July 18th, 2019
Leads to LeasesRSS Share Category: Customer, Data Science, H2O Driverless AI, Use Cases
By: Priya Jain
There is such a large amount of unstructured data being produced by companies. I personally find it so interesting that there is so much meaning and hidden value in text, audio, and visual content. Until recently, much of this data would go unused. However, since the rise of machine learning and artificial intelligence, it became possible to take this immense amount of data, find patterns, and gain knowledge that could transform business decisions. G5, one of the leading marketing optimization companies in the real estate industry, recognized and benefitted from the power of machine learning.
G5 provides solutions for property manager by creating leads so they can fill their properties. G5 works with over 7,000 properties all over United States and Canada. A significant portion of possible leads come through calls. Phone calls are a form of unstructured data and identifying the caller’s propensity to rent is a huge grey area because it is up to human discretion. In fact, only 1 out of every 7 callers have the intention to rent and so by applying machine learning methods G5 could identify true caller intent accurately. By knowing the caller’s propensity to lease, G5 can then more effectively identify leads and raise conversion rates.
It is fascinating how crucial deciphering this data is to the company’s efficiency and performance. G5, which did not have resources allocated for data science, was able to build a predictive model to score caller intent and have a real-world scoring accuracy of 95% all due to the wonders of automatic machine learning.
By using H2O.ai’s Driverless AI, an automatic machine learning program, and Amazon Web Services, G5 could take a training dataset, conduct feature engineering and from there build a predictive model. Of course, this is a very simplified explanation of the process, but get this: it only took 3 months to see this whole process through with a team of about 3 engineers to achieve such large-scale results. H2O.ai provided a small company like G5 a trustworthy, simple, and effective solution in a short amount of time. In fact, Martin Stein underscores “the level of trust we have seen and I have put into Driverless AI has gone from somewhat really good to oh my god I really trust that product and that now lets me speed up the process and be more confident.”
For someone like me who is deeply interested in decision making, the possibilities seem endless with automatic machine learning. Every business in every industry can benefit from analyzing data to reach strong decisions and effect change.
Being a newbie in the industry, I get inspired by hearing cases of AI enhancing business decisions, like in G5, or even saving lives, like in healthcare companies. I hope it excites you too!
Check out the G5 Case Study for more details.