Sila: more than social listening. It's social understanding.

Sila (‘connection’ in Arabic) helps D/A solve complex consumer insight challenges through a mix of natural language processing, machine learning and human smarts.


Arabic consumer insight, at scale.

‘Sila’ (Arabic for ‘connection’) is our proprietary AI-enabled consumer intelligence platform for insights into Arab-world consumers. It organises mega-data into workable structures from which insights can be called upon and analysed to inform audience-first strategy with features like:

  • Social data;
  • Dialect models;
  • Sentiment in dialects;
  • Personality analysis;
  • Topics/Predictive behaviour models

Proprietary data system

Sila uses an unbiased data-collection methodology which provides us with un-influenced, ambient feedback from multiple online sources.

Our machine learning operates natively in Arabic dialect so there is no dilution of vocabulary or sentiment through translation, giving clients the highest possible accuracy of Arabic insight.

Natural language processing enriches consumer data even further, creating addressable audiences. Identifying personality traits and consumption preferences allows us to predict likely future behaviour rather than simply assessing the past.

Sila's Proprietary System

Arabic-first AI consumer intelligence

Sila has numerous models that make up its AI core, most based on machine learning and natural language processing. These models give insight into the Arabic-language consumer better than most. Our models run across hundreds of millions of social data points and digital data to process:

  • Arabic dialect – understand 2 major dialects across the MENA region;
  • Sentiment– accurate (94.8% for Arabic) measurements of feelings across MENA;
  • Emotion – accurate (89.6%) emotion readings for social media posts across MENA;
  • Demographic understanding – age, gender, location models give accuracy to insight;
  • Audience segmentation – automatically segment audiences based on preferences, personality or shared interests.
  • Affinities – understand who shares interests in brands, competitors, influencers and places;
  • Personality – using the ‘Big 5’ model we’re able to determine the characteristics of audiences;
  • Preferences and drivers – what do consumers like? what might they use?;
  • Volume normalisation – frustrated at different measurements of social media recipes? We have a model to normalise these giving unmatched insight to trends;
  • Share of Search – As part of our new measurement suite, we measure Arabic share of search as a measure of campaign uplift.

How we use Sila to unlock insight

Our reporting platform (called ‘Sila Discovery’) allows for fast data visualisation while our data scientists manage the integrity of our data. Our researchers then can examine, filter and visualise the data to delve deep into finding insight that can make strategy and execution clearer.

We then use Sila by:

Get in touch with us and see how Sila can help you.

Get a demo of Sila’s power.

This field is for validation purposes and should be left unchanged.