Topsy Takes Social Analysis to the Streets

Leading social analytics company upgrades local analysis capabilities, gives brands ground-level view of local trends, sentiment and influencers

SAN FRANCISCO – Topsy, the real-time social analytics company, today announced major updates to its geo-inference model that will give brands the unique ability to identify and examine location-specific trends in real-time.  While localized social analysis has proven challenging for brands in the past, since less than two percent of tweets include latitude and longitude data, Topsy’s new geo-inference model adds location names from users’ profiles, tweet text, language, use of local websites and other signals to enable Topsy to infer location for a very large percentage of tweets, with over 90 percent statistical confidence.

Topsy can now identify the origin of over 95 percent of tweets at the country level, 50 percent of tweets at the state level, 30 percent of tweets at the county level, and over 25 percent of tweets at the city level.

Topsy’s new capabilities are a boon for brands who want to increase the impact of their local marketing.

“Marketing at the local level can often be a huge pain point for our clients because it’s impossible to have eyes and ears in every market when executing scaled, nimble marketing initiatives,” said David Shaw, associate director of social strategy at digital marketing agency Resource. “Topsy’s geo-inference capability has been a welcome addition to our social listening center, iCitizen Live, giving us the ability to uncover and understand local conversations.  Understanding social conversation down all the way down to the city level allows us to tailor our clients’ marketing messages and tactics to fit the local market.”

Using Topsy’s enhanced geographic social analysis, brands can now:

1. Create detailed local market profiles. Many large national brands have local presences across the country, but often manage social centrally from the national office.  Using Topsy, brands can now create detailed local market profiles that highlight differences in local customer’s interests, sentiment and views of the brand, allowing for much more targeted and relevant outreach.

2. Use local sentiment data in product launches. Companies can also use Topsy to spot where their brand or product is popular, as well as see which regions need improvement, prior to a launch. For example, an athletic retailer launching a new shoe can now measure buzz by location and tailor its outreach and media strategy to boost lagging markets, or a movie studio can plan where to do screenings based on which areas have the most positive conversation for upcoming films.

3. Launch highly relevant, real-time local campaigns. Brands can combine Topsy’s alerting capabilities with location-based information to jump on local events or trends in real time.  Setting up alerts for key geographies allows brands to get notified if conversation or sentiment changes, or as anything starts trending locally, before it even causes a blip at a national level.  For example, if a national coffee chain were alerted early to a crisis situation in one of its local markets, it might take the opportunity to offer free coffee to the personnel working overtime in the area.

4. Optimize Promoted Tweets for the local market.  Brands are already targeting Promoted Tweets to specific geographic regions; now, they can also use Topsy to optimize their Promoted Tweet content in real-time around local trends or issues, making the content much more relevant.

5. Get a jump on the local competition. Brands can also now identify areas where a competitor’s buzz or sentiment is weak in particular regions. For example, a cellular provider could use Topsy to uncover negative sentiment for their competition in the Chicago area, and immediately launch a regional campaign to capitalize on their competitors’ lack of customer satisfaction.

Brands can also discover where competitors are testing new strategies based on regional social chatter. If a clothing retailer is testing out a new rewards program in a specific region, a competing retailer can monitor the buzz and sentiment around the new program. They may want to start their own conversation with local consumers to gain additional competitive insights, or target that same market for their own testing.

“Twitter is used throughout the world as a way to converse and share information, and analysis of those conversations has proved immensely valuable to brands and agencies,” said Rishab Aiyer Ghosh, co-founder and chief scientist at Topsy. “Our updated geo-inference model continues to push the envelope with what is possible with enhanced geographic information.  Now you can monitor whichever portion of the world’s largest conversation is most relevant to you.”

Behind the Science of Topsy’s Geo-Inference Model

The most accurate geo-location data points associated with tweets are latitude and longitude, which are communicated when users opt-in to sharing their location from mobile devices equipped with GPS. However, only about one to two percent of all tweets contain latitude and longitude data, making it difficult to perform wide-scale geo-location analysis of Twitter data.

To address this issue, Topsy’s geo-inference model uses latitude and longitude coordinates as “anchor” variables and adds other variables such as location names from users’ profiles, tweet text, language, use of local websites, use of hashtags for local events and other signals to infer with high accuracy from where a user is communicating. Topsy’s geo-inference model uses machine learning to automatically discover which signals are accurate predictors of location. Topsy analyzes over 450 million tweets every 24 hours, and geo-encodes each tweet in real time. This geo-encoded data is the basis for geographic insights displayed within the Topsy Pro platform, and provides users with deep, rich geographic analysis that can be performed for multiple years in real-time.

About Topsy

Topsy is a social analytics company that gives you instant answers to critical business questions through real-time analysis of public conversations.  Operating the world’s only index of the public social web, Topsy is used by some of the largest global marketing, news, entertainment, and financial organizations to get real-time insights into a wide variety of business questions: everything from spotting breaking news and identifying customer satisfaction issues, to gauging response to TV programming and understanding political sentiment.  Many of the world’s largest brands also use Topsy as their day-to-day social dashboard to discover and track the impact of trending topics and predict future events.  Headquartered in San Francisco, Topsy is backed by BlueRun Ventures, Ignition Partners, Founders Fund and Scott Banister. For more information visit: