Six questions before you build an O2O platform

This is the summary of a speech by Diaoye, a serial entrepreneur in China who built a beef brisket franchise, an essential oil brand, and a marketplace for beauty services. The speech was made in 2015 but we feel the questions are still relevant for today – with so many O2O marketplaces getting funded in Southeast Asia in particular over the last 2 years. 

There are six basic questions entrepreneurs need to ask themselves before they build an O2O (offline to online) marketplace.

1.Are you in a high frequency business?

In general, high frequency beats low frequency in the same sector. One example is Didi – we knew that Did would not make money from taxis from early days. But it educated a very high frequency market (ride hailing) through taxis, and then moved on to do more lucrative private hire cars to satisfy the demand.

Yidao, an early player of private hire cars, found it difficult – because from the beginning it had to educate two sides of the markets: customers as well as drivers. Didi only had to educate customers, while taxi drivers were driving anyway.

Afterwards, Didi built more services such as hitchhiking, premium cars and valet – they are occupying the sector. It becomes hard for others who just focus on one of the services to survive.

2.Do you have solid demand from your customers? 

People need to eat, and people need to move around – that’s why Didi and Ele.me are occupying very good grounds (note: this comment was in 2015, and Meituan has thus overtook ele.me, which was fully acquired by Alibaba).

Why I built beauty services marketplace. Well, there aren’t many other good marketplaces to build. The demand of beauty services especially with manicure as a starting point, though only covering a small population, is solid. We could then move to do hair and other services.

3.One to One or One to Many? 

In the past, restaurants would deliver to their customers point to point, using their own resources. Both individually and collectively this is very inefficient. You need one headcount and the person can only do one to two orders per trip.

With food delivery platforms, we some times see one person bundling 5-20 orders at one trip – this is huge gain in efficiency. And of course with that, you can keep the prices low.

Manicure and hair care is a one to one business – because they can only do as much each unit of time. We believe that for such services, you can’t lower the prices in a sustainable manner. Because the beauticians would prefer to sit in a shop and wait for customers if the price is not high enough, rather than travelling on the road to customers.

So if you are running one to many marketplace, you need the scale to make money. If you are running one to one, you need to customise to beat the others.

4.B2C or C2C? 

JD.com is B2C, which focuses on standardised products, Taobao is C2C which focuses on variety of non-standardised items.

The essence of Taobao is to have unlimited amount of SKUs, with many small merchants matching with demand from customers everywhere. Fully customised.

Customisation is the only way for you to escape a potential price war. In a C2C marketplace, inventory risk is absorbed by millions of small sellers, and thus not burdening the platform.

For this reason why it is easier to sell hard liquor on JD.com but not red wine. There are only limited amount of good white liquor brands in China, while red wine comes with thousands of varieties and brands.

If you can’t customise, then you will for sure be dragged into a bloody price war.

5.Entry point or subset? 

Are you the customer entry point, or just a subset of services from another entry point?

For example, if you build a marketplace for bridal services, you need to analyse what is the customer entry point. In China, new couples have to register with the authorities, and take pre-wedding photos at the same time.

So the photos are the entry point – however, if you do not give them access to wedding services, you will still not have a good entry point. So you need to start by offering wedding photos, but also wedding services.

However, even with that, wedding is a low frequency business and it is quite hard to build a scalable marketplace.

Another example is Ctrip, which started by booking hotels and air tickets – now it does everything travel related. If you capture an entry point, you will need to quickly offer more services to make your entry point a truly sustainable entry point.

Back to the example of Didi. Valet service is a sub set of what Didi could potentially offer, with lower frequency. If you are offering valet marketplace, you would want to sell out quickly. A case in point is eDaijia (eValet), a valet marketplace which I used to use a lot. However, as soon as I realise that Didi also has a valet service, I started just doing that on Didi.

6.Extensibility? 

If you want your marketplace to be defendable and profitable, you need to try to move both up and down the value chain. It is usually very hard to make money from high frequency services – however, using high frequency services to attract the customers and offering them low frequency services to make money is usually a good path to take.

One example is Tencent – their most successful product is Wechat, but they monetise through gaming which converts a sub set of Wechat customers. In doing so, they not only killed the SMS revenue streams of telcos in China, but also hit Nintendo’s business in China really hard.

We will see more of this, and if you are part of any value chain, you need to think how you should defend yours by attacking others, otherwise you will be attacked.

It will always be a dangerous game.