What’s new at Factmata ?
Thanks for signing up for our newsletter! This is the first routine quarterly update we’ll be sharing from now on to let you know what we’re up to and how you can get involved!
First, some good news. Over the summer Factmata has won several awards across the globe:
In June, we won the 2021 CogX award for Best AI Product in Marketing and AdTech. CogX is the largest AI festival and conference in the UK and attracts major players from industry, governments around the world and celebrities. This year Robert Downey Jr and Matthew McConaughey attended.
In September, we won the 2021 MSDUK Innovation Challenge in the Advanced Digital Solutions category, sponsored by EY, GSK and Accenture.
In October, we got awarded Gold at the 2021 Globee Communications Excellence Awards for Artificial Intelligence Startup of the Year.
This month, we’ve won the £60,000 Kaust Challenge Grand Prize for Shaping the Future of Media, sponsored by Google, Microsoft and Facebook.
These awards are recognition that the last few years and almost $4m in funding has not been wasted!
We’ve also made progress on our product, having run multiple trials over the summer with clients of all sizes from small agencies to international brands. Here’s what one of our trial clients had to say:
“The team created a bespoke dashboard that allowed us to understand targeted conversations happening in real time as well as a broader social narrative on the subject. The knowledge and insight gained was incredibly useful in driving both our strategic and creative approach to the brief. The client fed back that they felt we understood their business and the barriers to communication with clarity, and we know that Factmata was a significant part of this. We’ve now secured the business, which is a big win for us both in terms of scope of work and breadth of clients serviced.”
We think our Narrative Monitoring product is now ready for releasing into the wild. Over the next few weeks we’ll be reaching out to a broader range of potential clients to get them to sign up and give it a try.
We’ll also be reaching out to journalists as well as some Not-for-Profit organisations to give them access to our ‘Threat Score’ product for free.
Threat Score is our freemium offering giving any company or organisation a snapshot of the quality of the conversations surrounding them online. Think of it like a Credit Score, but for a company’s online engagement. If there’s a lot of fake news or racism surrounding a brand, they’ll have a high Threat Score. We think this is a great way to get people used to the concepts of narrative monitoring but also raising awareness of harmful content earlier, so more organisations can get involved in tackling it, sooner.
We’ve created a quick explainer video to help potential clients understand how our product works, you can see it here.
New machine learning models
After four years of working on the problem of automatic identification of fake news and other forms of harmful content, one thing we’re sure of is that you can’t make good, automated decisions for all types of content in all circumstances. Context is critical, as is applying more than one kind of model to the content. That’s why we’ve now developed a total of 19 different models, each one capable of analysing a piece of text through a different but necessary lens, so we’re able to identify the harmful content better and provide much more context to humans using the results of our models for decision making.
Here’s the most recent additions to our suite:
Fake News – a model aimed at detecting content likely to contain misleading, low-quality content written to sound authentic.
Satire – A really important model to have in addition to Fake News to help prevent too many false-positive fake news flags.
Sarcasm – Another really important model to use in conjunction with the fake news model.
Subjectivity – Subtly different to our fake news model, this is aimed at identifying highly opinionated content, again, really useful in adding context to human decisions.
If machine learning models excite you (good for you) but also, we’ve just released our latest white paper on using ‘Stance’ instead of sentiment models for more accurate results.
Factmata as a brand is well known in certain circles and industries however, when it comes to the realities of building a global, hyperscaling business we need to be lazer-focused on our marketing and PR. However, our Narrative Monitoring product is not being promoted to the same kinds of audiences as our content scoring APIs. This makes our Factmata website larger and potentially more confusing for potential customers to navigate and also makes effective Search Engine Optimization (SEO) difficult.
Solution? We’ve set up a new company called ‘Content Score’, with a new domain www.contentscore-ai.com which more clearly states the purpose of our APIs and will allow much more effective targeting for each side of the business. Stay tuned for that website to go live as well as for the Factmata website to get leaner and more focused on our Narrative Monitoring product.
Factmata remains a popular company for people to want to work in due to the importance of our mission and the credibility of our name. We get to pick the best talent to work for us and here’s the newest additions to the team:
Sajad Muosavi – Senior Front End Developer
Ahmed Khaled – Machine Learning Researcher
Cinvi Tang – Strategy Director (Content Score)
Céline Mercier – Business Development (Factmata and Content Score)
Opportunities to get involved
Find your insights on a topic of your choice as our Self-Serve model is live and currently offering a trial for the duration a week !
To have a chance of winning a £100 Amazon gift voucher and a free insights report on a topic of your choice, complete this 3 minutes survey and a winner will be chosen in time for Christmas.