A critical analysis of current conversational platforms

Understanding chat

In a real-time in-person conversation

  1. Receiver : You know whom you gone talk 
  2. Content : You know what you need to convey
  3. Time : You know when to do that.

"The physical presence of the person makes it possible for you to validate his context of action and pitch the conversation accordingly."

In a virtual ( chat ) conversation


  1. Receiver : You know whom you gone talk 
  2. Content : You know what you need to convey
  3. Time : You don't know when is the right time to do that.

"The absence of physical presence of the person keeps you unaware of his context of action and state of mood. ​​​​​"

In a group ( in-person ) conversation


"Here instead of a single receiver, there is an audience. Everyone should be physically present in a closed proximity at a defined time for having a group conversation."

In a group ( virtual ) conversation , chatrooms


"Everyone is virtually available all the time. Everyone person has their virtual instance ( user profile ) in a group which makes all of them at very close and equal proximity."

"Here it’s just the instance of a user profile which is available but not the intentional  conscious attention of the user."

Based on my research and analysis I have identified the following pain points in group conversational platform currently available.

Understanding time as a dimension

Most of the chat interfaces are not designed to reflect this dimension in its actual proportion. Even if the messages are posted at different intervals of time, in a conversational interface they will be placed in equal proximity.

"Representing conversation with this dimension will add a realistic feel to the communication.But the scale should be dynamically identified with readability in mind."

Let me say it

Imagine a conversation is going on. Person B has mentioned a point on which you have an opinion. You want to convey it there after that message. You started typing........ But before that here comes another message from person C, because in a conversational platform every user has equal freedom and access to post a message irrespective of others. It can be relevant to person B's point or a different point, thereby triggering a context switch. You are, therefore, missing a chance to contribute effectively to that context. This scenario is depicted in the diagram below

"To solve this, we need to accept that for a  healthy communication it is important to give respect to other's space/effort/opinion and listen to them. A quality which we lost during the transition to a world of digital conversation."

Let's see how it happens in an actual physical scenario. If you see that someone else want to convey something, you will respect their effort by waiting till he makes his point.


" This happens because his act of conversation is visible to you.But in virtual conversation platform others behaviors and activities are not transparent enough, in fact, its hidden."

If it's possible to convey that visibility to others, it might tempt them not to interrupt in another person’s act of conversation. This can be possible by providing some visual cues to the interrupter.

Show a real-time instance of other users engaged in the act of typing. Thit instance should denote the start,end  approximate duration of the activity.

Despite that, if someone is trying to convey his opinion in between, the system should show how his is interrupting the other person.

  • Show how the two conversations are overlapping
  • How much noise is created

  • Where he is interrupting during other person’s course of conversation

  • This should be visible for all the users 

Even after typing give some visual cue again by shrinking the send button according to the amount of overlapping.

"If it is possible to bring more transparency on this interface with relevant visual cues, an interrupting user will be able to empathize on an interrupted user and thereby providing their required space and time"

Lack of defined context : Breaking the context

A thread started with a conversation about topic A can anytime switch to a conversation about topic B . This is not in the control of participants of conversation A. Anyone can pitch in any topic at any point in time. This can lead to a break in the initial conversation even an end to it. The main reason is the lack of visible context of a conversation. 

A conversation timeline is a single instance of representations of individual units of messages listed in a chronological order. Each unit of message reflects information regarding 

  • The person posted it
  • The amount of content , quantitative 
  • Type of content ( text , image ..)
  • Relative Position.

These information’s together forms the context of a single unit which you conceive easily without much effort. But to get the context of a conversation the user has to do a conscious effort of reading and understanding enough unit of a message. Sometimes these messages will be scattered around in timeline which makes it more difficult. 


A suggestion can be to have a temporary instant created for each context. Let it be active parallelly. There should be a maximum idle time after which the context will be closed.

  • Contexts will be aligned horizontally
  • It can be navigated with horizontal scroll
  • It will be arranged in the order of its relevance and activeness
  • User can tap a context area to start posting to that

"It has been identified that giving the user a preview of multiple contexts of conversations active at that point in time will provide better visibility. How to represent that in an effective manner needs to be studied and analysed"

Asking for attention by distraction

With the advent of technology and availability of the internet, it’s assumed that any message you send will be delivered to the person within no time. Now there are push notification which can push any such messages to your attention even if it’s out of your course of action. But by such a deliberate action, you are not respecting the persons time and effort which you would not have considered in a real physical scenario. 


To a certain extent, this can be handled by providing an option for both sender and receiver for selecting a priority level on sending and receiving the message.

“ This topic has been studied in detail in my case study about Notifications “ 

Assumed consent

At times, there are open-ended messages which seek reaction from users.These can be 

  • Statements which need approval from others to make it valid

  • Queries which need suggestions from others to be solved 

  • Orders which needs visibility and consent from others to be followed.


The Conversational system has a very linear structure. It is monotonous in the way that each message unit behaves the same. Because of that such open-ended messages are assumed to be perceived the way user wanted. This is called as assumed consent. Such messages should be equally visible to all the users so that it will be heard.

  • Equally heard : It should not get hidden with the scroll fold.

  • Monotonous nature : Messages should have different behavior based on the content. 

  • Effective feedback : Such open-ended messages should behave in a way that users can react over that to change its state.

  • Change in the state : Statements should be stated as valid with a given number of users approval action.

  • Change in the state : Queries should be stated as answered if the best and the correct answer is received

  • Change in the state : Orders should be stated as seen and agreed from a given number of people  

"Certain messages which need to be equally heard should behave differently.There should be a mechanism to substantiate the reach it had"