Amazon shareholder letter circa 2000

Classic innovation vs efficiency thinking

Per Scott Devitt from MS

 In the 2000 shareholder letter, Mr. Bezos stated, ""Price performance of processing power is doubling about every 18 months (Moore's Law), price performance of disk space is doubling about every 12 months, and price performance of bandwidth is doubling about every 9 months. Given that last doubling rate, Amazon.com will be able to use 60 times as much bandwidth per customer 5 years from now while holding our bandwidth cost per customer constant. Similarly, price performance improvements in disk space and processing power will allow us to, for example, do ever more and better real-time personalization of our Web site.

In the physical world, retailers will continue to use technology to reduce costs, but not to transform the customer experience. We too will use technology to reduce costs, but the bigger effect will be using technology to drive adoption and revenue. We still believe that some 15% of retail commerce may ultimately move online."

Efficiency vs Innovation

It seems that about every nine months I visit my Yahoo mailbox (Yahoo has the largest installed base in the US with 282mm identities and is second to Hotmail worldwide). On my most recent visit, I was greeted with an extensive upgrade. The home screen was touting a 2x speed improvement, better spam control, new social features and unlimited attachment storage; with a whopping 100Gb limit/attachment (in cooperation with YouSendIt). All these adjectives sounded great. Indeed, it's much improved, and it's as if the team went down a check box of features in a methodological way and executed in an efficient manner. Yet, sadly, like much of Yahoo today, it offers little in the way of real innovation or insight. It lacks passion.


The stagnation of Yahoo, while incredibly visible, is not a solitary event, in fact, after reading Jane Jacobs' Economies of Cities and Carlota Perez's Technological Revolutions and Financial Capital: The Dynamics of Bubbles and Golden Ages (thanks to Fred Wilson for suggesting it) there's a real science into why Yahoo is a shining example of markets ruthlessly punishing companies losing their innovative way, while not embracing the next logical developmental milestone; efficiency. Microsoft, as an example of successfully crossing this chasm. It is hardly innovative, however, under Steve Ballmer's leadership, it's undeniably an efficiently run organization.


Markets experience overlapping cycles. For example, we have the ebb and flow of financial markets, represented by the financial exchanges and innovations from fleet footed vendors spurring the advance/decline of production markets.  Each are related and often complementary, but with different foundations and drivers. Carlota Perez highlights that new markets are financed through production capital (much time and innovation and relatively little money), funding the ways and means of production. Later, financial capital (facilitated by the banking industry) supplies the means for massive deployment and broad sharing of wealth (ignoring for this discussion inevitable bubbles). While Jane Jacobs points to the drive to efficiency as the antithesis of innovation. Using the growth of cities, and the failure of planned urban renewal as examples, she gives a series of cases of how central planning stifles innovators. 


From the outside, I assumed that when Carol Bartz was brought on board at Yahoo, she would be a terrific hire. A no nonsense manager with a purposeful perspective, it seemed that she would bust through the layers of bureaucracy holding back the Company from building on its success and innovating within the social and mobile waves hitting the industry. The Company under her leadership did pare down the bureaucracy and had clear lines of authority, but it lacked a fundamental and axiomatic truth. In this industry a company must evolve and innovate in order to prosper. Yahoo did neither. 


Press reports about numerous Private Equity players preparing bids to acquire Yahoo are logical. With an abundance of Financial capital available, they surely will bring efficiencies new owners will demand. I just hope, that building on the failure of Ms. Bartz, there is a remnant of the culture that made this a once proud company and they uncork the beauty of Production capital which is so sorely needed.

Do It Yourself (DIY) vs Do It For Me (DFM)

Two unrelated events over the past few weeks highlighted for me an acceleration of change in the application and infrastructure ecosystems. Until I 'upgraded' to iOS5 on the iphone, and Apple killed the independent Siri application, I had been running Siri on my 4. Its feature set (along with Dragon Go!) is billed as a voice assistant, really a software layer that sits across a myriad of databases and retrieves the data it thinks you are looking for. Rather than opening a specific application and conduct a search, the application interprets and parses verbal instructions, searches various databases and returns an answer it thinks solves the question at hand. The application essentially provides a service for me (DFM) which replaces manual searching (DIY). 


A multitude of examples abound for DFM services which range from Box.net (storage/back-up), Google or Yahoo News (algorithmic filtering), to Amazon's suggested items to buy. Many are assuming the role of gatekeeper or curator for the personalized serving of information or commerce options. 


I was exposed to another example of DYM in the techie world while attending PuppetCon with Cloudsmith's management. Firms such as Puppet Labs are shielding administrators from layers of complexity via their automated IT management solutions, which enables IT professionals to manage dynamic cloud infrastructures. Companies, such as Twitter and Match.com, could not provide the massive global solutions they offer without folk like Puppet. Creative programmers and business folk are leveraging the latest infrastructure advances to mask complexity, enhance access, and increase utility. A killer combination.


Each time we get an advance in processing, bandwidth, storage or displays, programmers busily optimize/advance their programs by packing in more and different utility for users. It's one reason why older devices seem to degrade over time; the upgraded applications are no longer designed for their device and, the last generation(s) are ill equipped to keep up. 


Here's a video of Eli Pariser describing the potential downsides to enhanced filtering. It's really informative:

Techstars NY

Yesterday,  one of the NY based technology incubators, TechStars, hosted a session for its most recent 12 graduates to present their companies to a group of 500 investors, peers, and company executives. It was one of the most impressive young company events I have been to in many years and highlighted a few NY-centric points:

1. This generation of companies is far different from their brothers/sisters of a decade ago. In a similar venue (though before incubators) just about every presenting company would have highlighted a business that was perched, precariously it turned out, on three fragile letters....CPM.  This group of entrepreneurs were starting companies that ran the market gamut, ranging from infrastructure to application and business to consumer.
2. The number of second time entrepreneurs, with pedigree from great companies was notable, as were the dropouts from Penn and the one high school coder
3. The scope of early relationships/investors with Google, MSFT, Amex and Accel was impressive

Here's a brief summary of the presenting companies, in order of appearance:

Contently- marketplace which connects writers with brands. Thesis is that every brand needs to produce their own content. Issue is to ensure they get high quality content. Contently manages the matching, workflow and payments.  Organizes the market, like Bitwine, and is compensated as a % of the deal. Have 2,000 writers in market now. AMEX Open Forum is now a partner (not sure what this means).

Ordr.in- restaurant ecommerce like a Sabre for food. GOOG ventures is a seed investor.

Urtak- A polling infrastructure (think Q+A which you can embed on your site). The company objective is to organize the world's opinions, while increasing engagement (and providing metrics) on the sites which use their infrastructure

MobIntent-A tool for the creation, optimization and management of mobile ad campaigns. Mobintent simplifies the workflow across medias and claims 3x uplift in cost/click over Admob

Spontaneously-mobile application to enable people to meet friends at the last minute; or a tool to manage your free time

Piictu-a visual network based on a mobile app. Photos are now a foundation for interaction rather than static images.  Piictu creates a network around the photos. 130k downloads in 8 weeks. Users spend 2.5 min each app session. 62% of users post photos. SoftBank, BetaWorks and RRE were seed investors

Sidetour- 'Peer to peer market for experiences'. For example, booking a luge track session with an Olympic competitor. Or to paint graffiti with an artist. The company earns a 20% fee for running the marketplace. Avg price per booking has been $60/person with 6 people per session. Foundry and RRE are investors

Ambassador-social referral platform to help brands create and manage (incentivize) ambassadors. Www.getambassador.com Have an API which integrates into your site and they handle the tracking/management.

ChatID- let's companies engage with consumer via chat, but does not have to be on the company landing page (e.g. talk to a Sony rep via the BestBuy site). Feels like a LPSN competitor, but is more broadly architected to work across many chat front ends. Co-founder is on the XMPP (Jabber) committee. It works from any device and on any site. 70% of companies have chat on site, but conversation must start on their site, until now. The chat market has silos of incompatible systems, their click to chat, keeps customers on site, while engaging with multiple vendors.

Wantworthy-avg woman spends 25 minutes browsing shopping sites/day for entertainment. Company's objective is to pioneer 'time shifted buying' via a portable wish list invoked from a widget downloaded to your toolbar

Dispatch- centralized cloud management. Drag and drop amongst your services and share content www.dispatch.io In a world where I am now using five different cloud services, having one unified layer, with social sharing capabilities is compelling. I have signed up for the beta and anxiously await its release.

Coursekit- social network for education. Going after Blackboard. A place for students to download files, and stay to discuss/share content. Free download for individual teachers.

Platform philosophy

Attached is a great post from a Google Engineer, Steve Yegge, which appeared in SAI today. The author is extolling his managers/team to think of GOOG+ (and other applications) as a platform which enables others to write to it, creating a win/win for all (though not necessarily a proportional win/win).

A new set of ears and eyes

Over the past couple of months, I have noticed a series of iOS applications which utilize the camera attached to a phone or iPad that bring order to the 'chaos' around me. In each of the applications the image taken by the camera is compared against a specific database of objects (or GPS location) to give search results without ever invoking a browser. For example:

Zagat updated their search capability with an eye icon on the top of the application. If you scan your camera around a street, up comes the summary restaurant rankings, with no clicks, and no 'formal' search invoked

Fooducate uses bar codes to give you nutritional rankings, and suggestions, for packaged items

Skin of Mine uses pictures of your skin (moles, spots, etc) from your camera to search against their database of  skin disorders and either analyzes the issue, or offers to put you in touch with a professional for a consultation. Nice lead generation for the derm folk, and a streamlined process for the consumer

LeafSnap, referred by my farmer friend Ron, enables you to identify various trees by snapping photos of their leaves.

Google Shopper, is a mobile search tool that relies on a camera for searches on items outside of their 'Featured items'. For some items, such a books, no bar code is required, as it scans the image of the cover and matches it against their database. It's really amazing to see it in action and it's GPS abilities increases its relevance. eBay also lets you search for scanned products via RedLaser (acquired last year) or its own application.

I suspect that camera and/or voice enabled applications such as Siri (acquired by Apple) or Dragon Go! are now well along the way of fundamentally changing a user's interaction with their computer/phone/tablet. From past experience, we know that when the user interface substantially changes, such as with Windows, various browsers, or the iPad, tectonic market shifts follow.

Random bits

Interesting people have raised a number of points that I need to think about:

1. Now that Facebook has 600 million members, and has an ARPU (Average revenue per user) of $3.00 I suppose that majority of future revenue growth won't be centered on new member acquisition, but on ARPU. Shifting their emphasis to establishing a platform has energized a development ecosystem of tens of thousands. Inevitably, they will have to declare which applications (e.g. payments) will be subsumed into the platform. I would not be surprised if they want to grow ARPU by 5x over the next three years.

2. Vendors in the online commerce arena are innovating at a tremendous pace. The LA Times utilizes Groupon to power their platform, the NY Times is working on their own group buying product, and local papers, such as the Bergen Record have entered the market too . To top it off, MSFT has partnered with The Dealmap, which aggregates group deals by city. Moreover, Paypal is now offering 'exclusive' sales to customers via Ruelala Customers, (e.g. SMB's) now have a raft of choice and group buying margins surely will compress....no doubt more innovation is on the near term horizon

3. With the FTC about to open hearings on privacy, many behavioral advertising companies, which rely on cookies, will be under pressure as sites cleanly give viewers the opportunity to opt out of being tracked. At the same time, many sites are investing greatly in adding engagement tools, which enable far greater on-site profiling. It just may be this is the best thing that happened to many of the now protesting vendors.

4. Isn't it fascinating that people often choose less quality and greater price? For example, streaming video offers less quality than DVD's, MP3 is has lower fidelity than analog systems, and digital cameras have less resolution than analog. Luke Williams of Frog Design, has a number of provoking thoughts about innovation and the way that changing behavior is paramount to success by new vendors. He says that companies that concentrate on incremental change are eventually doomed (I felt he was talking about MSFT).