Who Put This in My Attention Cart?
I was thinking about Instagram and LinkedIn recently and realised that they have started to feel a lot like supermarkets.
You walk into a supermarket with a fairly clear list of things you want to buy. You know what you need, you probably know roughly what you want to spend, and you have no intention of buying anything beyond that. But then you start walking through the aisles. Something catches your eye. There is an offer on something you didn’t plan to buy. You see a product that you have never considered before, but suddenly it looks interesting. By the time you reach the billing counter, there are 2 or 3 extra things in your cart.
Nobody forced you to buy them. In fact, you probably didn’t even make a conscious decision to buy most of them. You just happened to see them at the right moment.
That is what I find interesting about social media.
I may open LinkedIn because I want to see what is happening in AI, read something interesting about software architecture, look at what people in my network are working on, or simply spend a few minutes catching up. I have, in my head, a very rough “list” of what I came there for.
But the feed has its own shopping list for me.
I get shown a company’s CSR initiative with photographs of the great work they have done. Then there is an internal company event that I have no context about, accompanied by several posts about what a wonderful time everyone had. Then there is the latest AI post, which may either be genuinely useful or just another piece of AI generated content or company marketing dressed up as something more interesting. There are product announcements, leadership lessons, motivational posts, celebrations, offsites, awards and an endless stream of “excited to share” updates.
There is nothing inherently wrong with most of this content. Someone may genuinely find it interesting. The CSR initiative may be meaningful. The company event may have been a great experience for the people who attended it. The product announcement may be important to a particular customer.
The problem is that the same content is being put in front of me simply because I happen to be connected to someone who posted it.
And I start wondering: Who put this in my attention cart?
The supermarket analogy becomes even more interesting when you think about what the actual transaction is.
In a supermarket, the extra product costs me money. On social media, the extra content costs me attention. And because the cost is not immediately visible, it is much easier to overspend.
I might look at something for 20 seconds. Then another post catches my attention. I spend another minute there. I read a couple of comments. I move to something else. Before I realise it, 30 or 40 minutes have gone by.
None of those individual decisions felt particularly significant.
That is probably the most interesting part of the whole thing. Social media doesn’t necessarily need to show us things that we desperately want to see. It only needs to show us things that are interesting enough for us to spend a few more seconds on.
And “maybe this is interesting” is a very powerful mechanism.
We are quite good at controlling our money because we see the transaction. We know that buying something costs ₹500. We can have a budget and make a conscious decision about whether it is worth spending that ₹500.
Attention is different. There is no little notification saying, “You have just spent another five minutes on something you never intended to consume.”
So perhaps we need to start thinking about an attention budget in the same way we think about a financial budget.
If I have a limited amount of high quality attention every day, I would much rather spend it on my work, my family, learning something new, thinking through a difficult problem, or reading something that genuinely adds value. Yet a surprisingly large portion of that attention can disappear into content that I never actually intended to consume.
This becomes even more interesting on a platform like LinkedIn because so much of the content is generated by organisations and their employees.
Consider an employee sharing a company CSR activity. It could be completely relevant to their colleagues, potential employees, customers or people in the local community. But does it necessarily need to reach every professional connection that employee has accumulated over the years?
Or an internal company event. For the people who were there, it probably means something. For someone outside the company who has no idea what the event was about, it may simply be another piece of content occupying space in their feed.
The issue isn't that the content is bad. The issue is relevance.
Get Varun Mehra’s stories in your inbox
Join Medium for free to get updates from this writer.
And relevance is personal.
This is where I think there is a much bigger opportunity for personalisation than what we have today.
Most recommendation systems are essentially trying to answer a question like: “What is this person likely to click on, watch, like or engage with?”
What if we started asking a slightly different question: “Is this actually worth this person’s attention?”
Those two questions sound similar, but they could produce very different systems.
Imagine LinkedIn understood that I am interested in AI architecture, cloud engineering and certain types of technology discussions, but generally don’t care about internal company celebrations or generic corporate marketing. It doesn’t necessarily need to block those posts. It could simply understand that they are low priority for me and keep them out of my primary attention stream.
And there is an interesting extension of this idea for the person creating the content as well.
Suppose an employee is about to share a post about an internal event. Instead of simply giving them a Share button, the platform could provide some context: “This appears to be primarily relevant to your colleagues and people familiar with your organisation. Would you like to share this with your company network or a selected group instead?”
That isn’t censorship. It isn’t telling someone what they can post. It is simply introducing a little bit of friction before something gets amplified to hundreds or thousands of people who may have no reason to see it.
The same principle could apply to companies.
Companies understandably want employees to become advocates. They want people to share product launches, CSR initiatives, awards, events and company achievements. But if everyone shares everything with everyone, the system eventually creates its own problem. The more content that gets pushed into everyone’s feed, the harder it becomes to distinguish the genuinely useful content from everything else.
Maybe companies should care about this as well.
Instead of only measuring impressions, perhaps we should eventually care about something closer to relevant reach. How many of the people who saw this actually had a reason to care about it?
That would require a different philosophy from the platforms.
Today, much of the optimisation is around engagement. More time spent. More clicks. More interactions. More content consumed.
But what if the best personalised feed was not the one that kept me there the longest?
What if it was the one that helped me get what I came for and then let me leave?
That sounds almost counterintuitive for an advertising driven platform, but I think AI makes this question much more interesting.
We are moving towards personal AI systems that understand our context, our interests, our goals and increasingly, our information history. There is no reason such a system couldn’t eventually sit between us and the enormous amount of content being produced every day.
Instead of asking only “What might you like?”, it could also ask “What do you not need to spend time on?”
That distinction matters.
The next level of personalisation may not be about giving us more of what we like. It may be about getting much better at filtering out what we don't need.
Because we don’t really have an information problem anymore. We have an attention allocation problem.
And perhaps the most valuable feature a social platform could eventually offer is not another recommendation, another notification or another piece of content.
It could simply be the confidence that when something does make it into our attention cart, there is a good reason for it to be there.
So, the next time I find myself 30 minutes into a LinkedIn or Instagram session wondering how I got there, I think the question I’ll ask is a slightly different one:
Who put this in my attention cart?
