APAC Marketers

The Cold Start Problem – also known as the “chicken and egg” problem – describes a paradox found in two-sided business models:

Your business only works when you have both supply and demand. But when you start, you have neither, and to get one, you need the other.

In this post, we’ll look at how tech companies in SEA solved their cold start problem, successfully navigating the supply and demand paradox to become household names in the region.

But first, credit where it is due. This article is a mashup of two posts: Cheng Zishuang on how famous SEA companies got their first 1000 users, and Lenny Rachitsky’s analysis on how to kickstart a marketplace business.

Let’s get started.

For Most Businesses, Supply is the First Problem

Supply problems are more common because most marketplace startups are attempting to introduce a new operating model to their industry.

The reason the startup came into being is because they identified pent-up demand from consumers aching for a better solution.

However, in order to capture that demand, they first have to educate and convince the supply side to buy-in to their new model. This usually means lots of human effort, relationships, and persuading people to take a chance on a new way of doing things.

We can see that when we look at how companies in SEA kickstarted supply in their early days:

Grab knocked on taxi windows to onboard drivers

Grab’s early days were marked with intense competition with Uber. Recognizing that driver supply and availability was critical to keep users coming back, the Grab team leveraged their hyperlocal knowledge for driver onboarding.

They would go to auto repair garages to convince taxi drivers to download their app:

“Uncle give us a chance lah, your income is sure to increase!” Mr Tan recalls saying. “We were basically begging,” he says. “Maybe out of 10 drivers I pitched, two would agree to try it.” (Today)

Even though Uber also had teams on the ground, the Grab team always had an edge when it came to local context. For instance, they realized that natural gas petrol stations in SEA were disproportionately visited by taxi drivers:

In Malaysia, the Philippines and Thailand, the taxis use natural gas, so we have counters set up in petrol stations that sell natural gas. (500 Startups)

Gojek started as a dispatch for motorcycle taxis

Another way to solve the cold start supply problem is by having one foot in the old operating model, then migrating everyone over to your new paradigm:

Initially started, as a call centre in 2010, Nadiem Makarim was able to bring efficiency to the process of Ojek (motorcycle taxis) riders and customers by using a call centre to reduce waiting times. In 2015, when they launched the app to pent up demand, the rest was history. (Today)

Having relationships with relevant industry players in “old guard” also helps you sidestep opposition from entrenched interests and potentially suffocating regulation, both of which Gojek successfully avoided.

Bukalapak attracted sellers with community events

If you’re dealing with micro-SMEs like Bukalapak in its early days, there’s no choice but to increase supply by meeting and onboarding them in person:

“We are close with our customers. I, myself, sometimes go to the sellers’ community events, and they really appreciate it when we go to them, encourage them, and motivate them to sell more. We also listen to the problems they face. They appreciate it, and they become evangelists for our brand.” (GGV)

The face-to-face element was unavoidable as many of the warungs they served needed support with digitizing their business, something they only wanted to do with a trusted partner they could see in the flesh.

Carousell went to universities & flea markets to acquire sellers

C2C apps like Carousell can use a strategy of targeting a younger, tech-savvy crowd to build up supply:

“[…] the app gained traction by marketing on university campuses, placing magazine advertisements and looking for potential sellers in flea markets and online forums.” (TechCrunch)

In addition to requiring less manual onboarding, younger users typically transact lower-cost items, which means higher volumes of transactions, an important catalyst in the early days of a peer to peer buy-sell marketplace.

How To Solve The Demand Problem

Startups operating within an existing paradigm can expect demand to be the primary constraint.

That’s because there’s usually a base of suppliers who know the rules and will play ball – as long as there are buyers. Therefore, when facing a demand constraint, there’s an emphasis on getting to scale as quickly as possible.

Let’s see how these SEA companies rapidly grew demand on their marketplaces:

PropertyGuru: a land grab via SEO and PPC

When PropertyGuru launched, real estate listings revolved around print newspapers and classifieds. As for online listings, they were on all-purpose websites like Craigslist and Gumtree.

There was an empty white space in the market for a dedicated real estate platform like PropertyGuru. They solved their listing supply problem via brute force, taking listings from newspapers and classifieds, then inputting them manually into their database.

Not only did this solve for supply, this was a (very) indirect form of user-generated content. Putting these listings online gave them a significant long-tail footprint on search engines.

Combining this with search advertising was enough to kickstart demand and solve their cold start problem:

We had listings, we had about 17,000 properties [on our website] … now we had some content. Because there was no real competition in those days, we got organic traffic. And we spent a very small amount on Google. CPC rates were very low, because again, there was no competition. (Startup Grind)

Traveloka offered a better UX than the competition

The spark for Ferry Unardi’s founding insight for Traveloka came from the pain he felt while booking flight tickets from the US to Indonesia.

It’s no surprise that UX is one of his core beliefs for the company:

“If you develop good services for people, they will eventually come.” […] we should focus on how we can make users love to utilise the whole service. (Prestige)

But while great UX can keep people coming back, it doesn’t get people there in the first place. Without demand, they couldn’t get hotels and travel partners onto the platform. With no demand-side leverage, Traveloka could only sell loss-making or break-even products.

Unardi pointed his team towards growing traffic, while continuing to iterate on UX, eventually getting the traction they needed to set up deals:

“At that time, our focus was to grow our website visits and make improvements, such as developing a clean and simple website interface. In just a short period, airlines and hotel groups became aware of us and started forming partnerships with us.” (Prestige)

Tokopedia spread the word with PR

Ecommerce in Indonesia was a small, but rapidly growing industry when Tokopedia launched in 2007. A lot of buying and selling was still centered around online classifieds, setting the stage for ecommerce marketplaces like Tokopedia and Bukalapak.

Unlike Bukalapak, Tokopedia needed to get brands and large retail merchants onboard. One of their key demand levers was to make a splash in the media, which they did very early in their life:

On the second day of its establishment, Tokopedia was visited and featured by Tempo Magazine, one of the media in Indonesia. (Life at Tokopedia)


Lazada and the Rocket playbook: PR and digital marketing at scale

Lazada spent most of its early life being labeled “yet another clone” created by the Rocket team, this time a copycat of Amazon. While I’m sure the Lazada team found this grating to hear over and over, it made their value very easy to communicate to potential stakeholders and suppliers.


Trace the path of any Rocket company and you’ll find the same playbook in the beginning: make a splash with a big launch that drums up PR and media, then keep momentum going with digital marketing.

Lazada entered SEA by launching in 5 markets simultaneously. Then they rode the PR and media wave of how they were an SEA company taking on Amazon.

When it comes to their digital marketing efforts, there’s not much writing out there about what they specifically executed, but I suspect it’s because it was mostly ads. This quote from their CMO gives us a glimpse.

We positioned Lazada’s brand and quickly increased our market share within new product categories. We were able to increase ad impression share by 50% within a new competitive market, which resulted in an overall increase in traffic of more than 30%. (Think With Google)

Recap: Solving The Cold Start Problem

Here’s what we’ve learned from these stories:

If you’re introducing a radically new operating model to the industry, your primary constraint will be supply. You’ll use high-touch tactics to onboard sellers. You’ll need to go where they are or tap into existing networks, then manually onboard them to make sure they’re good representatives of your platform.

If you’re competing within well-understood industry dynamics, then you’ll be constrained by demand. You’ll use tactics with high scalability. This could take the form of owning digital marketing channels where you have an unfair advantage, riding the wave of a media narrative, or localizing wider and faster than anyone else.

The Challenge of Southeast Asia

A final lesson not explicitly mentioned here is that the solving Cold Start Problem is already hard, but even harder in a fragmented market in Southeast Asia.

Solving it in one country doesn’t necessarily give you a tailwind to set up in another country, even if those countries are neighbors or can speak the same language. And when you think you’ve got it nailed, your market can suddenly turn into a red ocean.

For these companies to leave an indelible footprint in Southeast Asia, they’ll not only need to prove that they can solve the cold start problem in multiple countries, but that they can keep evolving with an ever-changing region.

Thanks again to Cheng Zishuang for letting me use his post as inspiration. Check out his Substack here.

Discuss this post on LinkedIn →

Leave a Reply

Your email address will not be published. Required fields are marked *