How Artificial Intelligence is impacting classifieds and middle men
The middleman or broker plays a vital role in many transactions. In India we have millions of them. Earlier set of internet pioneers thought this is inefficient and thousands of online listing platforms were built across the world to facilitate principal to principal transactions and disrupt the middleman. Online classifieds covered everything in newspaper classifieds or yellow pages like jobs (CareerBuilder, Naukri), real estate (Zillow, 99acres) and second-hand cars (Autotrader, Cardekho).
However, rather than disrupting the middleman, it is the middlemen who benefited most. Middlemen are the power users of listing sites using them to reach more buyers and sellers and quickly find the matching demand or supply. This has both grown their businesses and reduced costs. As a result, middlemen provide the majority of revenue for listing sites.
This surprising turn of events had three key drivers:
Listings are so numerous that even after using parametric search there are too many options for end users
Transactions have a service element because verification (e.g. softer elements like communication skill) or fitment is difficult to derive purely on the basis of a listing
Middlemen work on small margins and their families depends on the commissions they earn thus they fight tooth and nail for their survival
This was all good when the dominant model was listing like Google results. However, today, Apps capture much better user data and leverage personalization and AI to drive an exponentially better user experience. Uber sends you a cab rather than show you a list of cabs. People don't want result lists, they want results…a solution to their problem.
Such experiences are increasing users’ expectations. What worked yesterday is not good enough today. We see the evolution of artificial intelligence startups who are addressing this gap. Using tons of historical transaction data, AI can derive user needs and recommend the right fit much better than parametric search. The holy grail of consumer segmentation is a market segment of ‘1’ i.e. each customer is different, and services are personalized.
We find examples of this across verticals. In recruiting startups are mining social media profiles (like LinkedIn) to understand what kind of people fill a job role (say product manager). Now their recommendation of candidates to fill a product manager role is much better than just typical candidate filtering criteria. This can be further customized to each company based on their hiring preferences (like which college, what marks) which are derived from their own recruitment records. By reducing noise, AI reduces the need for manual review of hundreds of resumes.
The realization that data is the new oil means companies are collecting more personal information. This data can now be leveraged by AI to discover patterns which were not even a criteria before. People like staying near people who are similar to them (in terms of status, values etc). Thus, softer parameters like community and education of the property buyer are becoming input parameters to AI models in real estate.
AI capability to harness data and automate steps like communication, scheduling, reservation etc. adds up to massive change in the economics of a transaction. Historically the biggest impact of improved economics is to drive more transactions. The advent of electronic share trading led to a massive fall in brokerage charges and an order of magnitude increase in transaction volumes. Brokerages (middlemen) and exchanges (listing sites) who adopted technology and evolved their business models thrived. At the same there are many regional exchanges and smaller brokers who did not, and they disappeared.
It is time middlemen and listing sites imagined the future and their changed role in it. By adopting a disruptive technology like AI they can benefit from the paradigm change or as they say, if you can't beat them then join them!