Updated by Dr. Hugh Kelley, Chief Economist, Chief AI Officer on 1-APR-2020
​
Our approach works by fitting an AI model to a training dataset, with the goal of identifying fast-growing companies also known as Gazelles and applying the results of these fits to a prediction dataset. This allows us to identify companies likely to expand in the next 12-18 months. These companies typically:

This means that for G-scores of 5, there is a minimum 75% chance that the company is a Gazelle, and therefore has the highest propensity to expand domestically or internationally within the next 2 years.

But after years of R&D, our research shows that at any given time, roughly 60% of companies are not in expansion mode. So, we designed an algorithm that allows you to find and target Gazelle companies first in order to minimize your outreach efforts.
​

G-score 1: Company does not look likely to expand

Depending on the industry, an average of 62% of companies are not looking to expand in the near term, and this increase to up to 81% for some industries. While this is a reality of the economy, it doesn’t mean you have to spend too much time reaching out to these companies.

G-Score 2: Your odds just beat 50%.

Half of our companies scored with a G-score of 2 resemble Gazelle companies. By calling G-scores of 2 and above, the chances of that company expanding in the next 18-24 months are better than a coin toss.

G-Score 3: Getting warmer...

From there, the signals only get stronger. Think of it as a cell phone signal - you’re still connected, but three bars of service are better than two! G-Scores of 3 become Gazelles 25% more often than G-Scores of 2.

G-Score 4: Warmer….

Now we’re in the rarified air of that 10% of companies in growth mode. G-scores of 4 resemble Gazelles up to 65% of the time, meaning they’re essential to contact whenever they show up on your list.

G-Score 5: Red hot!

These companies show the same signals as Gazelles up to 75% of the time. You’re now contacting the best possible leads in that industry.

G-Score 6: Jackpot.

We help you out by calling as many G-scores of 5 as we can to gather even more information, such as exactly where they’re looking to expand and how that expansion will take place. We typically add about 30 of these projects a month.

We recommend building a diverse portfolio of G-scores 2-6, so you can tip the odds in your favour. Now instead of spending two-thirds of your time on low-growth companies, focus on high G-score companies who display the same signs as expanding companies.

While this shows you the big picture, each score is also customized for the company’s internal metrics, the industry it’s classified in, and the region it operates in. It’s calculated using up to 940 different data signals/variables over 11 years gathered from dozens of data sources in order to make sure that the company’s outcome is forecast as precisely as possible.