A Computer Might Be the Friendly Face Someone’s Always Needed

We all know that online shopping has revolutionized how companies do business. People often forget just how much it’s changed the lives of customers though. Online shopping has been the norm for long enough to change the face of modern culture. There was a time not too long ago when people had to rush in order to get to the store after work.

Today people can actually think about what they’d like to buy. They can even wake up in the middle of the night to purchase impulse buy that they’ve been holding off on. It’s a method of shopping that works wonders for both customer and company alike. There’s little wonder that so many stores have shifted more and more resources toward online sales.

But there’s aspects to this shift which are a little more difficult to manage. One of the more significant issues is lack of real interaction during the shopping experience. It’s not uncommon for sales to fall through because a customer has unanswered questions. People have tried to address this issue in a number of different ways.

Advertising is of course one of the more popular ways to get information to people. But it puts a heavy limitation on how much information is delivered to customers. There’s also an option to simply put a huge amount of information on any given product’s description. But too much information is often just as bad as too little. When people see a mountain of text they tend to ignore it rather than sort through it all.

But this is where people start to use high tech solutions to tackle high tech problems. Most online stores use some form of database to store information about their products. In the previous example we considered a store which simply dumped all of that information into a single plain text description for potential customers to read.

If a customer were in a store they’d instead just ask about a specific issue and get a reply from the salesperson. It’s a fairly simple interaction when one thinks about it. Basically all the customer is doing is asking for a match between a keyword he provides and specific facts that the inventory system is aware of. The only problem for an online sale is that a database isn’t really capable of understanding plain English. It needs some form of translation layer to take in a request and output a human readable reply.

Various forms of AI are now meeting that demand. For example, any conversational AI software solutions usually try to mimic a human salesperson’s behavior. They can parse a customer’s question into specific keywords, match it to information in a database, and then provide answers in a similar way to a normal salesperson. Artificial intelligence might not be at the level of science fiction yet. But we’re at a point where it’s able to properly understand simple questions and answers.

One of the more interesting things about this process is that it can easily grow and evolve. AI is in many ways as capable as the speed and memory of a particular system. As computers grow faster the underlying software can take advantage of those changes. This means that as companies grow an AI based system can grow along with them.