This article first appeared on my medium –


Digital commerce has evolved drastically and is undergoing a major flux right now. Amazon’s dash to IKEA’s catalog to Amazon Echo to InnerChef hungerTap, companies are trying their best to reduce the friction and reach closer to their customers.

The current iteration is the 4th coming of digital commerce:

Version 1.0: Browse-led — Listing sites of 1995, craiglists of the world

Version 2.0: Catalog-led — ecommerce, pioneered by Amazon

Version 3.0: Graph-led — Marketplaces — Flipkart, Amazon (in India)

Version 4.0: Assistant-led — UBERized model (more on this)

Version 5.0: AI-led

There is not much to talk about version 1 to 3, and our focus in this article is going to be version 4.0.

Assistant-led digital commerce

Let’s go through the current user-experience of purchasing online —

  • Figure out what you wish to buy
  • Visit,, or what have you
  • Visit few price comparison sites
  • Look at reviews, do some price comparisons
  • Once convinced, add the products to cart
  • Make payment, add address
  • Get the product delivered

This process, depending on the product, could take an hour to few days. The effort taken is just not worth, plus a satisfactory outcome is heavily dependent on the research expertise of the buyer. Yes, some people enjoy this experience but I strongly feel, that’s only transitionary.

Instead, let’s look at assistant led user-flow

  • You wish to buy a new product let’s say, a headphone
  • You have your budget sorted out, let’s say Rs 3000
  • You open an app and post the query “I want a headphone, my budget is 3000. I listen to a lot of music and so a headphone with greater range of frequencies will be perfect.”
  • And an assistant — digital or human, responds to the query, does some processing and suggests few options
  • You take a pick
  • Your mobile wallet is already connected
  • A payment confirmation push alert is sent to you
  • You approve it
  • Payment goes through and order gets delivered

Let’s look at this once again, All you did was —

  1. Place a query
  2. Approve payment

This user experience isn’t too far away,

  • Uber is essentially a digital assistant to get cabs
  • Amazon echo is a digital assistant to order products off amazon
  • Magic is an order anything text service, powered by humans
  • Haptik is a partially automated and partially powered by humans, customer support assistant

This paradigm will be heavily adopted in near future. As users get more comfortable with buying online and the information overload goes beyond comprehension, the need for assistance is going to become the norm. Depending on the product, the assistance will be either

  • fully human
  • partially human
  • fully AI

This is particularly useful for sectors where customization is an important component in the buying process. For instance, hyperlocal services. Hyperlocal service apps are mushrooming in India and I find their user-experience particularly bad. It’s really bad design, that while hyperlocal services are becoming mobile-first, they expect users to fill form after form to put up a simple request. Filling forms on mobile is a particularly painful experience. I will rather call than filling up forms. A chat based assistance will be so much better and frictionless for their users.

When I was designing Tushky’s mobile apps, I was faced with this issue. I built the first version of the app, pretty much like an ecommerce app — browse the experiences in cards and then book the experience. After building the app, in our beta tests, we realized how painful this experience is. Unlike product booking, where you know what you wish to buy, you don’t know what you wish to do this weekend. Add to it, you don’t want to spend an inordinate time to plan a quick weekend. That’s where, we decided to scrap off the app and build a new one centered around the new paradigm — Experience booking should be assistant led. Our new apps are chat led — whenever you wish to plan a weekend, just pop up the app, put your query and Tushky’s city insiders will respond in few seconds with up-to-date information of interesting experiences and things to do in the city. Once you are happy with the option, just approve the payment via mobile wallet and you are done. All our early beta tests have given encouraging results and we are looking forward to release the apps in public on 23rd June 15.

Sign up for Beta access here:

Download Chrome extension from here: TUSHKY CHROME EXTENSION

UPDATE: 13 Nov ’15

Above links will not work as I have discontinued working on Tushky. The apps were launched the hypothesis was proven right but it was too late. I had lost team members and was out of cash.


Please enter your comment!
Please enter your name here