correlate content to incremental sales data provided by our advertisers.