Redefining Classifieds One Step at a Time

The classified ads industry is an atypical business. It relies on email and telephones in an era of SMS and social networks. It sticks to anonymity when it is seen as suspicious to not show your real name or have a social media presence. It is as if the clock stopped in 1999 for this industry.

Classified ads is an industry that has barely innovated and is, therefore, outdated, becoming more and more inefficient and dangerous. Every single day there are thousands of perfect transactions, but there are also many scams, rapes, robberies, and murders linked to almost every single classified ad service.

Craig Newmark once said that, “craigslist works because people are good.” Sure, most people are good but evil people have a bigger impact than good ones when the technology is on their side. So the question is, is this really the best we can do?

Today, the classified ad industry is dominated either by craigslist in the USA and Canada or a craigslist’s clone in the rest of the world. Many other services have tried in vain to take them down, no one could. The common explanation is that these services have been around for a long time and they have massive user bases, that locks customers in even if the experience is better somewhere else.

The reality is much more nuanced. Classified ads suffer from two major issues – trust and noise. No one knows who he or she is talking to until they meet and the over saturation of most classifieds sites makes it difficult for members to find what they want or be found easily.

Every new classified ad player resolved only a fraction of the first issue through strategies such as ID verification, community ratings, social networks, etc. but no one cared about the noise. The result is that every year, dozens of startups come along to defeat Goliath without any success because they were either empty or noisy. Hence, users didn’t see much value in simply trying the new services.

In order to disrupt this field, a radical new approach is needed. With chatbots and AI advancements, we can think about classifieds differently.

At its essence, a classified ad is just a piece of text that express an offer or an ask. We can simply let people express themselves in a natural way to a chatbot; after all, markets are nothing more than conversations, right? A chatbot can then recognize the user’s intent, ask him or her to provide more information if necessary, and help him or her to get in touch with the right person.

Here is a potential scenario for a typical transaction:

a - a sell scenario :

Seller: I want to sell my iPad 4 Mini for $500
Bot: Can you send me some photos, please?
Seller:
Bot: Kewl ☺️. It’s noted. I’m gonna do my best for you now. 😌

b - a buy scenario:
Buyer: I’m looking for an iPad 4 Bot: I think I may have something for you. IPad Mini 4, 500$
Bot: Based on my experience, this seller seems legit (5 stars), made 3 successful transactions last week, price is ok too and he is pretty active on facebook, do you want me to schedule a meeting with him ?
Buyer: Yep
Bot: No problem, tuesday morning at 10am, 833 S Spring St, Los Angeles (Verve Coffee, 2miles from your location), is that ok for you ?
Buyer: Perfect

All bot users are already registred on a social platform hence the bot can apply machine learning algorithms and evaluate the seller’s credibility based on his social activity, his previous transactions, the average price of the item on the market (you know we all have seen that brand new iPhone 6s for 200$…). The bot can also plan and schedule a meeting between the seller and the buyer in a safe place that is likely to suit both parties.

If we go a little further and apply machine learning to users classifieds we detect common scams and other suspicious classified patterns, which will prevent users from scams and other potential threats. And since almost all bots are implemented on top of major messaging platforms, the environment becomes hostile for potential scammers or swindlers.

Local business is part of our daily lives, by applying current technologies, we can already fix trust and noise issues. Sooner or later A.I and bots, will enable a big disruption in the field and will change the way we do business locally.

This is why we started building and experimenting this approach a few months ago on Bly : a smart virtual assistant that you can chat with on Messenger to buy, sell, rent around you. Our first bêta users from France and UK had a great experience - “it’s so much more natural and intuitive”.

Here is a preview of the main tasks that Bly can perform