May 7th 2019
It’s so easy to do brand-related research with our tools at CivicFeed. I like to generate word clouds because they have a lot of visual punch and a lot of relevant information. And the information is conveyed effectively (not all charts would get this thumbs-up from people like me).
Just looking at linkedin, I ran across a former student, co-author, co-patentor, great colleague. Probably the best cpu hardware expert I knew. Then the MBA at Wharton. And now someone with a great tech leadership job that would impress just about anyone.
Of course, his company was in the press a lot, almost a year ago, with some issues. Thought I’d check how that company’s being talked about today, with some statistical/big data-based intelligence, not just traditional market research anecdotal evidence (we do poll social media often, but in most of my searches I limit to news outlets).
Well, happy to say the associations are looking good. Of 534 phrases containing the company’s name (minimum of 4 document nontrivial impressions in the past month), not one mention of the prior issue. Not one.
That has to be satisfying. What really satisfies me is that it took less time to generate that image than it took to write this sentence.
The underlying stream-oriented NLP term-extraction and crowd-sourced AI-informed IR does take some time to do its work each night. And yeah, I honed those skills on an intelligence agency project post 9/11 for five years, so not everyone can build the back end as quickly as we did. Something about n-grams for n bigger than 3. Good luck to people who try. But once the data are preprocessed, yeah, queries like this take about about as much time as it takes to trip a camera shutter. Speaking of which, let’s check Nikon and Canon. That just took about two seconds to check. Interesting, yes?
Monitor your PR impact across millions of stories. It's free to start.