What is the business and how did it start?
Back in 2014 I was investing and found it impossible to easily gather data on the market. I knew there were solutions for this, but they were not cheap. They still aren’t. Few people can afford up to £2000 a month to trade or invest their own portfolios. I wanted to change that, and so CityFALCON was born.
I learned more about coding and knew automation and APIs were the keys to bringing down costs and still tackling all the content that was out there. As the company grew, we realised the more people and companies that had access to information, the fairer and more accurately priced the markets would be. It also enabled people like myself to benefit from all the collected content. As the content accumulated, we knew we could also generate much-needed analytics on it, which now includes similar content, sentiment, and NLU extraction.
That’s where we are today: a news and data distributor trying to provide access to data for as many market participants as possible, from the retail user to big enterprises.
How are you different?
That’s a broad question. I think we are walking a path only lightly tread so far. To counteract machine learning’s less-than-perfect accuracy, we’ve taken a semi-manual approach to some of the work, especially in source curation, which has kept costs low but made the accuracy more acceptable in finance. We’re aggregating data from anywhere we think useful, too: news, Twitter, Trustpilot scores, regulatory filings, corporate notices. If you have a suggestion, we’re open to it.
We go beyond the data, though. We encourage individual curation (like/dislike) and apply that across user segments. We provide a lot of analytics and insights. Our proprietary CityFALCON score indicates how relevant a piece of content is to the user. Our Similar Stories feature groups similar content together, across languages, which makes research easier. Our very popular sentiment analysis (via API now, web/app soon) rates the outlook on everything from companies to people, sectors, and locations. NLU extraction gives enterprise clients an opportunity to understand their internal textual content. We plan to do a lot more with the data we glean from filings, too, including on private companies.
All of that comes at a steep discount to the incumbents in the space, in large part thanks to our embrace of automation.
Automation and free-flowing data access are the future, and we’re getting there early. We like to think we’re making the world a fairer place by giving everyone access to that financial content, and we’re helping to explain it with analytics and insights.
What challenges have you overcome?
There are many. We’ve written a blog on the mistake we made in the past. Since I am a first-time and single founder, there were a lot of challenges I inadvertently caused for the team, like less-than-optimal code and misallocating tasks to people who were not the best fit (like a UI person to do both UI and UX). As a product company, the years spent building the product instead of focusing on user acquisition were difficult, too.
We’ve overcome a few times where we nearly ran out of money and we were saved by angels coming in at the last minute. Our salary-to-loan scheme has also helped us mitigate Covid challenges.
Now the challenges to overcome are sales chokepoints. At this point, we have plenty of features to offer, but due diligence and legal reviews from big enterprises can take quite a while – not ideal for SMEs trying to grow revenue. Slow API integration on the client side translates to slow conversion, too. As always, we’re working on solutions here, like Single Sign On (SSO) and automated onboarding to reduce the integration times.
What are your plans for growth?
Now that the development work has produced results on the frontend (API and web/mobile) and our infrastructure is pretty robust, we think the product has become quite attractive. We’re adding more data every day, from filings and new companies and asset classes to new languages on our NLU engine. We are aiming for more than 90 languages, partially driven by our R&D project, so we will have global appeal and insight. We also keep adding more sources and build relationships with publications to source licensed content.
We’ve seen a lot of interest in our NLU extraction from banks to publishers, and sentiment always comes up in conversations about market analysis. Covid cash crunches have also incentivised a lot of people and companies to cut costs, and the traditional financial data solutions are looking less attractive at their sky-high prices.
Enterprises are where the money is at for now, particularly with their huge amounts of data just waiting for analysis and their thirst for analytics. Democratising access to financial content has always been our mission, though, so SMEs and retail users will not be forgotten. Besides, any improvement on the retail side, which tends to be for the web platform, is simultaneously useful for SSO enterprise groups.
As we add more non-English content and cover more countries, we expect to branch out from the UK and the US, where the bulk of our userbase resides. The future is looking bright.