In the podcast (Can Shopee / Garena Beat Lazada? Tencent vs. Alibaba In SE Asia. (Podcast 29)), I teed up the idea of an escalating “arms race” for demand online. I argued that there is an increasing fight for the attention, engagement and spending of consumers. And that this is a natural result of our age of abundance, where there are more and more products, services and content than consumers could ever consume – but there is a limited amount of consumer time, attention and money. It’s just getting more and more difficult and expensive to reach, capture and retain consumers. And we are seeing this struggle in business after business. For example:

• Everyone is struggling with the increasing marketing fees at Google, Facebook and Amazon in order to keep reaching their customers.
• In my article and podcast about Ctrip and Meituan, I talked about the increasing struggle online travel companies are having getting their customers from search engines.

If you read the annual reports of digital companies, which is my favorite thing to do in life, (for better or worse), they will always tell you about their DAU (daily active users), MAU (monthly active users) and probably GMV. They will tell you about customer acquisition costs, retention and churn. And the first thing that will be listed in their Risk Section (my favorite section) is their need to continually attract, engage and retain consumers. User engagement is the central activity of most digital businesses. Absent physical assets like stores, users are usually their main asset.

This risk is particularly acute for platform business models, where the value is almost entirely from user interactions and the data this generates. Matching and network effects rise and fall with engagement and interactions. And activity data is the fuel for matching, AI and improving the user experience.

I had a nice discussion with two heads in AI at JD. And the conversation quickly shifted to this same idea of an arms race to thrill and satisfy consumers. They have too many options. You have to keep offering them more and more to get them to pay attention to you. If JD is now offering conversational AI for multi-modal engagement (which we talked about), then their competitors have to as well. And if they offer personalized services and inventory, then competitors need too as well. The fight for attention just keeps getting harder. It’s a never-ending arms race.

It’s Mostly About Demand-Side Scale

In my article on how platform business models evolve (here), I used the below graphic a lot. It argues that you need to get to demand-side scale ASAP. That is the key to most digital business. And that means building three intangible assets: users, engagement and data.

 

From there you can build in lots of directions, but without demand-side scale you are in a weak position.

There is also a continual fight to stay at the front of the consumer relationship. To be the primary interface for users. For example:

• Mozilla browser displaced the Microsoft operating system as the front of the cue for consumer attention on PCs.
• But then Google Search displaced Mozilla as the key point of interaction for users.
• Then Android and iOS displaced both browsers and search when the world switched to mobile devices.
• Then WeChat displaced the mobile operating system in user attention in China.

Basically, whoever is the front of the cue tends to be in a strong position with the consumer. And as the supply of goods and services keeps growing, everyone else struggles behind this function. Are you serving the customer or the Google search algorithm?

There are a lot of reasons why supply is growing so fast. But we can basically put supply into four categories:

• Digital products and services
• Physical goods
• Standardized services
• Differentiated services

In digital goods (movies, videos, texts, ebooks, SaaS, ERP, etc), it usually costs little to create, mass produce and distribute new products. So there are billions of user generated videos online you can watch on YouTube and TikTok. There are millions of ebooks available. The number of free plugins you can use on WordPress is crazy. And so on.

For physical goods, the explosion of supply began with book stores. It used to be that retail book stores might have 10-20k books in their inventory that you could browse. Then Amazon began offering millions of titles that could be shipped. This massive increase in available titles was awesome for consumers. But it made being an author a much tougher business (and one you should avoid).

Services (standardized or differentiated) are the latest frontier. We are going from simple services like transportation (Uber), food delivery (Meituan) and rentals (Airbnb) to more complicated and advanced services like medicine, legal and accounting. We are also seeing lots of smaller niches like ghost kitchens. Travis Kalanick is apparently now focusing on ghost kitchens, which means a restaurant just for delivery (i.e., no real estate costs and seating) in every apartment near your home. In theory, this could be every kitchen in America and China. The choices available for dinner within a local neighborhood could go through the roof. And restaurants based on having a nice, expensive location could be in big trouble.

***

A final comment on this. This increasing supply was nice but not staggering for US consumers. Yes, Amazon has tons of stuff, but so did Walmart already. However, in much of China / Asia, there has been very little selection for consumers up until recently. Your typical small town in Indonesia had a few small markets and convenience stores. Suddenly, Indonesian village consumers can buy virtually anything online from anywhere in Asia. It’s a pretty amazing jump forward.

Filters and Matching Are Critical for Consumption

I remember when the web started in the mid-1990’s. You could suddenly search on Mosaic browser for little websites for Nasa, Stanford, individual writers or whatever. It was awesome for like five minutes. And then it was bad. There were just too many webpages. And most of them sucked. You couldn’t find anything. It was like entering as massive library with no card catalog and tons of garbage books.

Enter search engines like Alta Vista and Yahoo, which provided filtering and matching. It turns out in a sea of supply, you simply can’t consume efficiently or effectively without filters and matching. Google is a filter. So is Airbnb. So are the home pages of Netflix and Alibaba.

Abundance is amazing but it can be a real negative in terms of value if consumers have to swim around in a sea of irrelevance. You need filters and matching that ensure relevance and quality.

In my podcast, I mentioned several filter types. These are from the book Platform Revolution by Sangeet Choudary.

• The Facebook newsfeed is a “push” filter based on past actions, social signals and other factors. It does the filtering for you and promotes passive consumption, which is something WeChat creator Allen Zhang argues strongly against (Allen Zhang On How Information Flows Are Shaping Our Lives).
• The newsfeed of Twitter is also a passive-type push filter but it is based on recency in time. And based on who you you choose to follow.
• Google search is a pull filter based on active intent and lots of factors.

So filters and matching are a key part of all of this. Although I tend to think of this in terms of curation and customization.

Curation is about maintaining a baseline of quality in the vast sea of products or services you offer. If you have a ecommerce site with a bunch of fake products, people aren’t going to come back. If people are staying in hotels and having bad experiences, they will stop booking through you. If half the guys on a dating site are behaving like pigs, the women will all leave immediately.

You need to curate the supply to make sure the user experience is good. And curation becomes harder and harder as the supply increases. Facebook and YouTube simply have no way of checking what all +2B people are uploading on their sites every day. These companies have amazing scalability is terms of content creation and sharing. But their curation process doesn’t scale nearly as well. Hiring lots of editors isn’t feasible. Having AI and reviews do it makes the curation process stupid and mindless (and annoys everyone). There simply aren’t any good answers to this curation at scale problem right now. Which is why Mark Zuckerberg gets yelled at about censorship every week (among other reasons).

Customization is different.

This is about making it personal. People also call this a data network effect. Customization began with tailoring the products or videos you are presented on Alibaba and YouTube. But it has now moved far beyond this. Now companies are tailoring their entire website and user experience to each individual. What Taobao looks like on your screen depends on who you are. Taobao on fourth-tier Chinese cities looks more like Taobao of 2012 than what we see in Beijing and Shanghai. Customization is the deep well of value add for your customers.

Customization is now expanding to what communications you receive. The content and the tone of communication is being altered. What conversational tone the chatbot uses is changing. Note: JD uses a 19 year old cheerful girl as the voice and personality of their chatbot.

Why Data is the New Oil

Filtering, matching, curation and customization all require data. You literally can’t do any of the above without data, mostly from consumers.

Which brings us back to the arms race for attention and engagement. Not only is it becoming harder to attract and keep users. But the tools you need to do this well depend on data from consumers. It’s a vicious or virtuous cycle for a company.

A Quick Summary

That was a bit of theory. But my argument is pretty simple:

• The supply of services, products, and digital goods is rising. We live in an age of abundance.
• The fight for limited consumer attention, engagement and spending is therefore getting harder and harder. It’s a never-ending arms race.
• And unfortunately, achieving demand-side scale is critical to be in this game.
• It is also critical to get consumer data to offer compelling and improving services.

That’s it for today. Cheers – Jeff