Interview with Dmitry Tsyplakov, CEO/Product Owner, Fincase – #1 Property Technology Company across Russian market
FintechNews (FTN): How does big data and machine learning affect the real estate market?
Dmitry Tsyplakov (DT): What is real estate investment? Can artificial intelligence be present there, or is it profiteer too? I can say for sure that scrupulous search for profitable offers using the old methods in a rapidly changing world and the requests of potential customers do not work. If you want to get ahead of your competitors and invest in profitable premises, then it’s time to search for new ways to find them. For example, use big data and machine learning.
Machine learning (ML) is a so-called artificial intelligence, which task is to collect and analyze a huge amount of information in a minimum amount of time. Therefore, it solves one of the main problems that I think almost all companies, investors, and just people face today – the search for something valuable in an endless stream of information. Machine learning combines and processes various sources, removes unnecessary data, and outputs the result according to your query.
(FTN): How do customers and businesses benefit from market transformation?
(DT): As I have already said, information is very important in real estate. It is used for 90% of all decisions and its quality determines whether the investment will be profitable or unprofitable. At the same time, information must be provided quickly and in a way understandable for customers. They should not think about how much data, sites, tables, and documents were analyzed, or how many unsuitable offers were rejected. This should be similar to a phone app. You enter the necessary parameters and immediately get a list with the relevant offers.
Even just finding a suitable apartment for living is a complex and very responsible process. But standard search engines can only filter offers by the number of rooms, price, or location. This also works, but not effectively. The client still has to look through a huge list of offers and rely on intuition in choosing one or the other option. Now we use personalization algorithms that identify the user’s preferences and based on them produce a result with apartments where, for example, the walls are painted blue or are located in an area where families with children live. Moreover, these algorithms take into account the preferences of other users with a similar query. This precise determination of needs and tastes allows offering what the customer needs and sealing a deal much faster.
Machine learning together with big data, integrated into advanced technologies, take the real estate business to a different level of quality. Investment risks are minimized. Companies are forced to consciously approach their work and work with clients. There are cases when competitors were specifically connected to the big data in order to ruin their reputation in the event of foul play. Such digital intelligence changes the campaign itself and business strategy.
(FTN): How does artificial intelligence help you invest profitably?
(DT): When investing in real estate, the condition of the room itself, its purpose, be it commercial, residential or industrial and its square meters are not the only important things to consider. Today, it is important for an investor to know whether there is a kindergarten and a school nearby, whether it is possible to get there, how far is Starbucks from the planned office, so that employees can run for coffee in 5 minutes, how often the bus runs and whether there is a subway nearby. These are the parameters that are not easy to find, but they determine the benefits of future deals. Location and environment are of great importance. All these, what appears to be little things, help you understand where and how much to invest.
Another advantage of artificial intelligence is that it can predict the future, which is extremely important in an industry such as real estate. For what price to buy today and will it be profitable to sell tomorrow? Companies that use artificial intelligence and “predict” fluctuations in rental or sale prices do not go bankrupt. And this is today’s reality.
The success of the deal depends on whether the agent evaluates the property correctly. If with standard premises and apartments, the number of rooms and the sole fact of a repair is sufficient for the initial assessment, then elite real estate with designer repairs and non-standard layout is much more difficult to evaluate. Market leaders use artificial intelligence to estimate the value of real estate based on photos. This allows you to adequately evaluate even the most complex interior details that actually “sell” it to the buyer. In other words, they work with the client’s emotions, showing interesting details to a particular buyer. Someone wants a retro-style bathtub, someone wants a functional granite countertop, someone wants parking for 500 cars. This approach, combined with standard parameters, allows companies to set the price with a margin of error of only 2%.
Real estate is a perfect place for the development of artificial intelligence. Everything is important here. For example, a McKinsey study showed that having two grocery stores within walking distance increases the cost of an apartment, while having more than four leads to a decrease in the cost. And there are many examples of this, and people are not always able to track it, while the “machine” opens up new opportunities in data search and analysis. It is able to find the most unusual factors that affect the final price of real estate.
(FTN): What are the prospects for the development of technologies in the real estate industry?
(DT): The intrusiveness of artificial intelligence into real estate is not surprising. Companies around the world are making significant progress in data collection and analysis, but there is still a long way to go before we take full advantage of it, especially in Russia. Artificial intelligence is not what we have right now; it is a continuous process of learning and development every day. And the faster companies learn to systematize and standardize the information they receive, the faster we will be able to evaluate the benefits of working with AI. These are not the cheapest investments, and they are unlikely to be profitable in the short term, but they will definitely give advantages over competitors and increase profits in the near future. Because artificial intelligence allows you to see a bigger picture of offers, evaluate their real value and predict prices.
To effectively implement AI, companies need to learn how to share information and help each other. We are just at the beginning of such a relationship, but I believe that it will become a reality as soon as companies realize that it is more profitable to share and receive than to quietly watch from the sidelines.
Our lives, tastes, and preferences are changing, and companies must also quickly adapt to new demands. With the help of artificial intelligence, it will be easier for real estate agents to find the perfect home for anyone.
(FTN): What solutions does Fincase offer?
(DT): The company Fincase has developed a unique product for the real estate market – the scoring value analysis (SVA). As we have already informed, when buying/selling real estate, you need to know its exact market price, that is, to make an assessment of the premises. With traditional manual assessment, you need an analyst and a large amount of time to collect data for a single object. At best, it will take a few days or even weeks of work to prepare a detailed report with the area crime statistics, pedestrian and car traffic, data on investments in this area, whether it will become more appealing and comfortable for people or in a year there will be another highway that may be an advantage for offices, but it is unlikely to appeal to residents. SVA makes this analysis deeper, and most importantly incomparably fast. It takes no more than 3 seconds to collect and compare numerous data sources that give a real understanding of the surrounding area of the property and its value to the investor.
Fincase Data mining Department uses a unique parsing technology for data processing. The shared centralized database regularly receives new ads from partners, and data is also extracted from open sources. During the collection, the information obtained is being analyzed – it adds links, ratings and finds analogues. All this allows Fincase to accurately and quickly build an estimate for each specific object, and, if necessary, offer an alternative. Every day, the parsing system processes up to 150,000 ads and up to half a million per month.
A few years after its launch, Fincase is now the market leader in real estate evaluation. Our product has already been evaluated by the largest Russian banks and development companies, and we plan to create a platform that will combine all real estate market analytics services. The next stage of the company’s development is to enter the Western market. My goal is to help as many companies as possible learn how to make money on real estate transactions, even in a turbulent changing market.