I Went Looking for Shoes. Instagram Took It Personally.

A few days ago, I needed shoes.

I did what any normal person does when faced with a problem of this magnitude.

I searched.

I looked at a few pairs. Opened some pages. Stayed on others slightly longer than I probably should have. Compared styles. Closed tabs. Continued with my life.

Instagram did not.

By later that day, my feed had undergone a hostile takeover by the footwear industry.

Shoes in Stories.

Shoes between Reels.

Shoes from shops I knew.

Shoes from shops I had never seen before.

Shoes that looked suspiciously like the shoes I had already looked at.

Apparently, I had clicked on a loafer and accidentally filed a formal request for proposals.

At some point I realized something.

I wasn't really shopping anymore.

The shopping was happening to me.

And that, I think, is far more interesting than the shoes.

We still imagine shopping as something we initiate

The traditional idea of shopping is built around intent.

I decide I need something.

I go somewhere that sells it.

I look.

I compare.

I buy.

The internet initially made the same process faster.

Instead of walking into five stores, I could open twenty tabs and destroy an afternoon with unprecedented efficiency.

Google made this even more direct.

You type:

men's black loafers

and Google essentially says:

Excellent. Here are enough black loafers to make you reconsider the concept of footwear entirely.

But social platforms introduce something different.

They don't always wait for the fully formed search.

They operate somewhere between what you have done, what you appear interested in and what advertisers think you might do next.

So the journey starts changing from:

I want → I search → I find

to:

I signal → the system interprets → something finds me

That sounds like a small distinction.

Commercially, it isn't.

Maybe Instagram isn't a social network with shops inside it

Maybe we've been describing this backwards.

Maybe, increasingly, Instagram is a shopping environment with a social network wrapped around it.

Particularly in Albania, where the distance between seeing something and trying to buy it can be remarkably short.

And the numbers make that observation harder to dismiss.

In Albania, 42.4% of people aged 16–74 reported making an online purchase during the previous 12 months in 2024, according to INSTAT. Among recent online buyers, clothing, shoes and accessories were by far the most common category, at 90.3%.

So apparently I am not even original in my shoe problem.

More interestingly, Instagram itself has enormous reach in the country. Data published through Meta's advertising tools and compiled by DataReportal put Instagram's potential audience in Albania at around 1.35 million people in late 2025, equivalent to roughly 61% of adults aged 18 and over.

And the OECD reports that nearly 41% of surveyed Albanian SMEs use social media platforms for online sales, compared with about 45% using their own websites and only around 10% using third-party marketplaces.

Read that again.

Not for posting.

Not for awareness.

Not for “community building,” that magical phrase we deploy whenever we cannot quite explain what the social-media budget is doing.

For sales.

That's where this gets interesting.

In Albania, the DM is sometimes the checkout

If you've shopped from smaller businesses here, you probably know the ritual.

You see the product.

You open the profile.

You send a message.

How much?

They answer.

Size?

You answer.

Address?

You send it.

Two days later somebody calls because they are downstairs.

This would probably cause a Silicon Valley e-commerce consultant to develop a nervous twitch.

Where is the optimized checkout?

Where is the conversion funnel?

Where is the automated nurture sequence?

Where, for the love of God, is the abandoned-cart email?

Often there isn't one.

And yet the transaction happens.

That is important because businesses have a habit of confusing how commerce is supposed to work with how people actually behave.

Customers are less emotionally invested in our channel architecture than marketers might hope.

They don't wake up wondering whether today's purchase will be classified as social commerce, e-commerce, conversational commerce or omnichannel retail.

They wanted a shirt.

They saw a shirt.

They bought the shirt.

The taxonomy meeting can happen without them.

The algorithm is becoming part of the storefront

I don't know which exact signal caused Instagram to decide I needed more shoes.

And that's worth saying.

The fact that an ad appears after you browse something does not automatically tell you precisely which piece of data produced that particular ad.

Meta says it can receive information from advertisers and partners about activity on websites and apps, including visits and purchases, alongside signals from how people interact with Meta's own products.

And despite the enduring theory that our phones spend their evenings eavesdropping on us like badly behaved relatives, Meta says it does not use the microphone to listen to conversations for ad targeting; microphone access is used when people actively use features that require it.

Which, frankly, makes the reality more interesting.

The platform does not need to read your mind.

Human behaviour leaks clues everywhere.

What we click.

What we watch.

What we ignore.

What we buy.

What we return to.

What similar people appear interested in.

Prediction doesn't have to be supernatural to feel supernatural.

It only has to be good enough.

Search asks what you want. Recommendation tries to guess.

This, to me, is the bigger transformation.

For years, digital commerce was largely organized around search.

Search is wonderfully explicit.

The customer announces the problem.

I need running shoes.

The marketer's job is to appear somewhere near that declaration.

Recommendation changes the sequence.

The platform can put the running shoes in front of someone before they type “running shoes.”

That shifts commercial competition upstream.

You're no longer competing only for:

the purchase.

You're competing for:

the formation of preference.

That's a much more interesting place to be.

Because by the time I type a very specific product into Google, quite a lot may already have happened inside my head.

I may already know the style.

I may already have seen the brand.

I may already have decided what “good” looks like.

I may simply be looking for somewhere to buy the decision I made three scrolls ago.

So perhaps the feed is the new high street

Retail used to obsess over location.

You wanted the corner with the traffic.

The street people walked down.

The part of town where the right customer passed your window.

Digital advertising did not remove that logic.

It changed what location means.

Today, one of the most valuable pieces of commercial real estate available to a business can be approximately six inches tall and currently sitting in someone's hand.

The feed is a street where every shop window can rearrange itself depending on who walks past.

My Instagram does not look like yours.

Your storefront may exist prominently in my version of the street and be completely absent from somebody else's.

That means media buying is increasingly doing something retail landlords used to do:

deciding who gets the good location.

Except the location now moves with the customer.

Slight improvement over Via Montenapoleone.

The rent can still be painful.

But this does not mean “just run more ads”

This is usually where a perfectly interesting observation gets murdered by marketing.

Someone concludes:

Therefore, businesses should leverage social media advertising to reach consumers throughout the customer journey.

A sentence so lifeless it should come with a certificate.

No.

The lesson is not simply to spend more money.

Because there is one fairly inconvenient fact about sophisticated targeting:

It can deliver bad advertising with extraordinary precision.

You can now annoy exactly the right person.

At exactly the right time.

On exactly the right device.

Progress.

The platform can improve the probability that I see you.

It cannot make me care.

That still depends on things algorithms cannot solve on your behalf:

the product.

the idea.

the image.

the price.

the proposition.

the reputation.

the timing.

the experience.

the trust.

And, increasingly, whether buying from you feels easier than continuing to scroll.

Better targeting makes bad creative more obvious, not less important

There was a period when digital marketing seemed to encourage the idea that creativity might eventually become secondary.

Data would do the serious work.

Optimization would find the customer.

Automation would decide the placement.

Performance dashboards would tell us what worked.

And creative would presumably arrive at some point with the logo.

But when targeting technology becomes widely available, targeting itself becomes less of a competitive advantage.

Everyone gets access to increasingly intelligent machinery.

The question becomes:

What are you feeding the machinery?

If ten shoe brands can reach me after I demonstrate interest in shoes, the winner is not automatically the one with the cleverest targeting settings.

It's the one that gives me the strongest reason to stop.

Technology can make distribution more intelligent.

It cannot make mediocrity interesting.

The website isn't dead either

There is another easy overreaction here.

If Instagram can drive discovery and even facilitate sales, why bother with a website?

Because discovery and ownership are not the same thing.

Social platforms are extraordinary rented spaces.

But they are still rented.

Algorithms change.

Accounts disappear.

Reach moves.

Features arrive.

Features die.

Yesterday's essential social format becomes tomorrow's button nobody remembers using.

Your own website performs a different job.

It can provide depth.

Search visibility.

Credibility.

Structured product information.

First-party relationships.

A permanent body of content.

And, importantly, a place where your identity is not sitting between a dancing dog and someone's holiday in Mykonos.

For serious brands, the smartest model is probably not Instagram or website.

It is understanding what each part of the system is good at.

The feed creates discovery.

Creative earns attention.

The brand creates confidence.

Commerce removes friction.

The product decides whether any of this was worth doing.

That is a system.

Posting three times a week is not.

Albania is interesting here because the behaviour is already ahead of the terminology

The OECD data comes with an important qualification.

Although Albanian SMEs are using websites and social platforms for online selling, online commerce still represents a relatively small share of total revenue for many firms. The OECD describes adoption as increasing but still limited in sophistication.

That's useful because it stops us declaring that every Albanian business has become Amazon after opening an Instagram account.

They haven't.

What we have instead is something messier.

And probably more important.

Consumer behaviour has adapted faster than some of the infrastructure around it.

People are perfectly comfortable discovering products socially.

Businesses are perfectly comfortable selling conversationally.

Formal e-commerce systems are developing alongside those habits rather than replacing them overnight.

That means Albania doesn't necessarily need to reproduce every stage of e-commerce exactly as larger markets experienced it.

Some businesses may move directly into models where the boundaries between content, conversation, recommendation and commerce are barely visible.

Which brings us back to my shoes.

I searched once. The market kept searching for me.

That is the shift I find most interesting.

The old internet waited for us to ask questions.

The new one is becoming increasingly interested in answering questions we have not quite asked yet.

Sometimes badly.

Sometimes unnervingly.

Sometimes with twenty-seven pairs of loafers.

But commercially, the direction matters.

The future of shopping will not simply be about making products easier to search for.

It will be about becoming better at appearing at the moment a vague interest begins turning into a decision.

For brands, that changes the challenge.

You need to be discoverable before you're searched.

Recognizable before you're compared.

Trustworthy before you're contacted.

And interesting before a thumb moves another centimetre.

Because the most valuable shop window may no longer be the one on the busiest street.

It may be the one that appears just before the customer realizes they were shopping.