Author: Alex Rampell & Justine Moore, a16z

Translation: Felix, PANews (with some edits to the original text)

a16z General Partner Alex Rampell and Partner Justine Moore discuss how AI agents are changing business and their impact on Google's business model, affiliate marketing, online shopping, and more.

Why Think About the Role of AI in Business?

Erik Torenberg: You have been thinking about the role of AI in business for a long time. Can you first talk about what inspired the creation of this article and how these ideas came about?

Alex Rampell: A long time ago, I founded a company called TrialPay. In fact, I have been selling things on the internet for a very long time, even before the internet existed.

I was thinking about how Google would be affected first by the emergence of AI. Many people are considering this question: is search volume going up or down? Personally, my search volume is going down, but not for commercial searches, rather for all non-commercial searches.

That's one aspect, and TrialPay is one of the largest affiliate marketers in the world. The essence of affiliate marketing is that you refer customers to others, which is one of the oldest business models on the internet. It actually predates AdWords and AdSense; you direct someone to a place, and you can earn a share of the revenue, or you can receive a commission. But will the affiliate model really become the driving force in the new business landscape?

On the other hand, there are these very expensive items. You are doing in-depth research on them using AI. But there is no affiliate model. So how do you conduct business and transactions? So, the first point is that the ontology of business is very interesting. The second point is, will the entire affiliate model still make sense? Because this seems to be an area that ChatGPT and other companies are venturing into. The third point is, just personally, I now use ChatGPT possibly three orders of magnitude more than I use Google, which is interesting.

Justine Moore: There are currently several very large consumer markets, with online shopping likely being the largest. But I think, so far, there have been relatively few startups trying to use AI to enter this market, even though there are now more opportunities due to these very intelligent language models and agents that can help you make better decisions than you could on your own, even purchasing items for you.

You might think this would create more opportunities for people to package this content into products and offer them to consumers, but we haven't seen many people doing that yet. So I think part of this discussion is to delve into why this system is so complex? In what different types of procurement can AI play a role? And when thinking about the broader market, what do we hope to see?

To What Extent Will AI Achieve Dynamic Pricing?

Erik: Alex, can you imagine an extreme world of dynamic pricing? Like seeing the same item on Amazon, but it charges you more, possibly because you are wealthier.

Alex: Well, I mean, people have tried this many times. From an introductory economics perspective, this is definitely a very intelligent world. For example, how do you capture consumer surplus? Consumer surplus benefits consumers and disadvantages producers. Obviously, Delta Airlines has also tried this, or at least they attempted to.

It's like a poor person's scenario; for example, if you are using an iPhone, you should pay more than someone using an Android phone because iPhones are more expensive. This actually indicates that your demand elasticity is different from those with less money.

I think this could face regulatory challenges. Of course, it will definitely encounter strong dissatisfaction from the customer base, but someone has tried this approach. But generally speaking, this approach is hard to succeed.

Why is E-commerce So Small?

Erik: Before discussing this issue, let's review the previous platform transformations. Currently, e-commerce only accounts for 16% of total retail sales. If we had discussed and predicted the proportion of e-commerce in business 20 years ago, we might have thought this percentage would be much higher. But why is the reality not so?

Alex: It turns out that the demand curves for immediacy and non-immediacy are different.

I once served on the board of a company called Wise, which primarily deals with remittances, and I found that the real-time remittance market and the two-day delivery remittance market are completely different. Because for items delivered two days later, the demand may not be as high. It's surprising that you say it's only 16%, which seems very low.

Justine: I think that number is high. I'm not questioning your research on the data. I think this because I feel there are many situations where people do research online and then purchase in physical stores.

Alex: Actually, the hardest part, and related to this topic, is attribution. It's everyone's pain point, like how do I attribute sales for Justine's MacBook?

And I think on the internet, the most common and corrosive business model is this "last-click attribution." You attribute 100% of the credit to the last click. Many people can't distinguish between correlation and causation and fall into this trap, and this business model is one I dislike the most, like Honey. It’s the kind that pops up when you are about to make a purchase on a webpage. It asks you, do you want a coupon code? Of course, you do, why wouldn't you? A 10% discount. Click here. You click here. It redirects you to an affiliate page. It places a cookie on your device. Then it redirects you back to the page you were just on, effectively stealing the attribution for that visit.

Why Are E-commerce Brands Struggling?

Erik: Let's revisit this industry. It seems that the biggest winners are at the aggregation platform level, like Shopify or Amazon, and independent big brands like Allbirds or Casper. They seem to quickly gain huge revenues but have not managed to become sustainable businesses like before. As they scale, their performance has not improved. Can you talk about the overall situation in this industry and why this is the case?

Alex: Perhaps it's because they are not really manufacturing products; usually, someone else is producing them, and that is not a good business model. E-commerce brands are merely resellers of products, and I think that is the problem. It's similar to what happened in the internet 1.0 era, where the long-tail effect of product retailers basically began to fade away.

Justine: I think there are also shoes like Allbirds or cosmetics that are very trend-dependent, especially in the internet age. Nothing can stay popular for that long. For example, one year Allbirds is the hot shoe, and then the next year everyone is chasing retro Adidas. Now On running shoes are in vogue. Just like this year, I was watching TikTok videos of sorority girls at the University of Alabama, and every girl was wearing On running shoes, while last year they were all in Japanese-style New Balance shoes.

Will AI Threaten Google's Business Model?

Erik: Now let's talk specifically about this issue. Which functions of Google will be taken away? Which functions will be retained?

Alex: I think Google has always been a typical freemium business model, meaning they built a better search engine, which is much better because of how they link.

When Google first started, most searches were free because commercial activity on the internet was still in its infancy; all searches were free and consisted of various information. I remember when Google first came out, I used it and felt it was much stronger than HotBot and everything else.

Everything was free, with no profit involved. They basically copied the Overture business model proposed by Bill Gross. This was a project of IdeaLab, which eventually became part of Yahoo. That’s why Yahoo ended up owning a part of Google; if you understand the whole history, you will see that what made Google the $2 trillion giant it is today was AdWords.

Currently, Google is starting to abandon some of the "free" parts, but the "mium" (paid part) has not diminished at all. Google is abandoning some information queries, like who won the Oscar in 1977? Such queries cannot be monetized, but people still want to know. People are now asking ChatGPT, and ChatGPT has 800 million active users weekly, which is a huge number.

But for "mium," the paid part of the freemium model, it still exists at Google. How do I know? You can look at their financial data. Their financial data is still growing, but we also know that search volume is actually declining. So if they haven't lost revenue, what have they lost?

Alex: I want to say that the biggest threat currently facing the internet is that it is unhealthy, or rather, the World Wide Web is now unhealthy. It is unhealthy because many things that once existed on the open internet now exist in so-called "walled gardens." By the way, search has long been fragmented. This is not caused by ChatGPT. For example, if you want real-time search, you have to go to Twitter or X. If you want to search for information about friends, you have to go to Facebook. Various "walled gardens" are one of the unhealthy factors. Another unhealthy factor is the commercialization of the internet, which in itself is not a bad thing. But the problem is, for example, if you are looking for the best running shoes, who are the people writing content about quality running shoes? In 1995, if you had a blog, first, you hosted it on your own site. You installed Apache on your self-assembled server and did it purely out of passion. And affiliate links provided a profit model, but they did pollute the open internet at that time.

It's like the top ten running shoes list. Do you know what that is? To me, that means the top ten affiliate marketing revenues. Then I would find someone to write a bunch of nonsense and optimize that thing to make money. Think about the situation before the internet; there is still a publication called Consumer Reports, and one cool thing about it is that it is the only publication that refuses to run ads. It is entirely maintained by subscriptions, with the idea that you can trust those real reviews. We really need that kind of model, but the entire business model has disappeared. Currently, much content is locked away, and the content that is not locked is filled with junk information.

Justine: To be honest, I think the channel with the least amount of junk content I've seen is video. When I want to see an honest review, I go to watch unsponsored YouTube videos, which usually have a lot of views. Amazon is actually a huge search engine. And it has been polluted itself. Many of Amazon's practices are just arbitrage. It's like a polluted sea of garbage.

Costco's Business Model

Alex: So far, my favorite business model is Costco. I believe Costco is the greatest company in the world because Costco refuses to sell inferior products. They refuse to charge high margins. Why do they refuse… it makes no sense.

Erik: Is it because they want to return profits to customers?

Alex: No, it's because it would diminish the value of membership. They make money from membership fees, so they charge you about $100 a year for membership. If you look at their net income, it's basically the number of members… they have over 50 million members, which is a huge number, multiplied by the membership fee, that's their net income. Everything else doesn't count for much.

They started their own chicken farm because the cost of rotisserie chicken was too high. That's how they operate, refusing to sell anything they are not satisfied with. Ordinary brands are just as good. So Costco is the best… just like we discussed in the business realm, whether before the internet era, during the internet era, or the AI era, Costco is immune to all of this because they are like Consumer Reports… they treat customers very well, and that's why this company is worth hundreds of billions of dollars.

Which Shopping Categories Will AI Encroach Upon?

Erik: Let's talk about which businesses AI will change? You mentioned several types of purchasing behaviors that might be "eaten away."

Justine: Yes, so we studied the range of impulse purchases, and I think these purchasing behaviors… I mean, they still are, like the bottles of Coke in the aisle.

But now many people will see something while watching videos, like in TikTok shops, and think, "That T-shirt looks cool, I want to buy it," and they might even seriously consider purchasing items like houses, wedding venues, or cars, which can take a significant portion of your income.

It's like one-time or multiple purchases; you need to do a lot of research. So I think both situations are hard to be disrupted by AI. I think impulse buying happens because you haven't researched beforehand, and you don't go to any specific place to buy. Essentially, you decide to purchase immediately when you see it.

So, algorithms will increasingly pinpoint you more accurately, like in your TikTok feed, if a shirt with your dog's name on it appears, you will be more willing to buy it than other things. But this is somewhat different from what we call generative AI.

Then there's the part of the purchasing process that requires the most consideration. I think it's hard to fully implement end-to-end AI because while you might start doing online research on ChatGPT, Gemini, or any emerging AI-native products, the act of purchasing is significant; you might want to have some kind of personal experience, to see, touch, and experience it firsthand.

So this means that in the whole category of products in the middle, we think purchasing behavior might be disrupted by AI in several different ways. One is obviously the way of research, like, you know, my travel bag has worn out, and I need a bag that can hold my laptop, a large water bottle, and everything else, and fit in the overhead compartment on a plane… If you're busy, you don't have that much time to do this research yourself. You could ask an AI agent to watch all the TikTok videos, read all the Reddit posts, gather real consumer feedback, and then provide recommendations. You might also want to click through options yourself. But I would say in this case, if there is a good purchasing integration solution, you are very likely to complete it through an AI agent.

There are also some things you already know you want, like wanting the best price. So I think AI agents can do a lot in terms of price optimization. For example, if you always buy a specific type of laundry detergent, it can search online to find out where that laundry detergent is the cheapest.

Additionally, there is another type of purchasing behavior that I think can be mediated by AI to some extent, but it may also have some impact on humans. For example, for bicycles, sofas, or some higher-value items like laptops, you want someone to take the time to really understand all your criteria and help you make the best purchasing decision. This is likely something you will use for many years, and it's important to you that it works well, is the best choice, and won't become outdated quickly. And today, I think the only way to achieve this is for people to deeply research posts on Reddit, like the buyitforlife forum, etc.

Or they have a truly trustworthy brand, like Apple, and are willing to pay a premium. I think it will be interesting in the future to have an AI agent that truly understands you, so you can have deeper conversations about these kinds of things. For example, it could even be a call where they dynamically ask you a bunch of questions back and forth, and you provide them with the information they need to go back, do research, and make decisions. This is some of our thinking about how AI will influence purchasing behavior.

Related reading: E-commerce giants like Shopify, Walmart, and Amazon are rapidly turning to stablecoins; will payments be a killer application?

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。