FAQs

Citibeats is a unique platform. Public and private organizations should know the real needs of citizens. Citibeats gives you insights into context, trends, citizens' concerns, and much more. With Citibeats, your organization will detect changes in society earlier than any other traditional survey method, giving you the opportunity to proactively impact society in a positive way. As we like to say, to shape a better future, you need to start predicting the present.

With Citibeats, you don't just listen. You will understand what people are talking about on different social topics. Citibeats transfers learning for most of our deployed algorithms but also in multilingual intent detectors. Social listening tools work with brands and assess the sentiments of text. Citibeats, on the other hand, gives structure to texts based on the context in a much more general setting.

Citibeats is easy to use. The user doesn’t need a technical or data analyst background. Private and public organizations benefit from Citibeats by identifying and understanding pressing concerns and needs of citizens to make better-informed decisions. 

No, Citibeats' Platform does not have a minimum or a maximum number of users. Build, explore, and analyze with as many users as you wish. 

Yes, Citibeats' Platform does not have limitations about the number of users. However, additional users come with an extra cost. Be in touch with our team so they can keep you informed about the pricing.

Citibeats is a prepaid online subscription: All services are paid for upfront, either by credit card or by manual invoices. You can contact our sales team here on our pricing page. After any order or payment, an invoice / receipt detailing the transaction, and showing it’s been paid, is sent to the email address connected to your Citibeats account, or any other address you see fit.

If you have any questions regarding your invoice, you can contact our customer support team here.

Using filters, you can create a dashboard that best represents your needs. Moreover, you can also create reports that contain your customized text and content. When you get your license, you will also receive exclusive access to training.

Anyone on your team can easily access and use the Citibeats platform. You don’t need a programming, coding, or analytics background.

Once you have confirmed your account and paid, you’ll be able to use Citibeats in 1 working day.

Yes. You will have access to many materials (guides, video tutorials, reports) to show you how to make the most out of Citibeats.

Citibeats collects open-source opinions posted online: on social media (Twitter, Facebook, Instagram, Reddit, Youtube), from 1,000,000+ unique digital sources (forums, blogs, websites, comments on digital newspapers), and many other on-demand sources. All data has been previously anonymized.

When you start using any of Citibeats' monitors, you have access to millions of data points that have already been (and will continue to be) collected.

Social understanding platforms such as Citibeats serve a people-oriented purpose, aimed at positively impacting society by interpreting real-time societal changes, public opinions, and conversations to deduce macro- and micro-narratives. 

Social understanding tools have been developed to help you understand - and not just listen to - citizens’ real needs and opinions, giving you insights to create better policies.

This blog post illustrates the differences between social understanding and social listening. 

With the term "data," we refer to any citizen’s opinion or comment collected and analyzed by Citibeats' platform.

“Data” can be a tweet, a review on Google, an opinion in a forum or in a digital newspaper, etc.

Citibeats has access to more than a million data sources, which we divide into different categories:

• Public sources: tweets, reviews, opinions in forums and in Facebook’s public groups, public surveys, etc.
• Private sources: all the data (in text form) coming directly from our clients.
• On-demand data: adhoc surveys specifically created to reach areas where there is no Internet connection, collect opinions of citizens who otherwise would not be included in the decision-making process, and thus deliver models that are representative of the whole population. 

Citibeats’ AI platform is an example of how AI should be ethical and responsible from its very conception and design. Citibeats was one of the first four tech companies to implement the ethical self-assessment tool promoted by fAIr LAC and IDB Lab.  Our ethical technology is based on some fundamental principles: 

• Protection of users’ privacy. We only use anonymized data to protect the privacy of the users.
• Bias reduction. We have developed our own technology to draw balanced conclusions, representing the entire society while reducing age, race, or gender bias. 
• Use cases with a positive social impact. We only participate in projects with a positive impact on society, directly impacting people's well-being.

By analyzing aggregated, anonymized data, Citibeats’ artificial intelligence technology is ethical in nature, seeks to mitigate bias, and is utilized solely for positive societal impact. There are two kinds of bias that correspond to two different phases: the bias input and output (in the model).

The data collection thanks to which we gather the opinions of the global population is a process that is inherently biased (bias input). To “polish” and refine the models, we calibrate the results, dealing with both genders, and then we aggregate the results (as explained here) to provide a representative analysis of the whole population.

Citibeats uses the following official data sources: Twitter, web content, Facebook public pages, "On Demand Data".

It is also possible to integrate third-party data sources as .csv or via API connector.

Citibeats can rely on +1 million digital data sources.

When we refer to a “document” or “opinion” (expressed by a human being) we refer to any text-based data.

document = opinion = data

When we talk about the number of data we analyze or collect or a certain monitor offers, we are talking about the number of opinions expressed by citizens (that we can also call “documents”).

In our analysis, we can apply a segment filter (male/female/organizations), but we cannot detect if the document has been created by a bot or a fake account. In case a Twitter account was a bot, it should be stated in the description of the account itself. In any case, it is estimated that bots represent less than 5% of Twitter accounts, so they do not represent an issue in terms of “noise” for our data collection and analysis.

When we talk about “voice”, we are referring to the capacity of every single individual to express his/her ideas.

When we refer to “voices” we are NOT talking about opinions.

voice ≠ opinion

This is because, for example, one person clearly has one voice, but s/he might have different opinions. Following that, it is possible to have 1M documents but then 800k voices (because some of those voices expressed more than just one opinion).