Signal Media is an artificial intelligence media monitoring company that transforms the world’s information into accessible, actionable business knowledge. When the news cycle has shifted from 24hrs to 24 seconds, businesses need constant access to high-quality, hyper-relevant information to remain on top of a world that’s moving at breakneck speed.
They need the complete picture to be certain that the decisions they make reflect reality – not their assumptions. Signal Media, is the solution, as founder and CEO, David Benigson explains.
What does your business do?
All knowledge workers follow a similar process. They gather data, analyse it, extract insight from it and then act on it. But there are two problems. The first is that the digital revolution has flooded us with data, which makes it harder to find the signal in the noise. The second is that most of the expertise needed to effectively analyse this data is stuck in the heads of employees (80 per cent), and hard to access.
Signal is an artificial intelligence company that solves these problems. Our AI categorises, analyses and consolidates vast data sets to extract only the relevant signals/information. Then it digitises the commercial expertise of its users to offer the benefits of that siloed knowledge across the whole organisation.
Where did the idea come from?
When working in law and thereafter as a consultant, I was dealing with high volumes of information such as news media on a daily basis. The content was often very irrelevant, and required a lot of effort to sift through to find the valuable data in the noise. There had to be a better way! Machine learning had gotten to a point where it offered real potential as a solution. I partnered up with Dr Miguel Martinez and here we are!
How did you know there was a market for it?
Early in my career I watched people from all industries grapple with data overload and siloed information, every single day. Existing solutions such as traditional media monitoring have large client bases, but under-deliver. We’re a disruptor in the space.
How did you raise funding and why?
We raised a series A of £6.3 million in October 2016, in total we’ve raised over £10 million from fantastic investors such Frontline Ventures, MMC Ventures and Hearst Ventures. The expertise required to build AI systems is rare, and thus expensive, so we needed funding in order to run the business while we build our tools and go to market.
What is your business model in brief?
We use AI to offer unlimited services to our clients in real-time, something most of our market is unable to do when it comes to business intelligence. We offer a SaaS, self-service model, which means we can keep overall costs low while scaling our customer-base. As we grow, we bring in new forms of data to service new business units.
The highest point of the company was…
There have been so many high points in the 4+ years since I started Signal that its difficult to pick out just one. I think one very memorable high point was winning the Innovate UK KTP Best Business award, which was a culmination of our work with The University of Essex and some very cutting edge research we did in Machine Learning. It was about two years into starting Signal and was a fantastic recognition of all the efforts we had put into this business up until that point.
What advice would you give to other entrepreneurs?
My first response would be to ask: What about your business model, or the industry you are in, incorporates repetitive, data-heavy tasks? While the technology is making big strides, it is still limited to certain types of function. For example – baking a cake doesn’t need AI, but a big bakery company could use it to monitor and improve distribution methods, or communication channels. My point is that AI isn’t yet a panacea for all the ills of a business, but that it can free up the human capital of a business to do the things that it can’t do. Working out what parts of your business plan requires intense repetitive effort – that’s where AI can be most effective.
My next piece of advice would be to remember that AI is a service, not a product. It’s always improving, and can change the lives of users every day. And the more exposure it has to more stimuli the faster it will learn. Treating AI as a static product to do a specific thing (and no more) is a guaranteed way to fail. The more you push the envelope, the more things your AI can do, the better it will be.
My third would be to remember the data. The more data your AI has, the better its output will be. So make sure you have enough of it. An AI without data is like a broken pencil – pointless. And if you can build access to data at scale from day one, do it.
My last piece of advice would be to remember that AI is not valuable just by virtue of its existence. The value of any business service comes from solving customer/client problems, and AI is no different. Unless an AI solves ‘vertical’ business problems – problems that occur in business functions – and make clients’ lives easier, faster or simpler (preferably all three). then it isn’t going to be successful.
Where do you want to be in five years’ time?
Our eventual aim is to disrupt the $250 billion consultancy market with data-driven, AI-backed solutions and support for every role in every business function. We will be able to automate much of the research and analysis these giants undertake, eventually delivering insights with real business value far faster, far more effectively, and far more accurately.
If you weren’t an entrepreneur what would you be?
Working in social enterprise.
If you could go back in time what would you change?
It’s a challenge to find the balance between innovation and function. In such a fast-moving growth sector, it’s easy to spend years in R&D adding an extra percentage of quality in your algorithms or learning models. But without being in-market, it’s very hard to know whether that extra one per cent is really adding value to a client experience – and by the time you do get to market, you may already have been displaced by others with first-mover advantage. But go too early with a service that doesn’t deliver – as many have done with chat-bots over the last 18 months, for example – and you risk alienating whole audiences. Finding this balance between innovation and impact is key, and we could always have managed it better.