Pakistan's startup space has become practically meaningless

A host who is tired, and an opening confession

The episode opens with Muzamil unusually subdued. He tells the audience directly that he is starting the conversation in a depressing register, and that the reason is personal. He has spent years rooting for Pakistan’s startup ecosystem to mean something — partly out of computer-science loyalty, partly out of the hope that the country might do for services what India did. What he keeps running into instead, he says, is fraud. “I am a little tired,” he tells his guest, before introducing Mutaher Khan, founder and CEO of Data Darbar, a Dawn columnist, and head of the Insight Lab at the Karachi School of Business and Leadership.

Muzamil names the texture of his frustration carefully. Generic businesses in Pakistan are rent-seeking — someone holds a license, someone took a government contact and turned it into a concession, there is no innovation to cover. Services, particularly IT services in Lahore, were supposed to be different. He looked at one company closely. What he saw was social-security-number fraud being justified by a Sharia board. “Yaar, yeh ji chuna hai ki ek parhi likhi cheez thi, woh bhi out hai” — even the educated part of the economy turned out to be a con. So he wants Mutaher in the hot seat to answer a specific question: where is the journalism? Why does the person whose role it is to surface this, not do it?

The state of the economy, the state of the startup

Mutaher answers in two registers. On the headline numbers — the budget, fiscal management, the official story — he says things look more or less under control, oil shock notwithstanding. On main street, “there is blood.” Inflation has not actually fallen, only the rate of change has slowed. Purchasing power is in clear decline. Manufacturing volumes are down even where value looks fine on the back of pricing.

Then he draws the distinction that frames the rest of the conversation. There is the service economy — bigger, less talked about. And there is the startup space — much smaller, much more talked about. The startup space, he says flatly, “practically ab meaningless ho gayi.” That is why Data Darbar, which began covering startups, eventually had to broaden into digital economy and financial services at large. Startups were never large enough as a share of jobs or revenue to justify the airtime they got. What they had was narrative — startups were sexier than service businesses, and funding rounds got press regardless of revenue.

The asymmetry he describes is exact. A company doing real revenue could not get into an international publication. A company with a five- or ten-million-dollar round could. “Regardless of the revenue that you are doing,” he says, “usme koi newsworthiness nahi lagti media ko.”

Why VCs broke the expectations market

Mutaher’s diagnosis of the VC problem is precise and worth quoting. Pakistan, he argues, has spent the last seven or eight years in what may be the longest stretch of macroeconomic turbulence in its history — historically the country has lived in boom-bust cycles, but the gaps between them have collapsed. Booms got shorter; busts got longer. Some of that was external — the Russia-Ukraine grain shock, commodity super-cycles, the Iran war — and some of it was self-inflicted. Either way, no business operating in Pakistan got eight years of breathing room. Layered on top of that absence of oxygen was a VC class whose expectations were, in his word, “outrageous.”

He recounts a conversation about a US-based company raising thirty million dollars. The investor in the room said it was “not that big a number.” Mutaher’s point: when investors in a country whose total forex reserves sit at roughly eight billion dollars are setting thirty-million-dollar rounds as the floor, founders inevitably bend toward those expectations. Pakistan, he says, is an investor market, not a founder market. Five sources of capital decide what scalable looks like, and the deals follow.

Muzamil sharpens this further. He started interacting with the Pakistani VC scene nine or ten years ago, when Plan 9 was the marquee operation. What he saw, in retrospect, was that many of these investors lacked self-awareness. “These were mostly middle management in international companies who cosplayed investors.” Operators with bootstrapped, real-world models — he gives the example of a contractor approaching Plan 9 — were told they were idiots for not chasing the high-risk, high-narrative idea. The result was a cultural lock-in: the same diaspora that already loved doubling money on plot files now had English-speaking suits telling them startup gambling was the upgrade.

The Zameen problem, and the refusal to claim a unicorn

When Muzamil asks why, after all this, no real success story has emerged, Mutaher offers a sharper answer than the question expected. “Part of the problem, Muzamil, is storytelling itself.” He picks a specific example: Zameen, founded in Lahore in 2006, funded around 2011 or 2012, expanded internationally as Emerging Markets Property Group, became a unicorn in 2020. By every reasonable definition, Pakistan’s first.

Muzamil pushes back. The money, he says, was not really about Zameen — it was about high-margin markets the company had moved into after leaving the country behind. Mutaher refuses to concede the point. He argues by analogy: Uber’s later growth came from Europe, the Middle East, South Asia. Nobody says Uber stopped being an American company. Ola’s expansion does not strip its Indian-ness. “Why is it not a Middle East company? Why is it not a Latin American company or Southeast Asian company? We need to be more accepting of the fact that a company doesn’t have to make all its money in Pakistan to be Pakistani.”

The deeper issue, he says, is structural. Data Darbar itself struggles to classify funding rounds — capital arrives via Singapore or Dubai holding companies, because that is simply how the global system runs. If Pakistan only counts capital that landed in Pakistani legal entities, the entire startup space has effectively received zero equity. The choice to look past Zameen, in his telling, is a choice to look past a story Pakistan should be claiming.

The journalism gap, and why it never closed

This is where Mutaher’s argument becomes the most uncomfortable. Pakistani business journalism, he says, has never been attractive as a career. Journalism in Pakistan has historically meant opinion writing — editorials, columns — not reporting. There have been good reporters, he acknowledges, but the reporting muscle has never been deep. Dawn is the paper of record and a policy-stance paper; in-depth reporting exists but is not recurring.

Then comes the financial-incentive layer. “There are media organizations in Pakistan, large media organizations — not startups — that probably pay less than a Unilever peon. And this is not against the peon.” Reporters who have spent fifteen years in the system are not earning a living wage. Naturally, the incentive to chase a startup story dies. Status quo wins: write the column, get the four quotes from the same friends you have known for ten years, file it, go home.

Muzamil layers his own experience on top. He recalls the death of Airlift — a story he believed any half-functional business desk should have eaten. Nobody, he says, even messaged its former employees on LinkedIn for an anonymous comment. He had founders telling him things off-record that no journalist ever surfaced. “The amount of chuna jo aakar mujhe CEO mere muh par lagata tha” — the brazenness of the lies he heard in his own seat — was, in his telling, evidence that the seat the journalist should have occupied was simply empty.

Mutaher adds the last piece. Pakistan’s business reporters are concentrated in print, the web is staffed by desk editors and aggregators rather than reporters, and the pipeline of new business journalists has been thin for over a decade. The same five people carry the load until they get bored.

The two contradictory AI reports

The conversation pivots when Muzamil flags two reports he had read that day. India’s IT stocks are down roughly twenty-five percent and sitting at a three-year low, with the market pricing in an AI hit to the big five outsourcers. On the same day, an IBM report argued that India is prime for the AI revolution and could capture up to five hundred billion dollars annually in IT services value — more than double the current run-rate of the sector. Two opposite signals. He wants Mutaher to read what that means for Pakistan.

Mutaher begins where Pakistan begins: small. The country’s entire IT services industry, a few years ago, was the size of one Indian giant’s fresh-hire intake. That smallness has historically meant incremental growth was not the constraint. Pakistan is now somewhere around four and a half billion dollars in measured IT exports, and Mutaher suspects the real, including-unreported number is closer to nine or ten.

Then he makes his core call. “AI, I think, will get you to eighty percent of any task — pretty much any task. The real value add is in that remaining twenty percent. The question is whether Pakistan’s average engineer can do the twenty percent on top of AI.” Firms that can, will do well. Firms running pure BPO models will struggle. He sees evidence of the first pattern around him: companies he knows, who started recently, struggled with cash flow, and then the AI boom gave them more money than they could spend right now. Not lottery numbers, but enough buffer to start product development for the first time — to do R&D where previously they only sold time.

The micro-agency play and digitally deliverable services

The more interesting bet, Mutaher argues, sits one layer down — in micro-agencies. Pakistan’s freelancing economy has a well-documented valley of death: a freelancer who tries to institutionalize, hire, and turn into an agency tends to fail because hiring costs outrun the revenue ramp. AI flattens that. A six-thousand-dollar-a-month operator can plausibly reach thirty or forty thousand without the same headcount.

He widens the lens further. The relevant category is not IT services — it is digitally deliverable services. Does Pakistan have an accounting skill set? Yes. A legal skill set? Yes. Management consulting? Less so. The structural opportunity is that several of these professions require a human licensing layer that AI cannot legally cross — an audit needs the auditor’s stamp, legal advice needs the lawyer’s standing. Y Combinator, he points out, has been explicitly calling for AI-native service firms. If Pakistani firms can establish footprint and clear the licensing requirements in foreign jurisdictions, the play is significant.

Muzamil agrees and pushes it further. The biggest hit in the next few years, in his read, is the entry-level IT worker — because anyone with even moderate experience can become a manager of agents. He sketches a worked example: a former Pakistani standard lawyer who later worked in fintech could now spin up an outbound agent that cold-calls small and medium businesses globally and offers five-hundred-dollar legal advisory against a market rate of fifty thousand. Same for accounting. The cost gap, against the global SMB ceiling on what is affordable, is the opening.

He closes the AI section with a story from Seattle. He had asked a Pakistani vice president at AWS what was actually happening on the ground. The answer, paraphrased: lifestyle businesses, ten to twelve employees, all of them rockstars, average per-employee revenue of two hundred to two hundred and fifty thousand dollars. With a team in Pakistan, where ten thousand dollars a month is a delighted senior salary, the unit economics become absurd. The fear, Muzamil argues, is masking the opportunity.

What 2050 looks like, and the family-office play

Muzamil closes by asking Mutaher, at thirty, what he actually thinks Pakistan looks like in twenty-four years. Mutaher resists the easy answer. He concedes that information overload makes a Pakistani analyst gloomier than the facts warrant. But he points to a Vietnam analogy — a country whose trajectory bent meaningfully only after 2001 or 2002. “Generally it can take moments to just change your trajectory,” he says. “It might be the most pessimistic way of being optimistic. But there are certain right ingredients.”

Three he calls out. First, Pakistani entrepreneurial culture has genuinely improved — not VC-style culture, but the lifestyle-business kind, where people want to build something that pays for a house, a car, and a holiday. Second, an export orientation is finally entering the conversation. Third, and most concretely, the country’s existing capital pools — family offices, established business groups, the seths — already have the money to go and buy AI companies abroad and position Pakistan in the application layer without building anything from scratch. Post-2023, he argues, those families have realized that the state they used to seek rent from is not reliable enough to keep doing it; they need alternatives. Acquisitions could be that alternative.

Muzamil overlays his own evidence. The Lakhanis have just set up a services company under the Fatima Group; Jazz has announced an export-focused services arm. The same families who used to operate exclusively in generic, rent-seeking businesses are now positioning for export. He reads a report comparing Pakistan’s real exchange rate over the last twenty years to its peers. Asia’s average was 91. China kept theirs at 85 or 86. Pakistan’s twenty-year average was 116 — meaning the rupee has been structurally overvalued for two decades. His hope is that once the Fatima Group lobby and others like it have skin in the export game, the policy conversation will finally bend toward controlled depreciation instead of the three-year cycle of artificial control followed by free-fall.

He ends on a note that recurs in his work and that he names directly. He does not want to antagonize anyone. He does not want to name names while investigations are running. But he believes Pakistan’s culture of censorship and self-censorship does more harm than good. “I have self-censored quite a lot in my life, and I feel like it just ends up doing a lot more harm because the cycle keeps repeating.” Every two years, he says, the same story arrives in a new shell. Mutaher’s reporting, and conversations like this one, are how that cycle gets named out loud.