Where does BrandGraph Data come from?
BrandGraph collects and analyzes data from all the major social platforms, however, the majority of our data comes from Instagram. IZEA has focused our discovery efforts on Instagram as a reflection of the demand for Instagram content over the past few years. IZEA’s Instagram API allows us to pull information directly from Instagram into BrandGraph.
How do we pull data into BrandGraph?
We use the same proprietary systems that we use to search and surface content in IZEAx, tracking over 1 billion pieces of content, the majority of which comes from Instagram. We then scan captions for the common FTC disclosure to identify if posts are sponsored or organic. We note any Brand mentions to surface those pieces of content to the relevant Brands. In addition, BrandGraph will use a Brand’s associated social media handles, website, and their hashtags to identify content that mentions that Brand.
Enhanced Categorizing Measures
BrandGraph uses Phrase Matches and Proximity Identifiers related to the Brand to accurately identify content specific to that Brand, even if they are not mentioned or tagged.
Phrase Matches
Phrase matches are terms that are specific to a Brand but do not always have to include the Brand's trade name. The phrase matches help us accurately identify content specific to the Brand. McNuggets is an example of a phrase match for McDonalds. If someone posts “I was really craving some spicy chicken McNuggets”, BrandGraph will connect that content to McDonalds even though “McDonalds” was never mentioned. Another example for a sports team would be like “watching the Packers game” for Green Bay Packers.
Proximity Identifiers
Proximity Identifiers enable BrandGraph to disambiguate content that contains mentions of a Brands’ trade names. Many well-known Brands utilize trade names that are also words commonly used in everyday speech. For example, the word “apple” can be used in speaking about a technology company, a fruit, a city (The Big Apple), or even a person (Fiona Apple). The mention of the word alone does not indicate that an individual is speaking about the Apple Brand. Proximity Identifiers build on BrandGraph’s existing matching algorithms to further increase its ability to accurately identify Brands by analyzing surrounding text.
7-Day lag time
BrandGraph is pulling in data consistently every 10 seconds, but it takes the software about 7 days to fully filter, sort and categorize that data that is presented in Trends. The Trending Now data is updated regularly to capture an accurate picture of what is happening in real-time.
