When food is late, and startups are asleep, or why it's time to rethink food delivery from the ground up

Bakht Niyazov, city — Almaty, venture investor, Instagram

Soon I will officially announce the launch of my venture studio Justart. This will not be just an investment structure. I want my studio to become a place where I, my partners, and the founders of future startups would continuously learn to find opportunities in gaps, cracks, and pains that do not immediately look like a product niche.

To train this very "vision," I started looking at markets that seemed to be already captured to some extent by young technology companies. In particular, I wanted to explore the food delivery market — seemingly overheated, occupied, with no chances for new players. But the deeper I delved into it, the clearer it became: we are standing at the edge of a systemic turning point, which means this is precisely the moment when strong and sharp breakthroughs for startups are possible.

Experiment: food not for food's sake

For a week, I deliberately ordered food through various services — Glovo, Wolt, Yandex, Kaspi, Arbuz. I made more than 20 orders. Only Kaspi Magnum delivered on time — perhaps simply because they do not initially specify a fixed delivery time. They have a floating timing. All other orders were delayed by 30-50 minutes, and sometimes even longer. Contact centers are all dormant.

Overall, I came to the conclusion that the delivery market is, to put it mildly, not in the best state. There is a shortage of couriers, traffic jams do not decrease even in summer, transport regulation is tightening, and the aggregator business model is overheating and can no longer withstand current loads. Meanwhile, the companies themselves do not seem to consider it necessary to improve customer interaction. Today's delivery services are like a bridge that has long been overloaded, but no one wants to admit that its load-bearing capacity is at its limit.

Why aggregators are strategically uninteresting to restaurants

I love communicating with restaurateurs. For me, these are people who shape the image of my favorite city. Remove the restaurants from any beautiful city — and what will remain? I notice that more and more restaurants are creating their own delivery services. Even if they are simple, even primitive for now — but at least restaurants stop depending on aggregators.

What do establishments gain by creating alternative delivery infrastructure? What is the problem with working with aggregators? With these questions, I turned to Yuri Negodyuk — in my view, Yuri explained that aggregators take a commission of 30-35% of the order. This "eats up" almost all of the restaurant's profit. Currently, restaurant EBITDA has already dropped to 15-25%. Own delivery is a way to preserve business margin, especially for chains with large order volumes.

Through their own services, restaurants can control packaging, delivery times, courier service quality, integrate branded service, and receive direct feedback from customers.

In turn, aggregators do not share customer data. By delivering directly, restaurants can build their own CRM, launch personalized offers, develop loyalty, and stimulate repeat sales.

Another important aspect is that aggregator algorithms are unpredictable: a restaurant can "fall out" of the top without explanation, losing a large unplanned number of orders. Another incentive to build business directly with customers. 

I think that even this information contains a huge insight: restaurants need a product — a super-modern approach to CRM and all customer interactions.

Restaurant managers are extremely conservative and often do not understand innovations. But they understand that with low margins, it is necessary to increase sales using modern solutions.

Currently, this is undoubtedly — sales using AI. The possibilities are huge. If you create an effective solution in Kazakhstan, you can immediately enter the international market with it.

Delivery as a failed disrupt

The most paradoxical thing is: during all this time, delivery services have not improved the main thing — neither the food nor the restaurants themselves. They did not make the food tastier. They did not make people healthier. They did not add pleasure to the process itself.

They simply became an intermediate link that takes a commission from the restaurant, reducing its profitability. They took away clients. This is not disrupt. This is arbitrage. And harsh arbitrage. 

The future? Of course, autonomy… but don't rush

Yes, in the future, food delivery will be done by robots, drones, neural networks. This is already happening in Japan, China, and the USA. Personally, I am not very well versed in hardware and do not try to get involved in it. I understand that sooner or later any driving will become autonomous. And that drones will solve a huge number of tasks. 

But it seems to me that startup founders should not be looking at this. It is much more interesting to find new opportunities where few people are looking for them yet. 

What is more important — the box or the intelligence? What concerns me is: how will all this delivery infrastructure work in the artificial intelligence economy?

Maybe I am wrong, but so far none of the aggregators seriously realizes that their main asset is not food, not speed, and not even couriers. It is data.

Currently, my AI team is negotiating with one of the aggregators. My feeling is: they are more focused on survival within the current business model built on eating into restaurant margins.

And now imagine: millions of orders every day. People tell what they want, what they eat, how often their preferences change, what they buy besides food, how they pay, where they live, and where they spend their time. This is real gold.

But go into the Wolt or Choco contact center — and it becomes clear how this "gold" is handled: not at all. They do not see value in customer interaction. They do not even use basic, relevant AI solutions.

Recent experiment: an acquaintance of mine ordered some simple food from Choco and had a long live conversation with an operator by phone. Processes inside are manual along the way. I doubt they have speech analytics.

Opportunity for startups: building the "brain" for delivery

Here is the real window of opportunity. Not in robots, but in the following:
— AI modules that analyze preferences and offer upsell
— smart chatbots that work for restaurants and know how to sell
— low-code solutions that embed into delivery microservices and make them smarter, faster, more human.

Yes, at first this will be a narrow market, and people will be reluctant to pay. But in two to three years, this will become the basic infrastructure of smart retail. Those who start now will be in the right position.

Delivery as part of a sustainable society

There is a deeper layer. Today, individual delivery is a product of the ego-economy: for me, now, quickly. But such a model is unsustainable — neither ecologically, nor logistically, nor socially. 

We complain about the city akimat, but we do not realize that we ourselves create problems for ourselves with our own egoism. True value lies in collective and marketized forms of delivery, where not only time but also load on cities, ecology, and courier incomes are optimized. This is a real upside that should be piloted within Almaty.

AI technologies will help predict that in this area at 19:00 there will be a wave of orders, and it will form aggregated delivery, not 20 trips for one package. This is a cool opportunity.

Final question: what should restaurants do?

They too must transition to intelligence: their own CRM system, their own vertical chatbots, recommendation mechanisms, their own understanding of the client.

And this is a field for startups that is almost unoccupied. I will be closely watching all this.

Conclusion: not armor, but brain

We have built armor for too long: platforms, warehouses, clusters. This is important, this is infrastructure.

But who today is building the brain of food delivery in the era of the new AI economy?

The world is changing. And I want to be among those who launch startups not for hype, but for creating new neural connections — between people, services, and cities. If you are a founder and this topic resonates with you — write.

No one knows what product will grow out of this pain. But one thing is clear: it will not be an aggregator that is 50 minutes late and still takes a percentage from the restaurant.