GeoIQ

More innovations would be needed to make sense of data, and draw value from: GeoIQ

GeoIQ, a hyperlocal performance prediction startup, is a locational intelligence platform that tells the data values of each location — in terms of people, behaviors, businesses, and potential opportunities — as easily consumable layers on maps.

GeoIQ utilises proprietary algorithms to layer government data with other 600+ trusted public sources and satellite imagery into generating 100m x 100m geospatial grids. The AI layer on top uses these data points to predict user behaviour, affluence and business potential at an address level.

Here, Devashish Fuloria, CEO, GeoIQ, tells us more. Excerpts from am interview:

DQ: What is GeoIQ, and how did the idea originate?

Devashish Fuloria: GeoIQ is a location intelligence company that captures real-world data to provide hyperlocal insights at the street level. We tell businesses where to be, where their target audience is, and where their markets exist.

Devashish
DQI Bureau | DATAQUEST Devashish Fuloria.

We source data from more than 600 government and public data sources, satellite imagery, and many other channels. This data is transformed and structured under more than 3000 location attributes/ variables such as population density, income groups, presence of brands, average rentals, average meal cost for two, presence of banks and ATMs, presence of commercial and residential setups, and others. These attributes give insights into user behaviour at a location.

Simply put, businesses can get a lot of information and insights about the users based on their presence. BFSI businesses can predict risk, fraud, claims propensity, etc. when they add the location element to their prediction models. Similarly, retail brands can identify where their next store location should be to maximize footfall and revenue.

The idea originated when as data scientists themselves, the co-founders were struggling with the accessibility and usability of real-world data. They could witness how they were losing critical insights just because the real-world data was not readily accessible and consumable. Once the problem of harnessing this data was solved, the next natural step was to decide how to use this data. That’s when an AI layer was built atop this data to draw insights from it and make sense of the massive volume of data that was being collected. That’s how GeoIQ originated.

DQ: What is location AI, and how can it help unlock growth potential for business?

Devashish Fuloria: Location AI refers to the capabilities that help draw insights and visualizations from raw location data. The real-world data is unstructured and needs processing and transformation into a structured form before consumption. Once the data is structured, an AI/ML layer works on it to draw critical insights for an address or a latitude/longitude co-ordinate.

In simpler terms, location intelligence or location AI is a way to get answers and insights from raw real-world data. Once you surpass the challenge of sourcing, collecting, and storing real-world data, a more significant challenge then is to make sense of it and utilize it to solve business problems. Location AI comes into the picture to understand trends, patterns, and correlations in the data to draw actionable insights. Before that, it also discovers which location variables impact your use case.

For example, you have data for the average rent in a catchment, population density, average household income, and average meal cost for two. Now when the AI layer works on this data, it will first identify which variables have an impact on your use case and which do not. Let’s say that only population density and meal cost for two are identified as impacting variables by the model. Now, it will also tell you that “a suitable location is where the population density is 17,000 per sq. km. or more and average meal cost is INR 500”.

This level of intelligence can be applied to solve some of the pressing problems across industries and achieve unmatched prediction accuracy. Some of the biggest problems that we are solving across industries are credit risk prediction, fraud prediction, site selection and retail expansion, lead prioritization, hyperlocal digital and OOH targeting, and more.

DQ: Do you see any specific sector that can phenomenally gain from data analytics?

Devashish Fuloria: Location data analytics or intelligence from real-world data has use cases across industries. It could be used for credit risk prediction, fraud prediction, claims propensity, retail expansion, and site selection, hyperlocal targeting, understanding user behavior, and affluence, lead prioritization, and so much more. For us, most of our clients currently are in fintech and insuretech, retail, and F&B sectors.

Major fintech brands like NAVI and DMI Finance augment their existing databases with location data to predict credit risk with better accuracy (especially for the users who are new to credit) and prioritize leads based on affluence prediction. We provide them with location data and intelligence for over 3,000+ location variables categorized under demographic, socio-economic, infrastructure, crime, and more. These are population density, household income, average rental, and many more variables.

For retail brands, we are revolutionizing their market expansion and site selection with our sophisticated ML capabilities. The retail major, Lenskart, describes our model scores as the IMDB rating for their prospect locations. The custom ML model built for them understands location attributes/ variables that form the success metrics for a store. Once this learning is complete, the model scouts for look-alike locations that possess all the attributes in the success metrics for a store. These look-alike locations across cities are the recommendations for them to open their new stores.

As a step further, we are shortly launching our subscription-based site selection product ‘Retail IQ’. It is a platform for retail brands across expansion stages, be it their 10th store or 1000th. The platform comes with an easy-to-use interface, an integrated app for the field force, and an interactive dashboard to monitor progress and access reports.

Brands can simply input the location parameters that impact the success of the new store and the platform will recommend locations in the selected city within seconds. Users can then shortlist or reject locations based on the scores, and their preferences. The shortlisted locations would reflect on the ‘Geowise’ app for the field force to scout.

DQ: How do you see the Indian data science industry grow in the future?

Devashish Fuloria: The problem with data initially was that there was not much data available. With digitization, and the era of big data, this problem was uprooted in no time. But, the new problem now was, how to manage, structure, understand, and utilize the large pool of data that is being generated by the minute.

The future of the data science industry is very exciting simply because, as we generate more and more data, more innovations would be needed to make sense of this data, and to draw value from it in the form of business insights and forecasts.

DQ: Why did you choose to become an entrepreneur, and how will you describe your journey and experience as a founder till now?

Devashish Fuloria: I did not choose to be an entrepreneur! I was just looking for problems to be solved. Everything, then, just happened.

It’s been a great learning experience. While speaking to a few fresh MBA grads a few months back, a realization came that every month, building a company is like an MBA degree in itself. Yet, every day, I feel like a fresh student, figuring things out.

Leave a Reply

Your email address will not be published. Required fields are marked *