Amazon Web Services Outage: Causes And Remedies

If been a big fan of Amazon Web Services (AWS) because they lower the costs of startup experimentation. I’ve sponsored their events, judged their startup competition, etc. I have friends on the team. I’ve also had frank conversations with them about service level agreements and what it means to be an infrastructure provider in a mashup world. Mashups increase the need for high availability and uptime. If the user experience of a mashup application requires, say, five web services from three separate companies to be available the overall probability of failure goes up subtantially. it’s the weakest link in the chain argument.The Net learned this the hard way yesterday when multiple AWS services (S3, EC2, SQS, Simple DB, etc.) had a multi-hour outage. The problem was exacerbated by the fact that, internally, various AWS services depend on one another and especially the storage service, S3.It looks like the cause for the outage was a particular use pattern of S3:

What caused the problem however was a sudden unexpected surge in a particular type of usage (PUT’s and GET’s of private files which require cryptographic credentials, rather than GET’s of public files that require no credentials).  As I understand what Kathrin said, the surge was caused by at least one very large customer plus several other customers suddenly and unexpectedly increasing their usage. 

I would highly recommend for anyone who is building a developer community or providing SaaS infrastructure or relying on SaaS infrastructure to take the time and read the many posts on the AWS forums about the outage. You hear the real pain and frustration of people whose businesses depend on AWS. The key complaint was not that the service failed–failures do happen–but that Amazon was not prepared to engage with the developer community around the failure.

It’s AmazING the fact of having no info on what’s happening. Absolutely unacceptable. Come on, people on this forum are all tech guys, so we understand that bad things happen from time to time. However, you MUST be transparent with your customers and give them details on what’s going on (yes, we want to know exactly what’s happening and not a standard response like ‘The issue is resolved’). In fact, it is not. So please, scale these complaints to the right person and post the technical explanation of the issue as soon as possible.

Jesse Robbins over at O’Reilly has a good post comparing how Amazon dealt with the situation to how Salesforce responded to its infamous outage a couple of years ago. I’ve also blogged before about how SaaS brings increases responsiblities.All in all, Amazon worked very hard to get the issue resolved and the community was thankful for their efforts.

As I said before, you need to be transparent with your customers. No service can provide 100% uptime. It’s a fact. No matter if u have a redundant anycast network or supercalifragilisticexpialidocious elastic clouds. I just want to get notified and know what’s exactly happening. Nothing else. That said, the issue was resolved very fast, so you should be very proud. Hats off to Amazon’s IT staff.

Posted in amazon web services, SaaS, startups, Web 2.0 | Tagged , , , , , , , , , , | 8 Comments

Consumer-Selected Sponsorship

Sponsorships are growing as an advertising concept and the most interesting sub-segment is the one where consumers select the sponsors they want to be associated with. One of our companies, AWS Convergence Technologies (the folks behind WeatherBug) was a pioneer in this space. They even pioneered the Sponsor Select Network, which opens this for third parties.

An interesting twist to the model comes when you combine it with social incentives. The latest arrival (I saw the news in PaidContent today) is ArcheType Media / SocialVibe. They let brands sponsor your social network profile. Here is an example (on their site and on MySpace). The bait is doing good and getting goods.

Posted in Digital Media, Facebook, MySpace, Social Advertising, Social Commerce, startups | Tagged , , , , , , , , , , , | Leave a comment

Shopping vs. Buying

It’s Valentine’s Day and blogging about e-commerce just feels right… Keep your fingers crossed for ice.com, one of our recent e-commerce investments. It’s a big week for their business.

I had a good discussion today with a friend of mine who was one of the founders of Mobissimo (travel search) about the difference between shopping and buying.

Much of the Internet is optimized for buying: if you know what it is you are looking for, everything from Google to comparison shopping engines will help you find it quickly and at a reasonable price. Search is a great metaphor for this. Tell me what you are looking for. Here is where you can find it.

Shopping is about the “if” part above. It is about product discovery. Discovery is a big deal because it happens before an intent to buy is formed. How a purchasing decision is framed during the discovery process may determine which product ends up being selected. Great sales people everywhere know that cold.

OK, everyone wants an iPod (search works great for that) but does everyone want a Bob jogging stroller? What if they are just looking for a jogging stroller but don’t know what kind they want to buy? The Google query results are just not that helpful to me. JoggingStroller.com is a great site but how do I know they cover enough of the universe of jogging strollers? How do I know I’m not missing that One Great Jogging Stroller? I can open a few more sites but they are all geared towards buying and not towards shopping.

So I go to the thefind.com and search for jogging stroller. I get 12,189 results from 540 stores. I’m not sure about you but to me that answer is both annoying and depressing. Annoying because I bet there aren’t 12,189 types of jogging strollers out there. It shows that thefind.com has good crawling tech and pretty poor equivalence matching algorithms. Depressing because those kinds of numbers just make me feel like I’ll never be sure I picked the right one. (That has to do with the paradox of choice.)

Determined, I go to become.com. The same query delivers 687 products. That’s better. I like products. I’m looking for products. I wasn’t looking for “results”, which is what I got on thefind. Beyond that, become.com doesn’t offer any meaningful help in finding the right jogging stroller. Also, as Siva from thefind points out in the comments to this post, there is a qualitative difference in the result sets between thefind and become.com, though I wonder whether the average consumer will know and understand that difference.

Jonah posted in the comments section that trying this out on shopwiki produced 387 results, all legitimate products. Per Siva’s comment above, I’m not sure about shopwiki’s business model so I can’t say whether this result is good or bad. The user experience on the site doesn’t help much with shopping, though.

These sites miss something that any good salesperson knows about. The right match between a buyer and a product is as much about the buyer as it is about the product. Ever tried to buy a digital camera at Best Buy? The good salespeople are curious about your life and how you want to use the camera. They are not just building a relationship (though that’s important also). They are restricting the product set that makes sense for you in order to simplify the comparison shopping process. They want to know enough about you in order to present 2-3 reasonable models for you to choose from. The bad salespeople give you an earful about the specs of all top-selling models.

Which sites out there do a good job of discovery, i.e., helping people figure out what to buy as opposed to helping people buy what they know they want to buy?

Tagged , , , , , , , , , , | 12 Comments

Folksonomies For Email

I gave up on using folders for email management a while back for a number of reasons:

  • No single hierarchy works well for me. I don’t think I’m unique in that respect. Humans generally have a hard time with cut and dry delineations, which is why uber-taxonomies have mostly failed but folksonomy usage is growing.
  • Folders complicate search and discovery as they create sometimes artificial buckets of information. When you type in a search query, would you like Google to ask you which parts of the Internet you’d like the results to come from?
  • Last but not least, putting things in folders requires constant action. For example, I receive in email about Allurent (one of my companies). Say, I’d have to put it in the Allurent folder. I reply. A response comes to my reply. I need to put that in the Allurent folder also. And so on.

The solution is to take a page out of Gmail’s book and use categories/labels/tags and the absolute minimum number of folders (for me it’s ultimately one as I’ve adopted Inbox Zero). There is no single hierarchy. Search/browsing works well. And I don’t need to always take an explicit action because categories propagate out with email messages. Well, at least they used to.

Enter Outlook 2007 where the good folks at Microsoft decided that showing the rest of the world what categories you apply to your emails is a bad idea. Not sure about you, but I tend you use fairly benign category names and wouldn’t mind sharing them with the world with my outgoing email, provided that I get them back with any responses and therefore don’t have take any extra effort to categorize (pun intended) incoming emails. For a discussion of how Outlook 2007 differs from Outlook 2003 in these respects, take a look here.

For those of you who, like me,  want categories to (a) go out with sent messages and (b) be accepted with received messages, you have to do the following:

  • By default, Outlook 2007 installs and enables a rule to eliminate incoming categories.  If you don’t want Outlook to strip out incoming categories, rule “Clear categories on mail (recommended)” can be disabled via Tools > Rules and Alerts.  This will enable Outlook 2007 users to see the category assignments made by POP users on Outlook 2003 or earlier (and I assume other non-Outlook email clients as well). (quoted from the article linked above)
  • You need to modify a couple of registry settings. In HKEY_CURRENT_USER\Software\Policies\Microsoft\Office\12.0\Outlook\Preferences, create (they probably won’t already exist) DWORD values SendPersonalCategories and AcceptCategories and set them to 1. I got this tip from Microsoft. You can download a registry file to make this fix from here.

There are two other annoyances with how Outlook 2007 handles categories. First, it doesn’t offer an option to set categories on outgoing emails (since they decided it was a bad idea in the first place). Second, the interface for picking categories is not designed for dealing with hundreds of categories. I developed a small module (technically, a form region–see pic below) that makes it easy to type categories during sending, reading and previewing of messages and tasks. It makes it really easy to work with folksonomies inside your mail system.

image003

Here is how you can use it:

  • Download form-regions.zip.
  • Extract it to %userprofile%\Application Data\Microsoft\Outlook
  • There are three files there. Double click categories.reg to load the appropriate registry settings.
  • Restart Outlook.
  • You can modify categories.xml file and set the showCompose, showRead and showPreview elements to 0 or 1 depending on where you want to see the categories form region.

Word of warning: this has only been tested on my system. However, there is absolutely no code–form regions are built using just configuration information.

Now I have to go back and deal with all the email that has accumulated while I built the form region and wrote this post. 😉

Tagged , , , , , | 10 Comments

Viral Dictionary

I met the folks at Pandemic Labs a few weeks ago at a MITX event. Great small team–a technology-leveraged agency specializing in viral and social marketing. Their new site launched over the weekend and the viral dictionary caught my attention–I guess the science of viral distribution requires a new vocabulary.

Posted in Social Advertising, social media, startups | Tagged , , , , , , | 2 Comments

MIT 100K Startup Competition Kicks Off

The mother of all university startup competitions, the MIT 100K business plan competition, kicked off last night at the funky Stata Center at MIT. The competition takes three months–the finals are on May 14. I had a chance to talk to many participants. It’s truly wonderful to see aspiring entrepreneurs attempting to tackle big problems without the burden of conventional wisdom–the knowledge that these particular problems “cannot be solved”.

MIT 100K Competition Kick Off

Polaris Venture Partners has a longstanding history of supporting this event because of our tight links to MIT. I am a judge in the IT/Web 2.0 track this year. My partner Amir Nashat is a mentor.

The format has evolved to multiple tracks in order to force more diversity in the later stages of the contest. (There was a tendency to follow either the hot trend of the day or go for “greater good” winners, as one organizer told me.) The current format is outlined below.

image001

Posted in startups, VC, Venture Capital, Web 2.0 | Tagged , , , , | 1 Comment

Strategies For Startup Success

I gave at talk this morning at Yankee Group. It was part of the SmartSpeaker series that CEO Emily Green has been organizing for a few months now. Many months ago, over breakfast, Emily asked me to give a talk on strategies for startup success. Foolishly, I agreed.

The trouble with the topic, as I’ve written before, is that there are very few strategies for startup success with a high batting average. And startups, unlike blackjack hands and day trades, take a good piece of an entrepreneur’s life. A strategy that is successful 55% of the time will work wonders on Wall Street and in Vegas but may cause an entrepreneur to repeatedly fail her entire life.

Hey, let’s not get gloomy. There is a silver lining. It’s called luck and timing combined with assembling a great team that continually performs quick & cheap experiments. Esther Dyson says: “always make new mistakes”. Make them quickly and cheaply, I’d like to add. 

I ended up giving three talks in one at Yankee today:

  • Strategies for Startup Success
  • Why Do Startups Succeed or Fail
  • The Startup Environment

The questions from the analyst team were excellent. I hope to continue the discussion here.

Posted in startups, VC, Venture Capital, Web 2.0 | Tagged , , , , , , , , , , , , , | 4 Comments

Getting Social Platforms Right

Since I’m a platform guy by background, I’ve been very interested in how social infrastructure, i.e., social platforms such as Facebook’s F8, Bebo’s clone, Google’s OpenSocial and MySpace’s recently launched developer platform, is going to evolve. Some recent good reading on the subject comes from Max Levchin who as founder/CEO of Slide knows a thing or two about this.

Max’s first post covers the launching of a social developer platform with ten common-sense rules. (My only question is: why always ten rules?) Two key takeaways:

  • A platform will only succeed if it makes its developers successful. I wholeheartedly agree and hence my Ecosystem Test.
  • Be humble, open and embracing of the developer community. That’s how we built the developer community at Allaire. It was and still is, one of ColdFusion‘s greatest assets to this day.

The second post, really an essay, goes deeper into developer incentives in social networking platforms starting with the hypothesis that “designing a social platform is in some ways similar to designing a competitive multi-player game.” Max offers a Darwinistic/behavioral view of ecosystem management which makes sense in a social environment where multiple developers/apps compete for the attention of consumers.

Perhaps because of the Slide perspective, the incentive analysis is primarily focused on valuable distribution.

It’s worth pointing out that ultimately, until non-advertising business models are devised for social applications (and probably even after they are) valuable distribution (reach + frequency) is going to be the main underlying goal for all developers, commercial and otherwise.

My take is that non-advertising business models are just around the corner. The key is social advertising/social commerce where one is leveraging the distribution power of millions of consumers. There the developer controls a social platform puts in place are somewhat less important.

Posted in Facebook, MySpace, Social Advertising, Social Commerce, social media, Web 2.0 | Tagged , , , , , , , , , , , , , , , , | Leave a comment

MySpace + Google >= Facebook?

The much expected, joked about and pre-announced MySpace developer platform is finally here. Well, not quite:

As Aber mentioned, we’re currently in a sandbox phase. Here are the important points to understand about this phase:

  • Your application profiles are private
  • Your apps install on real MySpace accounts, but are limited to three installations
  • If your app renders a profile module, it will only be visible to users who have that application installed

So, essentially, the platform is in a public beta allowing for some back’n’forth between MySpace and the developer community. This is a good change in stance for MySpace, which has in the past taken a rather confrontational stance with respect to third party developers. I hope they have hired (a) some great platform architects and (b) some great developer community managers or else it’s unlikely that the outcome of the beta period will lead to significant improvements.

Always more of a media/lifestyle company than a technology company, MySpace has taken a long time to come to terms with opening up. Some months ago I had a conversation with one of the CXOs there about this very issue. From his comments it was clear that there was tension at the highest ranks about how to proceed. I guess Facebook’s growth accelerating and the company becoming progressively closer with Microsoft helped MySpace, whose own growth declined, realize that they need to open up and partner with a tech savvy power (Google, which is helping with OpenSocial and Caja). Good move, though I expect Google to benefit more in the long run than MySpace as it sets up OpenSocial for success.

Although Facebook and MySpace remain quite different, it will be interesting to see what developers do on MySpace. Let the widget games begin.

Posted in Digital Media, Facebook, Google, MySpace, social media, Web 2.0 | Tagged , , , , , , | 2 Comments

Oh, Well…

…and a cheers for comebacks.

Posted in Life | Tagged , | Leave a comment