Pakistan's IT growth is flatlining — and the policy is making it worse

The slogan is loud, the growth is flat

The episode opens with Muzamil framing the conversation against the noise of the moment. The slogans are everywhere — ten billion dollars of IT exports, five-G, freelancer growth, skill centres — and a new IT minister is moving visibly. The question Muzamil puts to Syed Ahmad, recording at the TBT Dubai studio, is whether any of that is realistic or whether it is mostly photo opportunities.

Syed’s answer is careful. There is a real shift in attention. Back in 2015, when he and others coined the “Digital Pakistan 2020” slogan, no politician was paying attention to IT. Today IT is on the agenda of the Special Investment Facilitation Council itself. “In that way it’s good,” he says. “You feel like top-level support has started to come.” But the underlying machine is not moving. “Last year we have been on flat growth,” he says, “and internationally — America’s economy has been down, venture capital funding has dried out. So there’s an international scenario, there’s a Pakistan situation. We don’t have much growth right now.”

Muzamil notes that this is not the story the country is being sold. Syed does not soften it. The often-quoted nine percent export growth, he points out, looks impressive only because it is being measured against a minus-half-a-percent base from the year before. The real comparison is to 2020–2022, when the industry grew roughly seventy percent across two years, and exports posted a 47% jump in a single year — the highest export growth recorded for any Pakistani industry. From that line, the current numbers are not a recovery. They are a flat plateau the country has been stuck on while its policy makers congratulate themselves.

Brain drain, and why he refuses to call it the real problem

Muzamil asks about the brain drain — the senior IT talent moving abroad in numbers nobody seems to be counting honestly. Syed confirms it from inside his own company. “Every company, my own company — every month you get a few resignations,” he says. “Someone has an offer from Portugal, someone is moving with family to Saudi.” He notes that the IT workforce is partly insulated from inflation by thirty-percent salary growth, more than the rest of the economy. But that only makes the comparison sharper: they earn more, and their ambitions become global. He looks at his own alma mater’s five-year-old batches and sees more than half the names sitting outside the country.

Muzamil presses the point further. The faculty is leaving too — and famously, NUST has been bleeding senior teachers because universities cannot raise fees fast enough to match private-sector or overseas offers. The pipeline is leaking from both ends: the talent that exists is moving, and the talent that was supposed to replace it is being trained by a thinning faculty. He then makes a sharper analytical move. In a free market with an uncompetitive rupee, talent leaving is not the disease — it is a symptom. The right worry is not how to stop people from leaving, but how to produce more people to take their place. Dollarising local salaries to retain them, he says, would defeat the entire economic logic of being a low-cost producer in the first place.

Syed agrees, and adds two things Muzamil’s framing does not. First, the expats themselves are an asset only if they remain “pro-Pakistan” — and on this front the country is losing them. “Our expats are not that kind of gung-ho about Pakistan,” he says. “They’re a little disillusioned, and they’re leaving on a bad note.” India’s IT sector, he reminds, was built by its expats. Second, on the supply side, the picture is more nuanced than the public conversation admits. The merit cut-off for IT has now overtaken electrical and mechanical engineering at top universities. One top institution has shut down most of its other engineering tracks to expand IT intake. The accredited universities can produce roughly 50,000 to 60,000 graduates a year — and there are unaccredited institutes on top of that.

The 1,800 interviews to hire 55 people

This is where the conversation lands its sharpest data point. Quantity is climbing. Quality is not. “Last year we hired 55 people,” Syed says. “In order to hire 55 people we had to do 1,800 interviews.” Muzamil immediately catches the implication — the cost of that filtration alone destroys company-level efficiency, and the same story is told by every CEO in the industry. “The quality of the graduate is really bad,” Syed says. “Investors will tell you the same about the quality of founders.”

The takeaway here is structural. The country’s debate is about how many graduates the system produces. The actual binding constraint is what fraction of them are employable. At a hit rate of roughly three percent, you do not have a workforce — you have a triage queue.

Why the WhatsApp dream is the wrong dream

When Muzamil shifts the conversation to investment — what a billionaire walking in today should actually fund — Syed pushes back on the unicorn framing the ecosystem has absorbed from American venture capital. He uses the cleanest possible illustration. “WhatsApp sold for 18 billion dollars and there were 65 employees,” he says. “The money didn’t go to America. It went somewhere — Cayman Islands. Sixty-five lives improved.” Compare that to Infosys or Tata Consultancy Services. They earn 16 to 26 billion dollars a year in services revenue, and to do it they employ 50,000 people — and roughly 60% of that revenue flows out as payroll into middle-class households.

This is the model Pakistan needs, he argues, not the model it talks about. India lifted four million families into the middle class through its services sector. Those four million people then became the professionals — the trained, ethics-bearing, internationally-fluent workforce that the next layer of the economy could build on. Pakistan, by contrast, has not yet produced its services-economy cohort, and the workforce that exists has not yet been put through the professionalism cycle that creates one. “Pakistan needs five to ten companies doing a billion dollars of services revenue,” he says. “That’s how a middle class is built.”

Muzamil agrees but pushes on the practical question: where does the investor actually put the money? You can fund a startup because the founder hands you a problem statement and a thesis. With IT services, the typical pitch is “give me capital and I’ll quadruple it” with no infrastructure around the claim — no office space for the thousand people you would hire, no talent bench deep enough to fill it. Syed’s answer is direct: invest in people. Skill training, post-college education, university capacity. “Any amount of investment in Pakistan’s youth would not be enough,” he says. The country has chronically under-invested here. LUMS is thirty years old and no investor has tried to build the next one in every major city.

The tax regime is squeezing the wrong people

The conversation turns sharper when it reaches taxation. Both Muzamil and Syed converge on the same conclusion from different angles. The salaried class is now carrying a disproportionate share of the national tax burden. Anyone earning roughly half a million rupees a month is paying over a lakh in tax — a level high enough that, as Muzamil notes, it pushes IT professionals to open restaurants and drive Ubers instead of growing in their field.

Syed extends this to the structural decision facing every company. “Registering a company costs you fifteen percent of your revenue in Pakistan,” he says. That includes payroll tax, EOBI, compliance overheads, and the cost of harassment. The arithmetic is unforgiving. If a person can earn a thousand dollars on the books with a Pakistani company or as a freelancer with no payroll tax, no rational actor stays inside the formal system. “I am willing to match him — I will pay you three lakhs instead of one-and-a-half,” he tells Muzamil. “But he says: I don’t want the tax on top.” So the company that does the right thing loses to the company that does not.

Syed’s framing is the one that lands: “Business is like water. It finds its own way.” He has personally told finance ministers and bureaucrats this. He recalls a meeting with a recent finance secretary who, under IMF pressure, was about to add a quarter-percent income tax that would, by his own estimate, raise seven million dollars. “For that seven million dollars you’ll lose 500 million in growth and 25 million in payroll tax,” Syed told him. “They don’t look at any of this. They just think: we will catch them. I have not seen anyone get caught.”

The result, he says, is that Pakistan has become “a competitive economy” — but only if you are corrupt. Ethical companies are choosing offshore offices and running operations from outside. The informal economy is growing. A handful of multinationals and a small set of genuinely principled firms are still paying their full taxes. Most of the rest have made the other choice.

Freelancing as the symptom, not the success

Muzamil and Syed spend several minutes on the freelancing narrative. The official line — that Pakistan has become the third-largest freelancing nation — is celebrated. Syed’s reading is the opposite. “I call it a symptom of an underlying disease,” he says. “It shows that natural industry growth is not happening, so individuals have to do it themselves.” The platforms themselves, he points out, are structurally hostile to agency formation. They want isolated workers they can squeeze on price, not small companies that develop pricing power. The freelancer earning eight hundred dollars today, on this trajectory, will still be earning eight hundred dollars in five years — and the competition for that work will be denser.

He widens this into a critique of the current policy direction. Companies are now declaring their employees as freelancers to evade payroll taxes. Three-hundred shell entities sit underneath single operating companies, structured to look like independent freelance contractors to the system. “The government is unintentionally helping them,” he says, “because supposedly the policy is to support freelancing.” The result is a one-sided incentive structure that punishes the formal company and rewards the informal one.

Ethics, exact, and the courage that is missing

A long section of the conversation deals with what Muzamil calls Pakistan’s moral conundrum in tech. He raises the high-profile exact scandal as a symptom rather than an anomaly — and notes that CEOs of compliant Pakistani IT companies have told him directly that they cannot compete on pricing or talent with players who operate outside the rules. He frames the question every principled founder is now asking himself: do I keep my belief system and get destroyed, or do I abandon it and join the system, because the returns are extraordinary.

Syed’s answer is one of the most quoted-worthy passages in the conversation. He believes that in any society, only one or two percent are genuinely committed to doing the right thing, and another one or two percent are genuinely committed to corruption. The remaining 95% are sheep — they follow whoever is winning. “Our corrupt side has always been one,” he says. “The good guys are sitting in the corner protecting their dignity. They are not willing to fight.” He invokes Lee Kuan Yew’s three qualities for leadership — merit, honesty, hard work — and adds a fourth that he says Pakistan is most short of: courage. “Until a few good people stand up and fight, and maybe even sacrifice, this system cannot be fixed.”

He also notes a feature of the local culture that compounds this: questioning gets read as personal opposition. When he has asked the heads of skill-training programmes for outcome data, he has been called and asked, “What’s your problem? Why are you asking for results? Are you against me personally?” In an environment where asking whether a programme worked is treated as an attack on the programme owner, success and failure cannot be distinguished — and therefore neither can be learned from.

Special Technology Zones: the design, and the quiet sabotage

The most policy-dense passage in the episode is Syed’s account of the Special Technology Zones (STZ) framework, which he helped design while serving on the Prime Minister’s task force. Muzamil sets up the context for listeners: the idea was a piece of land with its own legal regime, on which IT and digital companies could register and receive tax breaks, infrastructure, cheap office space, and protection from the bureaucratic harassment that defines doing business in the rest of the country.

Syed reconstructs the logic from first principles. Pakistan’s legal framework, inherited from the British and accreted over a century, is too tangled to fix at national scale. The STZ approach borrows from the economic-zone playbook: pick a small piece of land, give it its own business-friendly law, and ensure that the rest of the country’s anti-business reflexes do not apply inside it. Critically, this includes policy continuity — the IMF can force changes to national tax policy, but inside an STZ the originally promised tax exemptions remain intact. It also includes cluster-based enforcement: when the country shuts down the internet, you cannot protect a thousand scattered IT companies, but you can protect three physical zones. Customs incentives only work if the goods imported under them cannot leak into the broader market — which only physical containment allows.

Then he names what is going wrong. The current direction, he says, is to virtualise the STZs — to let companies register as STZ entities while sitting in any plaza, any home office, anywhere in the country. “There is no concept of a virtual economic zone anywhere in the world,” he says. He has already seen NGOs trying to register as IT companies under the scheme. He has seen accountants registering as startups. He has seen the entire economy preparing to relabel itself as IT to claim the benefits. The moment State Bank, FBR, and customs see the first wave of fraud, every incentive will be withdrawn. “The single biggest one-window solution we had — with the potential for 200 to 300 percent export growth — is going to die very quickly in this direction,” he says.

Muzamil’s frustration is audible. The G Systems building has reportedly been declared an STZ. Companies are flooding in under the new model. The original physical-zone vision is being hollowed out from inside, and almost nobody outside the industry has noticed.

The fifty-percent dollar retention is a non-event

When the conversation reaches the IT minister’s most-trumpeted recent intervention — raising the dollar-retention limit for IT companies from 35% to 50%, allowing exporters to keep a larger share of their earnings offshore — Syed is dismissive. “I don’t think it has any effect,” he says. The previous 35% limit, he notes, was not even being fully used. Actual retention was running at around 8%. Moving the ceiling from 35 to 50 is irrelevant when nobody was hitting 35 in the first place. The real demand for this change came from three or four publicly listed companies that wanted to use IPO proceeds to acquire foreign businesses — and the current rules do not allow that anyway, because the 50% is usable only through debit and credit cards for cloud fees and similar small payments.

The “nine percent growth” the minister has attributed to this change is, in Syed’s reading, a coincidence of the November–December seasonal lift compared to a negative base. The real reforms required — policy continuity, ease of doing business, high-end skill development for the age of AI, market access, and country branding — are nowhere on the current agenda.

Why the Indian engineer in the Gulf is winning

Later in the discussion, Muzamil pivots to a workplace observation that has been raised on every Pakistani founder’s call he has been on. The Pakistani professional is, in his words, “a nawab.” Offered the equivalent of two and a half lakh rupees in Pakistan or sixteen lakh in Dubai, the Pakistani frequently chooses to stay home. The Indian doing the same job will take eight thousand dirhams, support a family in India, work twice as hard, and use the position as a foothold for the next decade.

Syed extends this with his own recent experience in China, which he describes as feeling like “a futuristic city” already ahead of the West. He had asked about salaries and was told everyone works six days a week as a baseline, twelve hours a day, often seven days, with offices that include sleeping and gym facilities. “No nation on earth can compete with that work culture,” he says. He contrasts this with the Indian model — less raw hours but a culture of “sirr jhukaa ke kaam karna,” head-down execution and long-horizon thinking. In the Gulf, the head-down worker is preferred. The Pakistani who arrives expecting deference, then walks away when offered a pay cut from ten thousand to seven thousand, loses both the job and the network the next decade was supposed to be built on.

The frame Syed uses is “value, not money.” Indians, Chinese, and Filipinos take the strategic salary because the value will compound. Pakistanis filter every decision through the money number on day one. “If we focused on value in a large economy,” he says, “converting that value back into money would not be hard.”

Saudi, leap, and what it actually takes to crack a new market

Muzamil presses Syed for a concrete playbook on opening a new market — specifically Saudi Arabia, where he had recently spent time. The answer is methodical and worth following step by step.

Step one is to travel. Syed is emphatic on this and says he tells every university audience the same thing: you cannot understand a market sitting in Pakistan. The first time he went to the LEAP conference in Riyadh, his company joined a PSEB delegation with a small booth, purely to read the market. “We were surprised by the level of traffic, the size of the requirements, and the level of interest,” he says. Step two is to read the local culture: Saudi prefers bigger and branded, so the next year his company took the largest Pakistani booth at LEAP, which generated genuine inbound interest from organisations like King Saud University and the Saudi Space Commission. Step three is the realistic lead cycle: roughly six to nine months from first meeting to mature opportunity. Step four is a joint venture with a local Arabic-speaking partner — a culturally normal structure in Saudi that solves the credibility and language problem in one move. Step five is sustained physical presence — five to seven trips a year, because no video call substitutes for being there.

He also lays out the macro case for why this window matters. Saudi’s Vision 2030 is pulling investment, including American and Western capital, into the kingdom because growth has slowed at home. Saudi’s AI authority is being personally championed by Mohammed bin Salman. Pakistanis still enjoy a residual cultural advantage in the Gulf — partly because so much of the Saudi military is run with Pakistani involvement, partly because of religious affinity, partly because the flight is three hours. “I think the opening is two, three, four years long,” he says. “After that, by the time the Indians have arrived in numbers, we will lose it, as we lost Dubai.”

Muzamil names two other adjacent windows. Japan has formally asked Pakistan for 100,000 programmers; Germany has asked for 300,000. Both are countries that historically do not accept foreigners. The bottleneck, Syed says bluntly, is language. Pakistanis do not invest in learning a new language. China is an even bigger opportunity blocked by the same wall.

The ten-billion-dollar question

By the end of the conversation, Muzamil walks through his own back-of-the-envelope sizing of the export prize. Using India’s $150 billion services export base as the reference and dividing by population, he arrives at roughly $20–22 billion as Pakistan’s theoretical addressable share. At $3,000 of export per headcount, that requires roughly 666,000 engineers, or — adjusting for attrition and emigration over a decade — eight to ten lakh engineers actually deployable into the services economy. Against an industry currently producing fifty thousand graduates a year with perhaps twenty percent employable, the gap is obvious.

Syed’s response is unsparing but not without hope. “The way we are going, even in ten years we will struggle to reach ten billion,” he says. “We need reforms at the scale of the Ayub-era industrial reforms — but for the knowledge economy.” No current leader has that political muscle. Which is why, he argues, the only realistic vehicle is the original STZ vision, executed properly. Inside a small piece of land with the right legal and policy framework, two hundred percent growth is plausible and ten billion dollars within three years is reachable. Outside it, with the current “superficial policy interventions,” ten years is optimistic.

Muzamil closes by asking whether Syed would take a policy role under the next government. The answer is a careful yes — conditional on real technocratic leadership, no political intervention in the technology portfolio, and a clear willingness to make long-horizon decisions in a country whose policy makers want results in three to six months. “Too many cooks spoil the broth,” he says. “Everyone in policy has an opinion. Even people who know nothing. We need clear leadership in the technology area, run by technocrats, without political angle. If those conditions are right — of course.”

By the end of the conversation, the picture has resolved. The slogans are loud. The underlying machine is flat. The one piece of policy infrastructure that could change the trajectory is being quietly hollowed out. And the people who could fix it are, in Syed Ahmad’s reading, mostly still sitting in the corner protecting their dignity.