Who we work for
A message to convey.
A brand to promote.
A policy to implement.
A point of view to popularise.
If people are talking about it online, we can measure what they say, how it changes, and who is shaping it.
Ad agencies
“Half the money I spend on advertising is wasted; the trouble is, I don’t know which half.” Today, i‑sieve could tell John Wanamaker.
Our sentiment analysis finds out what people say about a given campaign: what they like, what they don’t, and how public perception of the product has shifted because of it. Experience has taught us to expect the unexpected: the public often fixes on an aspect that seems peripheral to the campaign itself, and it’s exactly this intelligence that lets agencies deliver future messages in terms that make sense to the target market.
Media organisations
When audiences are measured in millions, keeping track of their reaction is daunting. Surveys are expensive and only ever answer the question posed.
i‑sieve measures public reaction close to real time. Once trained, we can deliver sentiment results daily. We pick up every relevant comment, positive, negative, neutral or unexpected, identifying key words, phrases and sentiment drivers. Because the system is automated it scales to large volumes of data. And it’s smart: we recognise opinion shapers versus followers, mainstream versus extreme, the considered versus the ill‑informed.
Brand owners
Do people see you as a lemon or a peach? Do they use your brand name as a verb? How do they refer to what you sell when they don’t use your name at all?
These are the questions i‑sieve answers. We can do it as a snapshot, but the real value comes from tracking perceptions of your brand over time. Once our system is trained to recognise comment about your brand, long‑term monitoring of sentiment is easy to maintain at low cost.
Standards bodies
How many people actually go to the bother of complaining? And how typical are those complaints?
Journalists work to standards of accuracy, films and games carry age ratings, restaurants have hygiene rules. Compliance is usually monitored by responding to complaints received. i‑sieve monitors public comment instead: a tweet about a rat in a restaurant, a blog post questioning a game’s rating. Here the system is trained to find comments about breaches of the standard rather than a particular brand.
Policy makers
What do people really think about climate change? Which measures of the economy do people most care about? How representative are the loudest voices?
Rather than polling a thousand individuals, we analyse the sentiment expressed in hundreds of thousands of blogs, micro‑blogs, videos and comments. We track changes in opinion and help pinpoint the underlying cause of any shift or hardening. Results are statistically valid and you receive the actual data. Long‑term monitoring of an issue is particularly cost effective.