We use Artificial Intelligence (AI) and Human Intelligence to recommend software products.
AI because humans can be biased and humans because AI is not yet intelligent enough to automatically recommend products on its own. Hence, we use both humans and AI to recommend software products.
Our AI-driven algorithm uses a combination of sub-algorithms that are made-up of text analysis, vocabulary analysis, machine data and user-behaviour data. These algorithms effectively analyse B2B software products on the web and look at over 236,000 nodes. This, purely data-led intelligence is the foundation of our recommendation engine. Allowing our recommendations to be personalised for each software buyer.
On the human side of things, we speak with software buyers to understand their experience of researching software products. We have surveyed software buyers at events and exhibitions in London and Manchester. We have digitally chatted with different types of software buyers – interns, analysts, admins, super users, operational decision-makers and C-Level decision-makers. We have also spoken with software buyers over the phone who have contacted us for software recommendations.
On the human-side, we also regularly speak with software vendors. Each product that is listed on our platform is listed after we’ve personally viewed their product demo from their product team. This helps us to understand and validate almost every feature of every software product before listing it.
With software vendors, their product demos to us last for about 2 to 3 hours. We typically interact with product managers, pre-sales engineers, sales teams and account managers to fully understand the strengths and the weaknesses of a software product.
Since we do not use reviews and Ads, our recommendation is purely based on data that is collected using AI, software buyer’s requirements and software vendor’s product strengths. The outcome is 73% of software buyers who use Comparesoft end-up communicating with at least one of our recommended software vendors.