Honesty in design of online conversations

About fake conversations and misleading UI elements in promotional messages

Honesty in design is not only about being transparent with your users and avoiding dark patterns. It's also about selecting the appropriate UI element for an interaction. Let's take LinkedIn's promotional messages as a case study.

Like many of you I periodically receive promotional messages from LinkedIn. Here are three examples that illustrate how UI choices affect perceived honesty and user emotions, ranked from most honest to least honest, and from less frustrating to most frustrating.

(please note: all names have been changed)

First example: setting clear expectations

This is from the "LinkedIn team" and the topic is clearly identifiable.

When I open the message I see it is signed by "Jessy". While I cannot engage with Jessy I do not really mind as this is clearly an automated message.

Second example: vague topic but clear CTA to the rescue

Now, this is from Ann. I can see she is from LinkedIn and I can identify the topic. For a moment I think I might be able to interact with Ann but when I open the message I see a big blue button at the bottom of the screen. This button is a clear indicator for me that this is no regular chat. So while I can finally not interact with Ann either the UI choice here makes this feel OK.

Third example: advertising in disguise

A message from someone called Peter. It’s a LinkedIn offer, the topic is unknown and it starts off as a conversation. I expect that I will be able to chat with Peter. I open the message and start reading.

Notice how, unlike in the two previous messages, the message header here is exactly like the one you get when you start chatting with a new connection. This subconsciously reinforces my belief that I will be able to engage in a real discussion. I continue reading.

Aha, it is about their Premium offer again. Great, now I can finally tell them why I have never accepted their 'free’ trials. My eyes move down, I see the same suggestion pill I get in real conversations, in the same location. I now expect to be able to type something too, but alas, I cannot. Not only can I not express myself, the only option I have is to reply in words that LinkedIn has selected for me. "Free trial? Great!"
I can only react with this naive enthusiastic phrase when all I want to answer is "Sure, I am interested in a free trial as long as I do not need to give you my credit card number. And, can you please remind me what LinkedIn Premium is about again? "

Instead of a neutral experience this is now becoming a negative one. The copy and UI choices promised human interaction, yet that promise was not met. On top of that LinkedIn is trying to put words in my mouth, words that I do not like. And finally, notice how the usual, albeit low-contrast, “replies are deactivated” disclaimer in absent here too.

This is a great example of how misleading design choices can create an emotional roller-coaster for your users.

So, what can we learn from this if we want to practice honest design in promotional messages that do not allow for interaction ?

1. Clarity: ensure users can easily identify the sender and the topic.
2. Avoid fake conversations: don’t frame promotional messages as conversations unless genuine engagement is possible.
3. Visually differentiate from conversation UI: avoid using the same UI elements as in real conversations (e.g. suggestion pills) and use clear indicators to signal that this is not a conversation.
4. Respect user expression: don’t presume users’ emotional states. Avoid putting words in their mouth and stay as neutral as possible.

But why not actually allow for a conversation?

A moment of failed conversion is a opportunity to learn. By allowing a conversation (or at least some form of feedback) you may actually learn what is holding your prospects back. This gives you the opportunity to clarify misunderstandings, present your offer differently and improve it in the future.