Hi there.

As I continue down the route to looking at what our products may look like as AI develops and changes consumer behaviour, today I start looking at what an AI-first user journey looks like. I will share a case study from India that I thought was a fascinating example of how we can leverage an AI product and have consumers surprise us with their responses. 

While we figure out the question of the new AI ecosystem and where our discovery comes from, one thing remains known: We can only benefit from our own loyal audiences. That’s why we have developed the Building Direct Audiences master class packed with best practices and new ideas that are showing results. The full agenda is here.  

What’s top of mind for you right now? I’d love to hear about any big questions you are trying to answer, especially where we at INMA may be able to rally contacts and resources looking at the same things. Drop me a line at Jodie.hopperton@INMA.org.

Thanks, Jodie

3 questions we should ask about what an AI-first user journey looks like

As chat and AI overviews are increasingly becoming an important — if not the most important — form of search, how does that change the user journey as an entry point to our products? That’s a question I posed to the INMA Product & Tech advisory council recently.  

For many of you, I suspect the current levels of traffic from platforms such as ChatGPT, Perplexity, etc. are just too small to focus on. But as this set of charts from FT Strategies show, these platforms are accelerating rapidly while others are declining at surprising paces:

One thing that keeps coming up is although the volume is smaller, the referral traffic may be more qualified. They have a basic answer and want to dig deeper. 

Assuming the above is true and continues to develop, we need to at least start thinking about what the AI-first user journey may look like when it comes to our own platforms. Here are some key questions we need to know to start thinking about adapting the user journey:

1. What is the traffic source? 

We need to start to get to know the sources and the differences between them. Is there a difference between a ChatGPT user and a Perplexity user? What are the different signals we can determine and how do we react to them? 

2. What data do we have about the individual and the context?

What is the search term or question that was asked and in turn led them to your platform?

Lastly, what information do we know about the individual user? I’ve heard people say the AI answer machines don’t know that much about you. But ChatGPT has a pretty good handle on their users, as you can see from the screenshots below.

Most of this is pretty accurate by the way. The data is likely not structured in the way we are used to it — especially when dealing with nuances such as “may be in Helsinki” (which I was, but am no longer):

3. What signals are there? Is there an opportunity with this reader to engage further? Is there a risk of churn? 

While this may be a question about user behaviour once they come to our platforms, there may be valuable supplemental data that gives us insights. For example, could there be a way to determine if it is a standard user, a paid for/pro user, or an enterprise account on the AI platform the person has come from? 

There are a lot of unknowns so this is the time to ensure we are asking the right questions. Have you started thinking about an AI-first journey? If so, I would love to talk to you. If you can spare 30 minutes, please find a suitable time, here.  

4 key lessons on chatbots and audience engagement from The Hindu

Every year, Chennai has a large festival with about 2,000 concerts over the course of 30 days. There is a lot of different music, and it’s complicated for people to navigate through. So The Hindu built a chatbot allowing people to make their own schedule. You can see it in action here.

The target market is relatively small at about 60,000 to 100,000 people. But those who attend are invested in the experience. In fact they were so invested, they often spent five to seven minutes at a time querying this database, asking all kinds of questions. 

The chat experience brings more investment from the user side. It’s an active media experience, not passive. And they found the more invested the user is, the deeper the questions go with people wanting to know more. 

Pundi Sriram, chief product officer at The Hindu, told us they learned a lot in that process, including what else our user wants to learn. 

One surprise was that when people were planning their schedules, it was assumed they would want to know about the concerts — but they also wanted to know what to eat in between concerts. They noticed a whole new set of questions around this topic and ended up uploading menus of canteens and places where people can eat. This made the experience deeper still and fulfilled a specific user need. 

The other thing they noticed was that questioning led to a lot of other content consumption.  People wanted to know about the bands and artists — coverage The Hindu had and was able to surface. 

Pundi finished off by saying: “Maybe the interactive nature of chat, as opposed to a more passive query, brings about a different type of engagement.”

Here are a few learnings I take away from this at a product level:

1. Starting from a specific topic — rather than a blank slate — encourages engagement as users see a specific output.

2. Chat can be useful as a top-of-funnel search mechanism for content.

3. Chat can show us a lot about user needs and enable us to adapt quickly.

4. Chat users are highly engaged users as it’s an active experience. 

About this newsletter 

Today’s newsletter is written by Jodie Hopperton, based in Los Angeles and lead for the INMA Product and Tech Initiative. Jodie will share research, case studies, and thought leadership on the topic of global news media product.

This newsletter is a public face of the Product and Tech Initiative by INMA, outlined here. E-mail Jodie at jodie.hopperton@inma.org with thoughts, suggestions, and questions. Sign up to our Slack channel.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .