Over a million developers have joined DZone.
{{announcement.body}}
{{announcement.title}}

Using Big Data to Value Startups

DZone's Guide to

Using Big Data to Value Startups

Learn about research that's improving the process of analyzing big data to accurately value how much startups are worth.

· Big Data Zone
Free Resource

Learn best practices according to DataOps. Download the free O'Reilly eBook on building a modern Big Data platform.

Apportioning appropriate value to a startup is often a very subjective matter, with little in the way of cast-ironed methodology for determining the growth potential of a new company.

A team from Oxford University believe they've come up with a better approach. Their research attempted to better value startups in the information and communication sector.

The team, who collaborated with Chinese tech giant Huawei for the research, utilized big data consisting of every UK startup in the ICT sector between 2006 and 2015. In addition to the 143 companies they tracked, they also recorded the patents registered by each company during that period, with over 1,500 in total.

This data allowed a model to be developed that was based on the particular characteristics of each patent, the characteristics of the companies themselves, and the niche they were hoping to break into. By combining these, they were able to assign a value to the startup and understand any variability in that value.

When tested on historical data, the model was able to accurately account for around 85% of the variation in the value of various technologies owned by the startups in the database. They believe that this could, therefore, provide an accurate means of predicting future valuations too.

Valuing Deepmind

One of the more noteworthy companies analyzed by the algorithm was AI startup Deepmind, who was famously bought by Google in 2014 for $650 million. The model estimated the value of the company at between $590 million and $660 million — pretty accurate.

"The model has already proven itself to have a highly accurate predictive power, especially for technologies at an early stage of commercialization," the authors say. "It also uses completely objective data, which is a significant improvement on existing methods."

The next step is to try and bring the model to the market and find some commercial applications for it. With significant interest in the startup sector from the finance community, larger corporations, and government innovation bodies, it's a model that should find some willing users.

The team does note, however, that at the moment, the model is only really applicable to the ICT sector, and so has some limitations — although they believe that they could easily develop similar models for other industries.

Find the perfect platform for a scalable self-service model to manage Big Data workloads in the Cloud. Download the free O'Reilly eBook to learn more.

Topics:
startups ,big data ,data analytics ,algorithm

Published at DZone with permission of Adi Gaskell, DZone MVB. See the original article here.

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}