My chat with Abrar Haq, Cofounder of Tazah
Cautious optimism from a Pakistani technologist
When the Airlift’s shutdown news story broke, there was no shortage of takes about what it meant for Pakistan specifically and quick commerce broadly on Twitter. Everyone with strong opinions had found a new punching bag to make their points. Airlift's shutdown was playing out amidst broader global discussion around the excess of the past two years and the present contrasting downturn we live in now. I saw everything from blind optimism to reflexive dunks as startup folks across regions (mainly South Asia and America) were digesting the news online. One of the very few people tweeting out nuanced takes at the time was Abrar Haq. I discovered he was working on his startup Tazah in agriculture and had worked at now Uber-owned Careem. I decided to reach out to see if I could learn more from him. We discussed what he is working on these days, his reflections from his time at Careem, and his perspective on Pakistan’s startup ecosystem’s past, present, and future.
Sar: Can you talk about the breadth of problems being tackled by Pakistani entrepreneurs using marketplaces? I'm just curious about the state of the union of the local ecosystem and how it has evolved.
Abrar: While B2C marketplaces were the first to attract founder interest in Pakistan in areas like second-hand cars, real estate, job boards, classifieds, etc., none of them became large enough to attract institutional investors that could provide the founders a sizable exit. Zameen.com, the real-estate marketplace that eventually had an exit, could differentiate itself through regional expansion and had a large B2B element.
The first organised wave of B2C marketplaces came to Pakistan somewhere around 2012 when Rocket Internet unleashed some of its globally successful models in Pakistan. While most of them eventually folded, Daraz, the B2C ecommerce company, thrived, only to be acquired by Alibaba at a reported value of $180 Million in 2018. The talent that worked in the Rocket Companies during their foray into Pakistan was key in developing the tech ecosystem in the country.
Then came Careem and Uber somewhere around 2015-16, and the mobility space heated up. The multiplier effect that Careem had on talent, entrepreneurship, and investments is well-documented and unparalleled. Careem’s fast scale-up, competition with Uber, and the eventual $ 3.1 B acquisition by Uber put the Middle East/Pakistan region on the map. The talent that Careem developed went on to start more than 50 startups in the region, similar to the Paypal Mafia in the US.
It wasn’t until 2020 that B2B marketplaces suddenly got into fashion in Pakistan. Buoyed by the success of Udaan in India, we suddenly had quite a few B2B marketplaces in the FMCG/Kirana space. To be fair, the term marketplace is being used quite liberally here as the business model for most of these startups is closer to e-distribution than to marketplaces. The predominant operation is buying goods, storing them in warehouses, and distributing them to the retailers as the orders are received through apps or call centers. Interestingly, most of the capital raised in Pakistan is concentrated in this space as these businesses are easier to scale GMV-wise.
B2B marketplaces in other spaces have been more challenging because of their inherent complexity. For instance, we don’t have as many marketplaces/e-distributions in agriculture because of the fragmentation on both supply and demand sides (FMCG marketplaces have fragmentation only on the demand side).
Today, many companies are trying to solve various problems through marketplaces. While most startups are in their early days, Zameen (property marketplace), Daraz (B2C ecommerce), Foodpanda (food and now grocery delivery), and the two ride-hailing giants have had the most amount of success in finding PMF and building sustainable businesses.
Sar: I had no idea about Rocket Internet’s role in planting the seeds in the ecosystem years ago. Besides marketplaces, what are some other commonly used business models to tackle problems?
Abrar: Besides marketplaces, the SaaS approach to solving problems has also worked well in Pakistan from a user engagement point of view. Various ledger apps have built a strong presence among SMBs in Pakistan. Monetization remains a challenge, but we’ve seen clear intent, especially among the ledger players we have closely worked with, like CrediBook.
One success story in the SaaS space has been the FMCG distribution and salesforce management product, Salesflo. It has seen mass adoption among the large and medium FMCGs within Pakistan, and the same companies are deploying the product in markets outside Pakistan.
Wallet, payments, and lending are exciting spaces where many players aggressively build solutions for an underbanked population. Regulation, lack of creditworthiness data, and user behavior are key challenges. It is super hard for companies to create a product hook, get proprietary data, and monetize it. Most companies are good at one or two of the above traits.
Sar: You’ve cofounded a startup and are connecting farmers and businesses. Let’s talk about the status quo first. Can you talk through the industry, the major players, and their positions in the supply chain? How does the product get from the farms to grocery stores and restaurants?
Abrar: The current ag supply chain in Pakistan ($60B-$100B) is extremely fragmented on both supply (>10M farmers) and demand side (>1M businesses). The produce moves towards the end consumers in a highly inefficient supply chain (post-harvest loss ~ 40%) dominated by many layers of middlemen.
The biggest pain points for the farmers are poor access to capital, markets, inputs, and information. Farmers, the biggest risk takers in the value chain, make the lowest returns among all players.
Most businesses buy produce from informal mandis (wet and dry markets) daily. For the buyers, an inconsistent and opaque market with volatile prices and no incentives in the supply chain to ensure quality means high procurement effort and poor business profitability.
Sar: There are a lot of aspects to seeing fruits and vegetables from the farm to a restaurant. Everything from payments, contracts, and pricing to logistics, inventory management, and quality control. What pieces are you guys taking on beyond a marketplace’s discovery and matching functions? You guys are still early in your journey and have tried a few models. What do you do now, and what were the learnings from what you used to do?
Abrar: Tazah’s vision is to build an operating system for the agriculture value chain of Pakistan. While Tazah started as an ops-focused company, we have transitioned into a product-focused company over time.
As we spent time in the market, we discovered that building a supply chain parallel to the conventional supply chain is an inefficient way to deploy capital. When you do that, you are trying to compete with thousands of micro-entrepreneurs whose daily livelihood depends on squeezing some margin out of their deals. The assumption in such situations is that while you are initially not as efficient as the traditional supply chain, technology will get you ahead in the long run, and the capital deployed will build an unshakeable habit. The issue with this assumption is that the returns eventually achieved in a thin margin space will not justify the capital deployed. We can see some manifestation of this problem in FMCG marketplaces that are late-stage in India and Pakistan.
With these learnings in place, we had to find a more capital-efficient way to make the supply chain efficient while assuming limited control - and that’s how we transitioned to a product-first approach.
We needed a tool that drove high engagement in a key part of the agriculture value chain, had a clear path to monetization, and provided options for further growth. We found our break with agri-businesses for whom we built a SaaS tool that simplifies their operations. These agri-businesses (~ 500,000) each work with hundreds of farmers, providing credit, selling inputs, advising, and purchasing outputs, among other services. As we build traction on the product, we will enable these businesses to perform some of these services digitally. The software allows us to build a marketplace on both the agri-inputs and agri-outputs sides (domestic and cross-border).
The marketplace is structured such that most of the operations, including pricing, logistics, inventory management, quality control, etc., are performed by the marketplace participants in the best position to make them. Tazah solves demand generation, payment processing, and financing services.
Sar: Talk more about how the software helps you build the marketplace. You are selling your software to one of the key middlemen in the chain right now, right? When you insert a marketplace in a low-margin value chain, you usually can’t get away with charging high margins even when eliminating layers of middlemen.
Abrar: Sure, the software, and all its hooks are built to drive engagement and growth among a large user base - the agri-businesses. Unlike the traditional subscription-based monetization model, our software monetizes through the marketplace. The use of the software is free.
Additionally, most marketplaces are built to reduce/replace middlemen. What we are building brings more context to transactions, reduces inefficiencies in processes, and empowers the existing middlemen to do more business better. Yes, the space has low margins on average, but we have seen willingness by the players to pay for our product based on the value it adds.
That, coupled with the sheer size of the space, makes us very excited about what we’re building.
Sar: How do you envision helping the farmers? You are trying to get to them through the software you have built for the agriculture businesses that work with the farmers. How would you change the economics of the industry at scale? I imagine the farmers get a small portion of the retail price their produce eventually gets sold at because of all the intermediaries taking their cuts.
Abrar: While we eventually want to build for the farmers, our current tools focus on agriculture businesses. As they operate with more transparency, capitalization, and capability, we see them improve their P&Ls. Not only that, but their ability to serve farmers has also considerably improved.
The ultimate goal and impact of Tazah will be achieved once our products start reaching farmers. The agri-businesses provide us with the cleanest path to acquiring farmers and serving them.
As the farmers’ access to
Capital improves, their input purchase cost will come down
Market improves, their produce selling price will go up
Information and inputs improve, the farmers will be able to increase their yield
At scale, we expect farmer incomes to improve by more than 50% due to the above factors. The impact will be most pronounced for the small-holder farmers. With a more streamlined supply chain, we also expect a reduction in food wastage.
Sar: SMBs are notoriously hard to acquire and retain at scale. Bringing them online is a hard problem. Early generations of companies that fight tooth and nail just to get them to use more software do a service to the ecosystem in the future. Would you say Pakistan is still in the first innings of SMB digitization? Your go-to-market perspective above hints at your answer, but I’ll let you expand more.
Abrar: On the scale of hard to acquire and retain, Pakistani farmers arguably would be right on top. Limited internet access, smartphone penetration, and literacy are some of the top detractors. Interestingly, many industry stakeholders, especially the government, banks, and agriculture suppliers, have been trying to digitise them for a very long time. Unsurprisingly, so far, there has been limited success.
While we believe that with our product and operations background, we have a much better shot at building and deploying for the masses (farmers), we are taking a more innovative approach to solving this puzzle. The agri-businesses we are building for have direct access to and relationship with the massive farmer base. As we digitise them and make them more efficient, we are doing two things:-
For farmers, who are digital-averse, on average, we are providing a physical channel of service provision.
The database we are building of farmers will enable us to provide some services digitally later.
Sar: Is there anything you disagree with conventional wisdom on building marketplaces?
Abrar: I feel that the promised economies of scale in marketplaces and path to profitability do not materialise unless they are thought through from very early days. Unfortunately, for some businesses, the path does not exist. This also puts into question the model of scaling fast under the assumption that economics will automatically improve as the business grows. There’s a long list of companies in the world & the region that have absorbed a lot of capital and have stupendous GMVs but have no path to generating significant cash flows.
For operators who don’t want to end up with large companies with no path forward or backward years into the building, it’s important to have the ability to build with great clarity on what makes companies valuable: Not downloads, DAUs, GMV, revenue, not even EBITDA - a company will eventually be valued on future cash flows. Thinking through how you’ll eventually get there is important. Building organisational capabilities early that solve for strong economics even at a low scale is important. Monetizing early as evidence of customer's willingness to pay for the problem being solved is important. Startups, like everything else, are organic. After a certain time, things become super-hard to change. Great habits, competencies, values, and culture are best built early.
I can go on, but the gist is that entrepreneurs who care about their time and the problems they are solving should be able to think clearly and way ahead.
Sar: For someone who has worked at Careem, competed with Uber, and worked in an insanely capital-intensive ride-sharing market, I’m surprised by how much you have talked about growing sustainably in low-margin markets in our conversation so far. Perhaps, that experience informs your thinking now.
Abrar: Haha, true; the experience of working at Careem and then founding Tazah has brought a lot of perspective. Look, we all know the power law in venture-backed companies. Of all the venture-funded startups, very few will succeed. This works for the investors as they are banking on the Ubers, Googles, and Facebooks in their portfolios to return the fund. For the founders who are putting everything on the line for their startups, this skewed probability does not work. If they only focus on vanity metrics and swing for the fences, behaviors incentivized in venture-backed companies; they are playing the lottery with a high chance of ending up with nothing after years of relentless building. Hence, I feel that founders should play the game with more sanity. Swing for the fences, but also think about what will eventually make the business you are building valuable.
In most cases, it will be the ability of the business to generate a disproportionate amount of free cash flow. In some rare cases (WhatsApp acquisition by Facebook), you will have high engagement and data that will be valuable to a large player. That can happen, but you can’t build just counting on this outcome.
Sar: Are there any habits or lessons you picked up at Careem that you have to intentionally unlearn now that you are a founder and starting from scratch in a very different capital environment?
Abrar: Careem was a B2C startup with a lot of investor momentum by its competition with Uber and a strong PMF. When I joined Careem, it was going through the painful change process from growth to profitability, people-driven to process-driven, and ops-driven to product-driven.
The irony is that Careem could only win against a much better-funded Uber in the region by strong people and operations. Had they competed with Uber on products or processes, Uber would have smoked them. But the competencies that helped them win against Uber had limited use in the post-acquisition world. The Careem organisation had to adapt, and the process was not easy. While I can’t put my finger on the habits we had to let go of, I believe that between Mohsin (my cofounder) and me, having the front row seat as Careem evolved from blitzscaling to profitability was super-insightful.
Sar: Can you give an example of what Careem found challenging to adapt to after its acquisition by Uber?
Abrar: Careem became a successful company because of its values-driven culture and high quality of people. These people did whatever they had to do to beat Uber. Middle Eastern/Pakistani markets are hard and very different from Uber’s typical developed markets. While Uber is known for its hustle and grind in developed markets, with Careem, they met a match who could hustle harder. As Uber tried to win markets through superior tech, Careem teams treated every city and car category as a battleground and tried to execute better.
Post-acquisition, as Careem prioritised profitability, it had to move away from its empower the frontlines strategy and be more centrally driven. However, their best talent was on the frontlines, not in the centre. To centralise, you also need great processes, structures, and tech, which can’t be built overnight. The resultant transition was difficult.
Sar: Let’s switch gears and talk about what has been a hot topic for the Pakistan ecosystem. Airlift raised over 100 million dollars, employed over 1000 people, and hyper-scaled in Pakistan in less than two years after a massive pivot. High profile folks like Sam Altman, former president of Y Combinator; Biz Stone, co-founder of Twitter and Medium; Steve Pagliuca, co-chairman of Bain Capital; Jeffrey Katzenberg, ex-chief executive of Disney and Quibi; and Taavet Hinrikus, founder of TransferWise were investors. There are a few angles to this story. The most dominant angle in the discourse has been the debate about what it means for others in the nascent ecosystem. I think the negative outlook regarding the sentiment of local founders and investors is overblown. Would you agree? If there were ripple effects, I think it would be mostly on foreign investment flowing into the country.
Abrar: True, what happened at Airlift was very contextual, and most stakeholders (operators & investors) who actively operate in Pakistan know the same. They do not connect Airlift’s failure to the wider ecosystem.
To be fair, Airlift was a key part of the Pakistan startup story for the longest time. Airlift’s $12M Series A led by First Round Capital in late 2019 changed the game in the Pakistani startup space. This kind of raise was unheard of in the Pakistani context. For decades, startups in Pakistan who wanted to raise capital had very few doors to knock on. Round sizes were generally $25,000 to $200,000, early stage valuations above $1M were rare, the diligence was quite extreme, and don’t get me started on liquidation preferences and reserve matters. And then, a few months into its incorporation, Airlift raised a monster $12 M Series A. Like Careem’s exit to Uber, this was a monumental event for the Pakistani startup ecosystem. Immediately, potential founders started paying attention. Foreign investors started paying attention. And then, one after the other, we started seeing startups popping up and getting funded. The terms kept improving; before we knew it, Pakistan joined the global tech stocks euphoria of 2021. Airlift, in many ways, catalysed the explosive growth in Pakistan’s ecosystem.
However, while Airlift continued to dominate the funding headlines, its operations were anything but successful. Running a quick commerce business means building competencies in growth, product, supply chain management, procurement, pricing, warehousing, last-mile delivery, customer experience, etc. Airlift built the ability to scale fast. However, some of the abovementioned competencies did not materialise as quickly as required. That, the model’s inherent weaknesses and the macros eventually did it for them.
Sar: Let’s talk about the inherent weakness you mentioned. The category of quick commerce is still unproven virtually everywhere in the world. Many companies like Airlift have gone through layoffs, slowdowns, down rounds, and shutdowns across Europe, LatAm, and North America. My default view is the model doesn't make sense, are capital burning machines and never would have existed without the capital environment of the past decade. A more nuanced view would factor in the company and country-level factors. The people who can sustainably afford a premium for such on-demand deliveries also have full-time help around the house in many South Asian countries. Would you agree? Do you think the model is fundamentally not workable, or will someone eventually figure it out? You were at Careem. Uber is still struggling to become cash flow positive! It feels like quick commerce is just playing the delivery game on an ultra-hard mode.
Abrar: I don’t believe q-commerce is harder than running a ride-hailing business (two-sided marketplace) or food delivery business(three-sided marketplace). Having control over dark store placement and dark store operations considerably reduces variability, and you only have to deal with variability on the demand side.
Also, interestingly Uber did turn up a positive cash flow quarter just now. I feel that running a quick commerce business with good unit economics is still a tough problem to solve (not impossible IMHO). A slightly longer delivery window (1 hour to 4 hours) allows order pooling, less warehouse density, and better supply planning and makes the business proposition way more compelling. But will there be customer demand for it?
Your point on cultural context is very important. Yes, there is a house-help culture in middle and high-income households. There’s also the weekly or monthly shopping culture which, for many families with limited sources of entertainment, is an experience. They are deterrents but not big enough to completely knock out the possibility of quick commerce play in Pakistan. I also align with your view that most quick commerce players would not have existed if not for the availability of low-cost capital over the past few years. But, I still am not ready to give up on the larger grocery delivery play.
Sar: When you talk to your founder friends, are there any recurring topics or patterns that keep coming up?
Abrar: Pakistan’s startup ecosystem is still nascent, so the community is small. Most people know each other. It’s also one of the most supportive groups I’ve seen anywhere. I’m very fortunate to have made some great founder friends who share similar values that I do. Having this founder circle with whom you can open up is very important. Founding early-stage startups are hard enough anywhere in the world. A support group going through the same pains as you improves life.
Most conversations are about how hard building is :) Fundraising and investors are often discussed. Macroeconomics is another topic. Large global startups in the same spaces as we operate are discussed.
Sar: What are the biggest bottlenecks in holding the startup ecosystem back in Pakistan? What changes need to happen to unlock faster growth?
Abrar: There are two sides to this; internal and external.
Let’s start with the internal ones. Pakistan ranks poorly on ease of doing business rankings, and rightfully so. The country’s tax regime is one of the biggest deterrents to entrepreneurship. The complicated and often unintuitive tax code incentivizes tax evasion. Two of the most problematic areas are income tax withholding and import duties. Thousands of businesses in Pakistan have found ways to circumvent the tax laws, which makes businesses abiding by the tax laws uncompetitive. Regulations, laws, and their enforcement do not help either. It’s tough to enforce contracts in the country or collect money owed. The country is also prone to endless political uncertainty. The lack of political continuity disincentives structural improvement in the economy.
To overcome this bottleneck, the government has to improve the business climate in the country through legislation, contract enforcement, reduction of bureaucracy, and a simple tax regime that does not put compliant companies at a disadvantage.
Pakistan is in the early days of tech. While there is a lot of hunger among founders and the quality of talent in the country is immense, we lack the maturity in the ecosystem you gain through multiple iterations of building and macroeconomic cycles. With time this is getting better. We are going through our first downcycle as an ecosystem, and the learnings have been extraordinary.
On the external front, the brand Pakistan is still underdeveloped. The majority of foreign investment that came into the country last year and the year before was not for structured reasons. The global markets were heating up with multiples and valuations reaching crazy levels. Pakistan was still a bargain for anyone willing to take the risk. Funds and angels co-investing brought in more names. Founders who were good at fundraising built strong narratives and brought in money. Pakistan funds and angels inside Pakistan and in the US also played a role in building the Pakistan story and bringing in more money. Had the music continued to play, we might have been able to cement our position and get proper fund allocations for investment in Pakistan.
With the slowdown, a lot of this has changed. Most of the global funds will retreat to safe markets. Airlift going down and other casualties we will see over the next year will not help. We need many success stories emerging from Pakistan to create a strong and enduring brand in Pakistan. As many great companies are built, these success stories will emerge over the next few years. Until that happens, founders in Pakistan will have to reach a higher bar to get funded - higher than the bar that a founder in, say, India will have to reach the same stage. That is not necessarily bad - measuring up to a higher bar helps you build and execute better.
Sar: What countries or trends do the local folks discuss or pay attention to the most outside of Pakistan?
Abrar: It’s mostly India. We have similar cultures and socioeconomics, which makes drawing parallels easier. The Indian tech ecosystem is far ahead of us; it’s instructive to see the late-stage outcomes for most business models. Most founders raising venture capital say they are building the X (insert Indian comparable) for Pakistan - the Meesho for Pakistan, Khatabook for Pakistan, or Udaan for Pakistan.
There’s also the US. If you are building startups, you have to stay on top of what’s happening in the US. What trends are emerging there? What kinds of companies are the top-tier VCs funding? How are the macros changing there? What’s the current favored balance between profitability and growth?
Sar: Let’s end with a fun topic. What would first-time tourists to Pakistan get surprised by?
Abrar: I recommend people who want to learn about Pakistan to visit Pakistan and enjoy the full palette of its natural beauty. Pakistan’s northern areas have splendid lofty mountains. Five of the world’s fourteen highest independent mountain peaks are in Pakistan. We also have a more than 1,000 km stretch of coastline with beautiful beaches in the country’s South. The centre of the country has green agricultural planes and vast deserts. Not a lot of countries in the world offer this kind of range. Most foreign tourists who visit Pakistan are surprised by its beauty and the locals’ hospitality.
Recent chats :