Amazon Leverages Web Growth to Build Infinite Book Catalog

Chuck Severance

Electronic commerce remains in its infancy but will ultimately be remembered as a transformative milestone in human history.

AI-assisted summary of Jeff Bezos 1997 Interview — original source. Check the original for full context.

Source published: 2013-12-22

Key ideas

Jeff Bezos founded Amazon after discovering the World Wide Web was expanding at an astonishing rate of 2,300% per year, prompting him to design a business around that momentum. (0:10)
Books were chosen over roughly 20 other product categories because the book market possessed over 3 million active in-print titles worldwide, enabling an online catalog scale impossible for physical stores to match. (0:10)
Amazon maintained a tiered inventory system, holding only a couple thousand bestselling titles in-house while using an 'almost in time' wholesale network to fulfill ~400,000 titles overnight. (1:26)
Amazon spent zero dollars on paid marketing during its first full year of operation, relying entirely on organic media coverage and customer word of mouth to build its initial mindshare. (2:48)
Bezos characterized the 1997 state of the internet as the 'Kitty Hawk stage of electronic commerce,' asserting the medium was still in its absolute infancy. (5:00)

Detailed notes

## Background Published on YouTube by Chuck Severance in December 2013, this video features an archival interview with Amazon founder Jeff Bezos recorded in June 1997 at the Special Libraries Association conference in Seattle, Washington. The video captures Bezos just three years after the launch of the NCSA Mosaic browser, articulating early strategic principles such as the 'long tail' well before they became mainstream digital commerce concepts.

## TL;DR In this 1997 conversation, Jeff Bezos recounts founding Amazon after discovering that World Wide Web usage was growing at an astonishing 2,300% annually. He explains why books were selected as the premier product category over music and other goods due to the sheer volume of active titles that physical stores could never house. Bezos also outlines Amazon's tiered inventory model—relying on a lean warehouse footprint coupled with 'almost in time' wholesale distribution—and details why customer-centric innovation, word-of-mouth, and trackable digital advertising outperform traditional print media. He concludes by characterizing the era as the primitive 'Kitty Hawk stage' of electronic commerce.

## Atmosphere & style The interview has the charming, unvarnished feel of an authentic 1990s tech time capsule, set against the low-key backdrop of a library conference. Bezos speaks with rapid-fire enthusiasm, youthful energy, and razor-sharp quantitative clarity, exuding an infectious optimism about the potential of the nascent internet. The conversational exchange feels casual and unhurried yet strikingly visionary, offering a rare glimpse of foundational tech doctrines being formulated in real time.

## Bulletin - [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) Jeff Bezos founded Amazon after discovering the World Wide Web was expanding at an astonishing rate of 2,300% per year, prompting him to design a business around that momentum. - [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) Books were chosen over roughly 20 other product categories because the book market possessed over 3 million active in-print titles worldwide, enabling an online catalog scale impossible for physical stores to match. - [01:26](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=86s) Amazon maintained a tiered inventory system, holding only a couple thousand bestselling titles in-house while using an 'almost in time' wholesale network to fulfill ~400,000 titles overnight. - [02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) Amazon spent zero dollars on paid marketing during its first full year of operation, relying entirely on organic media coverage and customer word of mouth to build its initial mindshare. - [05:00](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=300s) Bezos characterized the 1997 state of the internet as the 'Kitty Hawk stage of electronic commerce,' asserting the medium was still in its absolute infancy.

## Key takeaways - [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) Jeff Bezos left a quantitative hedge fund in New York City after identifying annual web growth of 2,300%. - [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) Bezos evaluated roughly twenty potential commercial categories before determining books were the optimal initial product. - [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) The sheer volume of book titles—over 3 million active titles worldwide and 1.5 million in English—far exceeded music, the second largest category with only ~200,000 CDs. - [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) Online retail's strategic advantage was only viable if it offered something that traditional physical retail methods could not replicate. - [01:26](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=86s) Amazon stocked only a few thousand of the top bestselling titles within its own physical warehouse. - [01:26](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=86s) For ~400,000 titles between the 2,000th and 400,000th bestsellers, Amazon operated an 'almost in time' electronic ordering pipeline across a dozen distributors delivering overnight. - [01:26](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=86s) Amazon maintained direct relationships with roughly 20,000 publishers to access an additional 1.1 million slower-moving titles. - [01:26](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=86s) Approximately one million out-of-print titles were included in Amazon's catalog through targeted specialized search and acquisition services. - [02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) Bezos identified customer attention as the true scarce commodity of the late 20th century. - [02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) Organic word-of-mouth recommendations and newspaper press fueled Amazon's growth entirely without paid advertising during its first year. - [02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) Print advertising presented severe tracking limitations, as testing multiple tracking URLs compromised brand recognition. - [02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) Online banner ads across 40 websites provided a 'marketer's nirvana' by enabling end-to-end tracking of click-throughs, sales conversions, and revenue per creative unit. - [05:00](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=300s) Bezos compared early electronic commerce to the Kitty Hawk era of aviation, anticipating massive future transformation over historical timescales.

## Detailed walkthrough ### [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) The 2,300% Statistic and Why Books Were the Ideal First Product

Jeff Bezos introduces himself as the founder of Amazon.com and recounts how the company began three years prior. At the time, he was based in New York City working for a quantitative hedge fund when he stumbled across an astonishing metric: World Wide Web usage was expanding at an annual rate of 2,300%. Recognizing the sheer momentum of that trend, he set out to formulate a business model designed specifically around that rate of expansion. He compiled a list of roughly twenty prospective product categories that could potentially be sold over the internet, ultimately deciding that books were the absolute best choice to tackle first.

```mermaid flowchart TD accTitle: The Selection of Books as Amazon's First Product accDescr: Diagram illustrating how 2,300 percent annual web growth led Jeff Bezos to evaluate 20 product categories and select books due to catalog size.

subgraph Catalyst ["Inspiration"] A["Working at quantitative hedge fund in NYC"] --> B["Discovers web usage growing at 2,300% per year"] end

subgraph Evaluation ["Product Selection"] B --> C["Evaluates ~20 potential online product categories"] C --> D{"Why Books?"} end

subgraph Comparison ["Category Scale Comparison"] D --> E["Music (Rank #2): ~200,000 active CDs"] D --> F["Books (Rank #1): >3 million active titles worldwide<br/>(>1.5 million in English alone)"] end

subgraph Outcome ["Strategic Advantage"] F --> G["Build a store online that couldn't exist any other way"] end ```

The primary rationale behind selecting books lay in the extraordinary scale of the category. As Bezos explains, books are unique because there are vastly more individual items in the book space than in any other commercial category. Music sits in second place with approximately 200,000 active music compact discs available at any single moment. By contrast, the global book market contains more than 3 million active, in-print titles across all languages, with over 1.5 million available in English alone. This sheer volume meant an entrepreneur could construct a digital superstore with a catalog size that physical brick-and-mortar stores could never replicate. As Bezos notes regarding early adoption: “Basically, right now, if you can do things using a more traditional method, you probably should do them using the more traditional method.”

[01:26](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=86s) Warehouse Inventory and “Almost in Time” Supply Chains

When asked about how Amazon manages its physical stock, Bezos breaks down the company’s layered inventory structure across its total catalog of 2.5 million titles. Rather than housing millions of volumes internally, Amazon stores only a few thousand of the top bestselling titles inside its own warehouse. For books falling roughly between the 2,000th and 400,000th bestsellers, the company relies on what Bezos terms “almost in time inventory.” When a customer purchases one of these 400,000 titles, Amazon electronically routes the order to a network of over a dozen book wholesalers and distributors. The wholesaler fulfills the request overnight, arriving on Amazon’s loading dock the following morning, where it is promptly packaged and dispatched to the customer.

Beyond those fast-turnaround titles, Amazon accesses another 1.1 million books directly through relationships with approximately 20,000 individual publishers, a process that typically requires a couple of weeks for delivery. Finally, the remaining portion of their 2.5-million-book catalog consists of roughly one million out-of-print books. For these titles, Amazon conducts targeted searches to locate copies wherever possible on behalf of the customer before shipping them out.

- **On-site warehouse stock:** A couple of thousand of the top bestselling titles. - **“Almost in time” wholesale network:** ~400,000 titles sourced from more than 12 distributors with overnight delivery to Amazon's dock. - **Direct publisher orders:** ~1.1 million titles sourced from 20,000 distinct publishers (typically taking a couple of weeks). - **Out-of-print catalog:** ~1 million titles tracked down and acquired through specialized searches.

[02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) Capturing Mindshare: Word of Mouth vs. Measurable Banner Ads

Addressing the fundamental challenge of building brand awareness on the web, Bezos broadens the premise: “Attention is the scarce commodity of the late 20th century.” He explains that Amazon’s strategy for earning attention was rooted entirely in building a novel, customer-centric service: “And we, one of the ways that you can do that, and it's the way that we did it was by doing something new and innovative for the first time that actually has real value for the customer.” Providing genuine utility triggered widespread organic media coverage from newspapers alongside rapid word-of-mouth customer recommendations. In fact, during Amazon’s entire first year open to the public, the company did not spend money on paid marketing, expanding purely on earned media and word-of-mouth fanout.

The interviewer mentions having noticed Amazon’s earliest print experiments—tiny blurbs positioned at the bottom of the front-page column in *The New York Times*, likening a displayed URL to a macro that expands into a storefront. While Bezos acknowledges that experimental print run, he points out that print campaigns carry a critical drawback: offline ads are exceedingly difficult to track. Using varying URLs to track response rates would hinder consumers from learning the core address. By contrast, digital advertising allows for complete accountability. Amazon runs banner advertisements across roughly 40 different websites, allowing them to track the exact ROI down to the creative unit:

[05:00](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=300s) The “Day One” Kitty Hawk Era of Electronic Commerce

Bezos closes the discussion with immense optimism about the future of the internet and digital commerce. He emphasizes that despite the company's early momentum, the medium is only in its absolute infancy. Comparing the contemporary web to the earliest days of aviation, Bezos declares: “This is the Kitty Hawk stage of electronic commerce.”

He reflects that both Amazon and the broader tech landscape are advancing rapidly across multiple disciplines. Looking back at the historical significance of the era, he concludes with enthusiasm that “the late 20th century is just a great time to be alive,” predicting that even a thousand years in the future, people will recognize the era as a transformative moment in human history.

## Notable quotes > [02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) Attention is the scarce commodity of the late 20th century. — Jeff Bezos

> [05:00](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=300s) This is the Kitty Hawk stage of electronic commerce. — Jeff Bezos

> [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) Basically, right now, if you can do things using a more traditional method, you probably should do them using the more traditional method. — Jeff Bezos

## Mentioned Amazon.com, Chris Anderson, Chuck Severance, Jeff Bezos, Kitty Hawk, NCSA Mosaic, Richard Wiggins, Special Libraries Association (SLA), The New York Times

## Video description This is an interview with Jeff Bezos - the founder of Amazon in 1997. It is very clear that Jeff is already on the right track in 1997 (three years after the introduction of NCSA Mosaic). While his words seem obvious now 15+ years later - in 1997 - he was way ahead of his time. The book "Long Tail" by Chris Anderson was published in 2006 (9 years later).

Watch the baby in the background escape from its mother at 2:16 and run towards the taping :) The mother came rushing through the frame to get the baby which caused the edit at 2:17. Taped June 1997 at the Special Libraries (SLA) conference in Seattle, WA. Video courtesy of Richard Wiggins.

[Source](https://www.youtube.com/watch?v=rWRbTnE1PEM)

中文简报

亚马逊借力网络增长打造海量图书目录电子商务仍处于萌芽阶段，但终将被铭记为人类历史上具有变革意义的里程碑。在这场1997年的对话中，杰夫·贝索斯回顾了在发现万维网使用量正以惊人的每年2300%速度增长后创立亚马逊的历程。他解释了为何在音乐和其他商品之前将图书选为首要品类，原因在于全球流通的书目极其庞大，实体店根本无法容纳。贝索斯还概述了亚马逊的分层库存模式——依靠精简的自有仓储配合“准即时”的批发分销网络——并详细阐述了以客户为中心的创新、口碑传播和可追踪的数字广告为何优于传统印刷媒体。最后，他将这一时期定性为电子商务初期的“小鹰镇阶段”。杰夫·贝索斯在发现万维网正以每年2300%的惊人速度扩张后创立了亚马逊，促使他围绕这一发展势头规划商业版图。 (0:10) 图书从大约20个产品品类中脱颖而出被选中，是因为当时全球有超过300万种在售图书，这使得线上目录能够达到实体书店无法企及的规模。 (0:10) 亚马逊维持着分层库存系统，自有仓库仅存放数千种畅销书，同时利用“准即时”批发网络在次日完成约40万种图书的履约配送。 (1:26) 亚马逊在运营的第一个完整年份中未投入任何付费营销费用，完全依靠自然媒体报道和用户口碑来树立最初的品牌认知。 (2:48) 贝索斯将1997年的互联网状态形容为“电子商务的小鹰镇阶段”，断言该媒介仍处于绝对的初创期。 (5:00)

Reading summary

Jeff Bezos explains how Amazon leveraged a 2,300% annual web growth rate to build an online bookstore, utilizing a layered inventory model and trackable digital marketing to capture customer attention during the infancy of electronic commerce. Selecting Books for Infinite Shelf Space (0:10) Inspired by a 2,300% annual growth rate in World Wide Web usage, Jeff Bezos evaluated roughly 20 product categories before selecting books as Amazon's initial offering. The decision was driven by scale: while the second-largest category, music, had only about 200,000 active CDs, the global book market featured over 3 million active titles, including 1.5 million in English. This vast selection allowed Amazon to build a digital superstore with a catalog size that physical brick-and-mortar stores could never replicate. The Layered Almost in Time Supply Chain (1:26) To manage its catalog of 2.5 million titles without massive overhead, Amazon developed a tiered inventory system. The company stocks only a few thousand top bestsellers in its own warehouse. For the next 400,000 titles, Amazon uses an "almost in time" system, routing customer orders electronically to over a dozen wholesalers who deliver the books overnight. The remaining catalog relies on direct relationships with 20,000 publishers for 1.1 million titles, and specialized searches for approximately 1 million out-of-print books. Capturing Attention Through Word of Mouth (2:48) Viewing attention as the late 20th century's scarcest commodity, Bezos focused Amazon's early efforts on customer-centric innovation to drive organic word-of-mouth and media coverage, avoiding paid marketing entirely during its first year. When Amazon did expand into advertising, Bezos favored online banner ads across roughly 40 partner sites over traditional print ads. While print campaigns proved difficult to track, digital ads offered complete accountability, allowing Amazon to measure exact return on investment, click-throughs, and completed sales per dollar spent. The Kitty Hawk Era of E-Commerce (5:00) Bezos characterizes the late 1990s internet landscape as the "Kitty Hawk stage of electronic commerce," asserting that the medium is only in its absolute infancy. Despite Amazon's rapid early momentum, he views the era as a highly transformative period in human history, predicting that future generations will look back at this time as a foundational milestone.

精炼阅读摘要

亚马逊创始人杰夫·贝索斯（Jeff Bezos）在1997年的访谈中指出，万维网2300%的年增长率孕育了前所未有的商业机遇；他通过选择选品极为庞大的图书品类切入，依托分层的“准实时”供应链与注重精准量化的数字营销，将当时的电子商务定位为处于起步阶段的“小鹰号时代”。品类选择：以超大规模图书突破实体零售限制 (0:10) 贝索斯回顾了创办亚马逊的契机：在纽约一家量化对冲基金工作时，他关注到万维网使用量正以每年2300%的速度爆发式增长。在评估了约20个适合线上销售的商品类目后，他最终决定将图书作为首要切入点。贝索斯认为，如果某种业务使用传统方式运转得很好，创业者就应当遵循传统；而图书品类独特的物理属性，为在线模式创造了不可替代的壁垒。选择图书的核心依据在于其品类规模的极其庞大。作为对比，当时体量位居第二的音乐类目仅有约20万张处于活跃发行状态的CD；而全球在版图书超过300万种，仅英文在版书目就达150万种以上。传统实体书店受物理空间限制，绝无可能容纳如此庞大的选品库。贝索斯由此确立了战略目标：通过互联网构建一个在物理世界绝不可能存在的巨型在线商城。供应链设计：自营仓储与分层准实时采购网络 (1:26) 针对250万种在册图书目录，亚马逊并没有采用全量囤货的重资产仓储方案，而是搭建了多层级的动态库存体系。在自营仓库内部，亚马逊仅常备数千种排名靠前的超热销书目；对于销量排名在第2000至40万名之间的大量在版图书，亚马逊实行所谓的“准实时库存”（almost in time inventory）机制。一旦消费者下单，系统会通过电子数据网络向十余家大型图书批发分销商发出采购指令，分销商次日清晨即可送达亚马逊装卸码头，随后立刻打包寄出。针对周转更低的长尾书目，亚马逊则进一步向外层供应链延伸。目录中约110万种图书直接对接约2万家独立出版商，通常需要数周的调配周期；剩下的约100万种绝版书，亚马逊则通过专门的搜寻渠道按需为客户追踪寻访。这种架构让亚马逊以极低的自身库存占用了横跨数百万书目的庞大目录。认知获取：从用户口碑传播到可量化追踪营销 (2:48) 贝索斯提出“注意力是20世纪末的稀缺商品”。亚马逊在成立的整个第一年里未投入任何付费营销预算，其品牌建立完全依靠提供极具价值的全新客户体验，从而激发了新闻媒体的自发报道以及消费者的口碑裂变。贝索斯认为，为客户创造货真价实的创新价值，是争夺公众稀缺注意力的最有效路径。在探索付费营销的过程中，亚马逊尝试了传统纸媒与网络数字广告两种路径。针对在《纽约时报》头条底栏投放的印刷广告，贝索斯指出线下广告极难进行效果归因，若为测试反馈而更改URL反而会干扰用户对主域名的记忆；相比之下，亚马逊在约40个合作站点投放的网络横幅广告实现了完全的效果量化，能够精确定位到单个广告素材的点击量、最终购买转化及每美元支出带来的实际收入，成为更可靠的营销投入方式。贝索斯将当时的电子商务进程类比为早期航空史上的“小鹰号时代”（Kitty Hawk stage），强调互联网商业仍处在极度原始的起步阶段。尽管基础设施与技术尚不成熟，但他坚信该领域各学科的演进速度极快，并乐观预测20世纪末的这场技术变革具有跨时代的历史意义，即便在一千年后仍将被后人视作人类发展史上的关键里程碑。

Detailed notes — zh

## Background 该视频由 Chuck Severance 于 2013 年 12 月 22 日发布在 YouTube 平台上。视频记录了 1997 年 6 月在华盛顿州西雅图举行的专业图书馆协会（SLA）大会期间对亚马逊创始人杰夫·贝佐斯（Jeff Bezos）的采访，展示了他在互联网发展初期远超时代的商业洞察。 ## TL;DR 在本次 1997 年的早期采访中，杰夫·贝佐斯阐述了在观察到万维网年增长率高达 2,300% 后创办亚马逊的契机，并解释了选择单品规模庞大的图书作为切入点的核心逻辑。他详细拆解了当时亚马逊覆盖 250 万种图书的四层履约机制，包括仅存畅销书的自营仓库、依托批发商次日送达的“准实时库存”、直采出版社以及绝版书寻检服务。贝佐斯还分析了 20 世纪末公众注意力的稀缺性，对比了报刊传统广告的局限与早期网络横幅广告可精准量化 ROI 的优势，并激情澎湃地将当时的电子商务时代定义为航空史上的“基蒂霍克时刻”与“第一天”（Day 1）。 ## Atmosphere & style 整场访谈洋溢着 90 年代互联网萌芽时期的纯粹热情与先驱探索精神。年轻的贝佐斯表现出极强的逻辑清晰度、商业预见性与感染力，其标志性的笑声与生动比喻让硬核的供应链与数据归因逻辑变得通俗易懂。现场环境带有一种展会走廊即兴对谈的质朴纪实感，贝佐斯从容笃定的表达与对未来千年的宏大展望交织在一起，给读者带来一种跨越时空见证商业传奇诞生的厚重宿命感。 ## Bulletin - [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) 杰夫·贝佐斯因看到万维网年均2,300%的爆发性增长而创办亚马逊，并在评估20种品类后选定单品数量远超其他行业的图书作为切入点。 - [01:26](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=86s) 1997年亚马逊通过250万种目录结合自营仓储、批发商电子直连与出版社直采，构建了线下无法实现的“准实时库存”履约网络。 - [02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) 公司首年未投付费广告完全靠口碑增长，随后放弃传统报纸广告，转而在40个网站投放横幅广告以实现从点击到销售全链路追踪。 - [04:09](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=249s) 贝佐斯将当时的电商发展阶段定义为航空史上的“基蒂霍克时刻”与“第一天”（Day 1），预言该时代将对后世产生深远影响。 ## Key takeaways - [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) 创办亚马逊的核心契机源于万维网使用量每年高达2,300%的惊人增速。 - [00:35](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=35s) 图书被选为电商首发品类，核心在于其在印单品规模庞大（全球超300万种、英文超150万种），远超音乐CD等其他品类。 - [01:05](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=65s) 早期技术不成熟时，网络商业的核心壁垒在于提供传统实体商业在物理空间上绝对无法实现的新体验。 - [01:26](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=86s) 自建仓库仅需常备几千种顶级畅销书，其余绝大部分长尾需求可通过分层供应链解决。 - [01:42](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=102s) 针对排名前2,000至400,000的约40万种图书，亚马逊通过电子化直连批发商网络实现了次日早晨到货的“准实时库存”流转机制。 - [02:04](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=124s) 对更长尾的110万种在印书籍直接对接20,000家出版商采购，并为100万种绝版书提供专门的寻书检索服务。 - [02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) 贝佐斯认为在20世纪末的商业环境中，注意力无论在线上还是线下都是最为稀缺的商品。 - [03:15](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=195s) 业务上线的首年未投入任何付费广告，完全依靠颠覆性的客户价值创新所引发的新闻媒体主动报道与用户自发口碑裂变。 - [03:38](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=218s) 虽然曾尝试在《纽约时报》头版投放微型广告，但由于传统纸媒无法精准衡量转化效果且需保持顶级域名统一，该模式被迅速叫停。 - [04:09](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=249s) 互联网横幅广告能够精确追踪点击率、成单率以及每美元投入产出比，贝佐斯将其称为营销人员梦寐以求的“终极乐园”。 - [04:58](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=298s) 贝佐斯将1997年的电子商务定义为航空史上的“基蒂霍克时刻”（初创起飞期），并坚定宣称当前仍是商业变革的“第一天”（Day 1）。 ## Detailed walkthrough ### [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) 创立亚马逊的契机与选品图书的逻辑在被问及身份与成名原因时，杰夫·贝佐斯（Jeff Bezos）自我介绍为亚马逊公司（Amazon.com）的创始人。谈及创办亚马逊的灵感来源，贝佐斯回忆起三年前他在纽约市一家量化对冲基金工作时的经历。当时，他偶然看到了一项令人震惊的统计数据：万维网的使用量正以每年 2,300% 的惊人速度增长。面对这种爆炸式的发展态势，他决定寻找一个在如此高增速背景下能够成立的商业计划。贝佐斯罗列了大约 20 种可能适合在线销售的商品品类进行评估，最终选定图书作为切入电子商务的第一最佳品类。他解释说，图书在一个非常关键的维度上极其特殊：图书品类中包含的单品数量远远超过其他任何商品品类。排名第二的品类是音乐制品，但在任何特定时间，市面上活跃流通的音乐 CD 仅有大约 20 万种；而在图书领域，全球所有语言活跃在印的书籍超过 300 万种，仅英文在印书籍就超过 150 万种。由于单品规模如此庞大，创业者完全可以在网络上构建一家在线下物理世界中根本无法存在的庞大商店。贝佐斯进一步指出，在互联网技术尚处于婴儿期的当下，如果一件事可以通过传统方式很好地完成，那么人们就应该继续沿用传统方式；而在线书店能够做到传统实体书店在物理空间上绝对无法实现的事情。贝佐斯在阐述当时的技术阶段与选品逻辑时指出：“基本上，现在如果能用更传统的方法把事情做好，你大概就应该用传统方法去做。” - **互联网增速数据**：贝佐斯在纽约量化对冲基金任职时，发现网络使用量年增速高达 2,300%。 - **初选品类对比**：筛选约 20 种潜在商品后选定图书；图书品类单品数远超其他品类。 - **品类单品量级对比**：活跃音乐 CD 约 20 万种；全球活跃在印图书超过 300 万种，其中英文书籍超过 150 万种。 --- ### [01:26](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=86s) 四级图书目录与“准实时库存”机制面对庞大的选品规模，主持人询问亚马逊如何管理库存。贝佐斯详细拆解了当时亚马逊总计 250 万种图书目录的分级履约体系。亚马逊自身仓库只常备销量最前列的畅销书，在库品类仅有几千种。对于排名在第 2,000 名至第 400,000 名之间的约 40 万种图书，亚马逊采用了一种被称为“准实时库存”（almost in time inventory）的高效流转模式。贝佐斯解释说，亚马逊与十几家批发商和分销商建立了电子化连接网络。当顾客当天在网站下单购买这些图书时，亚马逊会在当天通过电子网络向批发商下单采购，这些书籍次日清晨就会运抵亚马逊的装卸码头，亚马逊随即将其包装发往最终顾客手中。至于目录中另外约 110 万种长尾在读书籍，亚马逊直接对接多达 20,000 家不同的出版商进行订购，到货周期通常需要两到三周。目录中剩余的 100 万种则属于绝版书，亚马逊会根据顾客的需求开展搜寻检索服务，找到后再寄送给顾客。贝佐斯在解释这种库存模式的实际运作时总结道：“如果顾客今天从我们这里订购一本书，我们今天就向批发商订购这本书，第二天早晨书就会送到我们的装卸码头上，然后我们就能把它寄给顾客。” ```mermaid flowchart TD accTitle: 亚马逊 1997 年图书库存与供应链分层架构 accDescr: 展示亚马逊250万图书目录根据销量和获取难度划分的四层履约机制 subgraph FastTier["快速履约层"] A["自营仓库现货（几千种畅销书）"] --> Dock["自有装卸码头"] B["准实时库存（约40万种书籍，排名前2,000至400,000）"] -->|"向十几家批发商电子下单，次日清晨到货"| Dock end subgraph LongTailTier["长尾及绝版检索层"] C["直接对接出版社（约110万种书籍）"] -->|"对接2万家出版商，调货需数周"| Dock D["绝版书检索（约100万种绝版书）"] -->|"视情况搜索寻货"| Dock end Dock --> Customer["包装寄送至最终消费者"] ``` - **自建仓库库存**：仅存放销售排名前列的几千种高频畅销书。 - **准实时库存（Almost in time）**：覆盖销量排名第 2,000 至 400,000 的约 40 万种图书，依托 10 余家批发商网络实现次日晨间到货。 - **出版商直采**：覆盖约 110 万种图书，直接对接 20,000 家出版社，采购耗时约数周。 - **绝版书检索**：目录中包含 100 万种绝版书，根据订单提供寻书搜索服务。 - **目录总规模**：1997 年亚马逊全站目录收录图书达 250 万种。 --- ### [02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) 抢占注意力：创新驱动的口碑传播与早期广告尝试主持人提出，在互联网上建立影响力、抢占用户心智（mind share）极其困难，并询问亚马逊的秘诀。贝佐斯赞同这一观点，并将范围进一步扩大到整个社会大环境。他认为，抢占注意力不仅是互联网的难题，更是 20 世纪末期整个商业社会的普遍挑战，注意力已经成为这个时代最为稀缺的资源。亚马逊成功突围的核心，在于开创了一种真正为客户创造实际价值的新颖产品体验。只要能做出实质性的客户价值创新，新闻媒体便会主动报道，顾客之间也会互相推荐，从而形成巨大的口碑裂变效应，进而极大地推动并加速业务增长。在亚马逊面向公众开放的第一年里，公司完全没有进行任何付费广告投放，所有的业务增长全部依靠口碑传播和媒体的主动报道。随后，主持人提到曾在《纽约时报》头版专栏底部看到过亚马逊的小广告，并称赞这种将网址作为“宏指令”（Macro）的做法非常巧妙——读者在报纸底部读到简短文字并获知网址，进入网址后便能展开海量内容。贝佐斯笑着回应称那确实是亚马逊最早尝试的投放形式，但由于此类传统印刷广告很难追踪具体的转化效果，亚马逊已经不再使用。尽管主持人建议或许可以为不同广告位设置不同的专属网址来跟踪，但贝佐斯指出，品牌初期的首要目标是让大众迅速记住并熟悉统一的顶级域名，因此频繁更换投放链接在战略上并不可行。贝佐斯在强调注意力的时代属性时强调：“注意力是 20 世纪晚期最为稀缺的商品。” - **注意力的稀缺性**：贝佐斯提出在 20 世纪晚期，无论线上还是线下，公众的注意力都是最稀缺的资源。 - **第一年零广告费**：运营首年未投付费广告，单纯依靠媒体免费报道与口碑传播实现飞速增长。 - **纽约时报早期投放**：曾在《纽约时报》头版底部刊登极小尺寸的文字广告，引导用户访问主域名，后因难以精准衡量转化而停止。 --- ### [04:09](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=249s) 在线广告的可追踪性与电商的“基蒂霍克时刻” 对比传统报刊广告的难以追踪，贝佐斯分享了亚马逊当时在互联网数字广告上的最新实践。当时亚马逊在约 40 个不同的网站上投放横幅广告（Banner ads）。贝佐斯对网络广告的量化追踪能力赞不绝口：针对投放于不同渠道的每一组广告创意素材，亚马逊不仅能够精准掌握点击量（clickthroughs），还能追踪最终转化为实际购买的成单量（sellthroughs），并精确计算出在特定渠道投入的每一美元广告费最终带来了多少美元的营业额。贝佐斯将这种透明、可归因的数据反馈称作营销人员梦寐以求的“终极乐园”（nirvana）。在访谈尾声，贝佐斯对互联网与电子商务的未来展现出了极具感染力的乐观与远见。他认为此时此刻正处于极其关键的起点，是真正的“第一天”（Day 1），是电子商务的“基蒂霍克时代”（Kitty Hawk stage，意指莱特兄弟首次试飞的初创萌芽阶段）。各项技术与商业模式才刚刚起步，亚马逊与其他众多先锋公司正在各个技术领域向前探索。他感慨 20 世纪末是一个令人振奋的时代，并预言一千年后的人类回望现在，一定会惊叹于 20 世纪晚期生活在地球上的先驱所见证的伟大变革。贝佐斯在展望网络商业的前景时预言：“我认为一千年后的人们回望历史时会说：‘哇，20 世纪晚期生活在这个星球上真是一个伟大的时代。’” - **横幅广告（Banner Ads）布局**：1997 年在约 40 个不同网站投放横幅广告。 - **全链路追踪指标**：监控指标涵盖广告点击量、实际成单销售量、以及每单位广告投放金额产出的销售收入（ROI/ROAS）。 - **营销终极乐园**：贝佐斯用“营销人员的涅槃（nirvana）”形容数字广告全链路可衡量的特性。 - **时代定位**：贝佐斯将当时的电子商务比作航空史上的“基蒂霍克时刻”（Kitty Hawk stage），并重申“这是第一天”（This is day one）。 ## Notable quotes > [02:48](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=168s) 注意力是 20 世纪晚期最为稀缺的商品。 — 杰夫·贝佐斯 > [00:10](https://www.youtube.com/watch?v=rWRbTnE1PEM&t=10s) 基本上，现在如果能用更传统的方法把事情做好，你大概就应该用传统方法去做。 — 杰夫·贝佐斯 ## Mentioned 杰夫·贝佐斯, 亚马逊公司, 纽约时报, Chris Anderson, Chuck Severance, Richard Wiggins, 莱特兄弟, Special Libraries Association ## Video description This is an interview with Jeff Bezos - the founder of Amazon in 1997. It is very clear that Jeff is already on the right track in 1997 (three years after the introduction of NCSA Mosaic). While his words seem obvious now 15+ years later - in 1997 - he was way ahead of his time. The book "Long Tail" by Chris Anderson was published in 2006 (9 years later). Watch the baby in the background escape from its mother at 2:16 and run towards the taping :) The mother came rushing through the frame to get the baby which caused the edit at 2:17. Taped June 1997 at the Special Libraries (SLA) conference in Seattle, WA. Video courtesy of Richard Wiggins. [Source](https://www.youtube.com/watch?v=rWRbTnE1PEM)

Source quotes

[2:48] 杰夫·贝佐斯: 注意力是 20 世纪晚期最为稀缺的商品。 [0:10] 杰夫·贝佐斯: 基本上，现在如果能用更传统的方法把事情做好，你大概就应该用传统方法去做。

Source quotes

[2:48] Jeff Bezos: Attention is the scarce commodity of the late 20th century. [5:00] Jeff Bezos: This is the Kitty Hawk stage of electronic commerce. [0:10] Jeff Bezos: Basically, right now, if you can do things using a more traditional method, you probably should do them using the more traditional method.

Original transcript

[0:02] [Music] [0:10] Hi there. Who are you? I'm Jeff Bezos. [0:13] And what are what is your claim to fame? [0:15] I'm the founder of Amazon.com. Where did [0:17] you get an idea for Amazon.com? Well, [0:20] three years ago, I was in New York City [0:22] working for a quantitative hedge fund [0:24] when I came across the startling [0:26] statistic that web usage was growing at [0:28] 2,300% a year. So, I decided I would try [0:31] and find a business plan that made sense [0:34] in the context of that growth. And I [0:36] picked books as the first best product [0:37] to sell online, which making a list of [0:39] like 20 different products that you [0:41] might be able to sell. And books were [0:43] great as the first best because books [0:45] are incredibly unusual in one respect [0:47] and that is that there are more items in [0:49] the book category than there are items [0:51] than any other category by far. Music is [0:54] number two. There are about 200,000 [0:56] active music CDs at any given time. But [0:58] in the book space, there are more than 3 [1:00] million different books worldwide active [1:02] and in print at any given time across [1:04] all languages, more than one and a half [1:06] million in English alone. And so when [1:08] you have that many items, you can [1:10] literally build a store online that [1:13] couldn't exist any other way. And that's [1:15] important right now because the web is [1:16] still an infant technology. Basically, [1:18] right now, if you can do things using a [1:21] more traditional method, you probably [1:22] should do them using the more [1:24] traditional method. What kind of [1:26] inventory do you keep? We inventory uh [1:29] the bestselling books. At any given [1:31] time, we're inventorying in our own [1:33] warehouse only a couple of thousand [1:35] titles. And then we have we do almost in [1:37] time inventory for another 400,000 [1:40] titles or so. We get those from a [1:41] network of electronic we order [1:43] electronically from a network of [1:45] wholesalers and distributors. We order [1:47] those today. They're on our loading dock [1:49] the next morning. Then for another uh [1:53] 1.1 million titles, we get those [1:55] directly from 20,000 different [1:57] publishers. And those can take a couple [1:59] of weeks to get. And then the uh there a [2:03] million out of print books in our [2:04] catalog. We have a cal of two and a half [2:06] million books altogether. Those million [2:08] out of print books, some of them we can [2:10] get and some of them we can't, but we [2:12] find them uh if we can and then we ship [2:14] them to our customers. We do kind of a [2:16] search on those. What's almost in time [2:19] inventory? Almost in time inventory is [2:21] the phrase we use to describe a whole [2:23] selection of of books that we offer. [2:25] It's basically the things that are, you [2:27] know, below the 2,000th bestselling book [2:30] up to the 400,000th bestselling book. [2:33] Those are titles that we can get from a [2:34] network of more than a dozen different [2:36] wholesalers. So if a customer orders a [2:38] book from us today, we order that book [2:40] from our wholesalers today and the book [2:43] shows up on our loading dock the next [2:45] morning and then we can ship it to the [2:47] customer. They say one of the toughest [2:49] things to do on the internet is to [2:50] capture mind share. What was your [2:52] secret? How did you do that? Yeah, even [2:55] more generally I I agree with you that [2:57] you know capturing mind share on the [2:58] internet is extremely difficult. Even [3:01] more generally, it's the late 20th [3:02] century, not just the internet. You [3:04] know, capturing attention. Attention is [3:06] the scarce commodity of the late 20th [3:08] century. And we, one of the ways that [3:11] you can do that, and it's the way that [3:12] we did it was by doing something new and [3:14] innovative for the first time that [3:16] actually has real value for the [3:19] customer. That's a hard thing to do. But [3:21] if you do do that, then newspapers will [3:23] write about you, what you're doing. [3:24] customers will tell other customers and [3:26] you'll get a huge sort of word of mouth [3:29] fanout and and that can really drive and [3:31] accelerate businesses and that's what [3:32] happened with us in the first year of [3:35] opening Amazon.com to the public. We [3:37] didn't do any paid advertising and all [3:39] of our growth was fueled by word of [3:41] mouth and media exposure. I saw little [3:44] ads at the bottom of the column of the [3:46] New York Times. That was our very first [3:48] advertising. Um, we don't do that [3:50] anymore, but at the very at the very [3:51] beginning, we did little tiny ads at the [3:54] bottom of the front page of the New York [3:56] Times. I thought that was very clever. [3:58] It's sort of using a URL as a macro [4:00] because I I read then it expands. We're [4:02] a bookstore. Click here. Right. That's a [4:04] great way to think of it. And it worked [4:06] very well, apparently. I don't know. You [4:09] know, the problem with that kind of [4:10] advertising is it's extremely difficult [4:12] to track. um put a different URL for [4:16] every uh that's the problem is that you [4:18] want people to start to learn your URL. [4:20] So you don't want to actually use a [4:21] different one. And it's very uh easy. [4:25] One of the great things about online [4:26] ads. We do advertising today in maybe 40 [4:29] different uh on different websites. We [4:33] do banner ads and that advertising is [4:36] very easy to track in terms of knowing [4:38] how effective it is. So, we know for [4:40] each piece of creative in each venue, [4:43] not only how many clickthroughs we get, [4:44] but how many sellthroughs we get and how [4:46] many dollars of revenue it generates per [4:50] ad dollar spent on that creative in that [4:52] venue. And that is a sort of a [4:54] marketer's uh you know, nirvana in a [4:57] certain sense. Well, it's an exciting [5:00] place to be on the web right now. Oh, it [5:02] absolutely is. I mean, it's just [5:03] incredible. This is what's really [5:05] incredible about this is that this is [5:06] day one. This is the very beginning. [5:08] This is the Kittyhawk stage of [5:10] electronic commerce. We're moving [5:12] forward in so many different areas and [5:14] lots of different companies are as well. [5:16] And the late 20th century is just a [5:18] great time to be alive. You know, we're [5:20] going to find out that a I think a [5:22] millennia from now, people are going to [5:24] look back and say, "Wow, the late 20th [5:26] century was really a great time to be [5:28] alive on this planet. [5:31] [Music]