Pakistan does not have 240 million customers — and we made that number up

A Dubai studio, a missing hour of footage, and a guest with three businesses

The episode opens with Muzamil flagging, on camera, that something went wrong with his own footage during recording. He recorded the conversation at the Dubai studio with Raza Matin — founder of PayPro, Brandverse, and Chikoo — and discovered afterwards that part of his side of the shot was missing. He apologises on the record, asks viewers to bear with the editing, and moves on. It is a small moment, but it sets the tone of the whole conversation: matter-of-fact, no spin, no performance.

Raza was born in Dubai, spent his first six years there, moved to Oman for the next eleven or twelve, and then went to Canada at seventeen or eighteen. He studied economics at Queen’s University in Eastern Ontario. He has never actually lived in Pakistan for any extended period until he moved there in 2012. Muzamil notes the distinction directly — Raza did not move back to Pakistan, he simply moved there for the first time as an adult. Raza agrees.

When Muzamil asks what the relationship with the country was like growing up, Raza pushes back gently on the romantic version of the diaspora narrative. “I don’t think if you actually ever live here, you get a true appreciation for the nuances and you build that strong relationship with the culture, the people, the cities,” he says. “I think people overstate your connectivity with the country based on these small interactions or these short interactions. I don’t think you truly build that part of the identity.”

How Raza and Tania actually got YouTube unbanned

Before the businesses, there is the Google chapter. Raza spent close to five years as the country consultant representative for Google in Pakistan, starting around 2015 — roughly a year before YouTube was unbanned. He calls his contribution to that effort “one of my finer contributions to the country.”

The reason YouTube was banned in the first place, he explains, was a stalemate. The Government of Pakistan wanted oversight on content — including the discretion to take down content at a global level, not just in Pakistan. That was incompatible with Google’s policies. Neither side would move. Politically, nobody in government wanted to be seen as giving Google permission to do as it wished.

The fix, Raza says, was internal to Google, and credit for it goes “nearly entirely” to a colleague named Tania. The team had to make the case inside Google that Pakistan was worth proactively building a localised content lens for — something that required engineering work, product decisions, and many layers of stakeholder buy-in. Slides were not enough. So they flew Google’s senior leadership to Pakistan for a four-day “whirlwind, whistle-stops tour,” meeting government officials, business titans, creative artists, students, entrepreneurs, and startups. The argument they made was not only commercial. “At the end of the day, Pakistanis deserved the opportunity to learn and benefit from things like YouTube, just like everyone else.” It worked. Raza says the newspaper announcing the unban is framed somewhere inside his house.

”We made up the 240 million” — the most honest line in the conversation

When Muzamil asks how important Pakistan really is to global tech platforms, Raza gives one of the sharpest answers in the episode. He is willing to name the narrative for what it is.

“That for that narrative — the 240 million people, one of the youngest populations in the world, one of the lowest median ages in the world, highly connected or increasingly connected population, high proficiency in English relative to a lot of other markets — that narrative was unashamedly created by myself and my colleagues at Google when we were trying to sell Pakistan internally to get a lot of our product issues and challenges sorted.”

The narrative then propagated. It became folklore. Raza is not apologetic about having created it — it was what was needed at the time — but he is very specific about what is actually realisable underneath it. The headline number is 240 million. The actually-addressable population, in his read, is the subset that is connected, has disposable income, and has the digital literacy to transact: somewhere around 60 million people. That is the size of the United Kingdom. It is still a serious market. But it is not 240 million, and pretending it is leads founders and investors into bad decisions.

He adds a second layer. Those 60 million sit inside 240 million, distributed across a very large geography with many cultural nuances. So even where the demand exists, the distribution, sales, and marketing challenges of reaching it are complex. “Is the opportunity there? Absolutely. Is it as large as we would like to think or others would like to have us think? No.”

Parasitic technology versus beneficial technology

Muzamil asks why Raza left Google in 2020. The answer is partly COVID, partly the question of what he would leave behind. “If you were to die and people were to look back on your life, what would they say about you? Have you left footprints in the sand or have you left mountains?” He did not want his legacy to be “a bunch of reports or some advertising sales numbers.”

What he took from Google was a particular thesis about technology: that it can be a force of uplift, not just economically but socially, and that this had not yet been done well for Pakistani businesses. He frames the prevailing relationship Pakistani retailers have with technology as parasitic. The example he gives is the card-acceptance machine sitting on a counter.

“Go to a retailer today, and he will have this credit card machine that sits on his counter. Ask him how much that credit card machine costs in his business. He’ll say 1.6%, 1.8%. Ask him what he gets in return. He’d be like, I get the money in my account two days later. So it cost him 1.8% of a transaction to get the money two days later, and ask him what was the benefit of all of this. He would say, I don’t know.”

That, in Raza’s framing, is what parasitic technology looks like. It charges, but it does not help the merchant find new customers, increase the average sale, increase profitability, reduce losses, or manage inventory. The mission he set himself when he left Google was to figure out what beneficial technology for a Pakistani merchant would actually have to do.

PayPro — and why the rails existed years before anyone used them

The first business in that lineage is PayPro, which Muzamil notes preceded Brandverse and Chikoo sequentially. PayPro grew out of an earlier venture Raza’s father had started, which over time was reshaped into a fintech. The product, in plain language, takes a bill from a merchant or business, sends it to a consumer through whichever channel — SMS, email, WhatsApp — gives the consumer the option to pay digitally through card, debit, banking app, EasyPaisa, or JazzCash, and reconciles the transaction instantly. Two days at a bank counter becomes two clicks.

The point Raza wants Muzamil to sit with is not the product. It is the timeline. The rails PayPro was built on — specifically the 1LINK bill-presentment rails that allow a third party to send and receive transaction confirmations programmatically — existed for three, four, maybe five years before PayPro was founded. Nobody had used them. “I think it goes to my point about the fact that we need to build business solutions, not services. The same services that we built PayPro on existed for maybe three, four, five years before we did it.”

The implication is uncomfortable for the Pakistani fintech narrative. The bottleneck has not always been infrastructure. Sometimes it has been the absence of a founder willing to package what was already there into something a small business could use.

Muzamil pushes on the question that has dogged Pakistani fintech for a decade: why is digital onboarding for SMEs so weak? Why does every product get sold to “large, medium and large” rather than the small e-commerce seller or the social-commerce operator running live sales on Instagram? Raza’s answer is honest and unflattering to the industry he is part of. “Because we aren’t truthful with ourselves.” Products get created in a vacuum, then audiences get retrofitted to them. PayPro itself, he admits, had this problem — the original product required a level of digital hygiene (a CSV of students, a list of parents) that informal businesses simply did not have. Four years later, an individual with a registered sole-proprietor business and a national ID card can onboard in twenty-four to forty-eight hours and start accepting card, debit, wallet, and bank transfers. Informal businesses — a girl in Sahiwal selling embroidered clothes from her Instagram, in Muzamil’s example — are still out of reach, and Raza is direct about why: the payment schemes do not allow it, and PayPro does not yet have the marketplace-model intermediary structure to take responsibility for the KYC.

Brandverse — the boring infrastructure of Pakistani retail

Brandverse, the second business, is where the conversation gets genuinely interesting from an infrastructure point of view. Raza describes it as having two halves. The first half is, in his words, “the only at-scale creator of ecommerce content in Pakistan.”

The picture and information attached to a bottle of water on Carrefour’s website did not appear by magic. Someone photographed it. Someone wrote the name, the description, the ingredient list. Someone made sure the same product appears consistently across the three hundred other retailers selling it. Over five years, Brandverse has built what Raza calls Pakistan’s largest aggregated database of high-quality product information, structured to the GS1 global supply-chain standard. “If you give me a product, I can tell you based on its barcode, its ingredients, place of origin, height, size, width, depth. I can tell you its brand name. I can tell you where it was manufactured. I can tell you all of this information. I can even tell you how it’s sold in terms of bulk packaging.” Bottles of water and cans of soup were the starting point. The same content factory now serves some of the largest fashion brands in Pakistan.

Muzamil challenges him on the obvious failure case — Daraz, whose product metadata he calls “trashy” on the record. Raza does not flinch but he is fair. The reason marketplace metadata is bad, he says, is that creating good content is genuinely hard for the seller. “You’re asking the person who typically sits in a shop to take beautiful photographs, write fantastic descriptions with detail, do it in English, and provide all the information you’d find on the physical box. The answer is because it’s hard. That’s why it’s not there.” Amazon’s metadata is better because Amazon enforces high standards, but the trade-off is fewer sellers. The fix, Raza argues, is structural — give the seller a barcode scanner inside their phone and have the title, ingredients, descriptions, and dimensions fetched from a referenceable database. The seller types nothing. The platform gets clean data.

The language of commerce

The metaphor Raza reaches for to explain why this matters is one of the cleaner pieces of thinking in the episode. He compares the structured product database to what Google Maps did for location.

“Much like what Google Maps does for location — it takes those weird little numbers, those GPS coordinates, and translates them into something more meaningful. It translates them into the name of a place, a name of a street, a particular location you’re on. And what Google Maps has enabled are innovations like Careem and Uber. They’ve enabled Foodpanda, for example. They’ve taken many disparate actors and connected them together using the language of location.”

A barcode plus structured context, in his framing, can be the equivalent for commerce. Once retailers, marketplaces, dark stores, kirana shops, and delivery aggregators share a common product language, new businesses become possible. The example he uses is hyperlocal: instead of a bottle of Coke coming from a dark store three kilometres away in five minutes, the same bottle can be sourced from the kirana shop twenty metres down the street in one minute, picked up by Chotu and delivered. That kind of routing only works if every actor in the chain references the same product the same way. “But that’s only possible when you start with creating the language. And that’s what we’ve done at Brandverse.”

The Foodpanda Shops example Muzamil raises is the same problem in a different costume. Grocery shops with low margins, huge SKU counts, and prices that change every time a distributor walks in were a logistical nightmare for Foodpanda to onboard one by one. Their answer was to retreat to dark stores. Raza’s argument is that a shared product language with API-level price-fetching and a margin call on top makes the small kirana addressable again.

Chikoo — an operating system for the merchant, embedded inside the card machine

Chikoo is the third and most ambitious piece. Raza describes it as the packaging layer — the database is the chemical, Chikoo is the pill the merchant can actually swallow. The original idea, he says, was “Shopify for this part of the world.” But as his team worked closely with merchants, the list of needs widened: financial reporting, inventory management, a point of sale that ties together walk-ins, ecommerce orders, Instagram and Facebook orders, and Foodpanda orders into one window. Everything in one place, in the palm of one hand.

The product today runs on phone, tablet, browser, laptop, and desktop. And, in what Raza calls the announcement that brought him to this conversation, it now also runs on the Android card-processing machine — through a partnership with Bank Alfalah. “The device that’s sitting on so many retailers’ counters today and not doing anything other than processing payments and churning out receipts can do so much more,” he says. Chikoo embedded on that device turns it into an inventory system, an ecommerce front end, and a point of sale all at once.

The consequence of that embedding is one of the more arresting passages in the conversation. When a typical retail card transaction happens, the bank sees an amount, a timestamp, a merchant ID. Nothing about what was actually purchased. With Chikoo’s line-item data married to cardholder identity through a loyalty program or phone number, that changes completely. “If you’re buying diapers, based on the type of diapers you’re buying, I can tell what the age of your child is. I can also tell you whether your child is sick at any given point in time.” He acknowledges, on the record, that this is “slightly creepy but very powerful” — and exactly the kind of marketing signal that has historically been unavailable in Pakistani retail.

Muzamil asks why Bank Alfalah and not JS Bank — the institutional investor on Brandverse’s cap table. Raza’s answer is operational. JS Bank does not yet have an acquiring business; it is building one. Bank Alfalah, by contrast, has been one of the largest acquirers in Pakistani internet payments since the early days of card machines in the country. For a payments product, that footprint matters.

Why he did not raise from VCs — and why VCs are doing God’s work anyway

Brandverse has raised roughly $2.5 million to date, largely from institutional investors in Pakistan, with JS Bank as the named anchor. Muzamil presses him on why he did not go to the wave of Pakistan-focused VCs that emerged in 2020 and 2021 — Indus Valley Capital, Zayn Capital, and others.

Raza’s answer is one of the more self-aware lines in the conversation. “I don’t think as entrepreneurs or a business we are a good fit for VCs. The reason for that is we are very hyper-realistic in our outlook, myself and Faizan, my co-founder. We find it very difficult to lie to ourselves or give into delusions of grandeur. So given that, it’s very difficult for us to spin a fantastical tale that — I understand why people want it. We’re just unwilling to give it.”

He is at pains not to dismiss venture capital. He calls what VCs do “basically God’s work” — backing ambitious founders with the realisation that the money could go to zero. He just thinks the model is not for every Pakistani company, and certainly not for a slow-burn infrastructure business serving a market whose understanding of the product is itself still developing.

The blow-up: Swvl, Airlift, Carfirst, Dastgyr — and a defence of failure

Muzamil walks him through the Pakistani startup bloodbath of the previous eighteen months — Swvl, Airlift, Carfirst, Dastgyr’s redesign, the broader pullback. He asks what went wrong, carefully framing the question to avoid trashing the founders.

Raza is generous and clear. “I don’t think it was necessarily the fault of the entrepreneurs. Their responsibility is to dream big, to try to change the world. So they did, and people bought into the idea, backed it. Good for them. The only challenge was that the expectations were unrealistic. They were disconnected from the realities on the ground.” Money, in his framing, is not just an amount — it is also an expectation about how much comes back, how, when, and how quickly. The wrong expectations on the wrong timelines force founders to do things they should not be doing. “Growth hacking sounds cool, but it also sounds very dangerous. Hacking doesn’t always work out.”

His second point is the one Muzamil seems most struck by. Failure is not the disaster the Pakistani ecosystem treats it as. It is the mechanism through which the ecosystem learns. The right response is not to disappear and emigrate but to publish honest postmortems — written by the VCs and founders themselves, not by investigative journalists. That credibility is what makes the lesson stick.

The collaboration question that runs alongside this is just as bleak. “Based on the number of MOUs that are signed in this country, you would assume collaborations are happening left, right and centre. But if you really look past the MOUs, doesn’t really seem like much is happening.” He attributes the failure of data-sharing partnerships in Pakistan — the precondition for the credit-scoring revolution Muzamil keeps hoping for — to trust, possessiveness, and a culture of signing instead of delivering. His blunt instruction on data hoarding is worth quoting: “Why over-index on protecting your data when you don’t know what the quality is? You’re just assuming it’s very good. Put it out there. Test it. If you keep it to yourself in the hope one day that you’re going to leverage it to some benefit, we’ll still be waiting. Today is 2023. We’ll be 2030. We’re still talking about all the data we have and the fact we’re doing nothing with it.”

Pakistan at 2050 — and a father’s hope for his child

Muzamil closes with the question he asks every guest: what does Pakistan look like in 2050, twenty-six years out? Raza’s answer is careful. He thinks Pakistan will be stronger, hopefully more economically settled, hopefully more at ease with itself socially. He reaches for the framing that the country is, at 75-odd years old, still a baby in country terms — a very large country with a complex cultural fabric, repeatedly finding itself in the same quagmires not because it has failed but because the path through is genuinely long. “Maybe we are where we were supposed to be, because this is one of those paths that we have to cross towards where we’re going to be.”

He does not promise he will still be in Pakistan in 2050. But he is clear about the test. “I fundamentally hope it is a place that my child calls home.”

Muzamil closes the conversation by acknowledging that he had cooled on doing podcasts at all — Pakistan’s mood, the world’s mood, the sense that nobody gives much of a damn about startups when broader social anxieties are louder. He thanks Raza for restoring some appetite for the format. The recording wraps quickly. Raza has another meeting to get to. The promise is taken for a round two.