Friday, December 18, 2009

Google's Vertical Search

I have been writing for a while about how Google's search results stink, and how vertical search is far better. Some companies might continue on their existing trajectory towards failure, but not Google. Over the past six months, Google has been steadily adding vertical search features, although of course in a very Google way.

Rather than having the user pick a category such as music or stocks and then search, Google guesses what vertical your search is in and puts that result at the top of your search results.

Examples include:

Addresses -> Maps

Businesses -> Map & Reviews

Stock Quotes -> Stock Chart

Music Artists -> Artist Tracks

For some types of searches, Google just shows the results in the Google Suggest feature, so you don't even have to click search! Examples include:


Flight Status

Missing are Facebook & LinkedIn for people searches and shopping comparison for product searches. I can understand Google's reticence about integrating successful social networks since Google lacks one, and Facebook and LinkedIn results inevitably end up near the top. The AdWords free-for-all based on product searches is not one I think they will ever optimize for the enduser.

The Google vertical search progress is impressive and I have found myself using vertical searching less and less and going back to the Google mothership. Well done.

Thursday, December 03, 2009

The Microsoft Comeback

Everyone is writing Microsoft off, including the New York Times epitaph last month. I think that it is premature to discount Microsoft.

The Cloud

The overall transition to cloud computing has been incredibly slow. Most small and medium sized businesses are still running their own mail servers, file servers, wikis and applications, and the fact remains that most of these businesses are running Microsoft software on their desktops and servers. When these businesses finally transition to cloud computing, what are they likely to do? Try to migrate everything to GMail and Google Apps? That is a huge pain and requires a ton of user training.

Microsoft is going to come in and say "we will migrate your Exchange to hosted Exchange, your files to hosted NTFS, your Office to Hosted Office, your SharePoint to hosted SharePoint and your .NET/MS SQL Server apps to Hosted Azure. All for less than the cost of an upgrade cycle."

And then Microsoft will turn their license revenue stream into a monthly revenue stream. Which means they will no longer have to add a bunch of useless features and change file formats just to get customers to "upgrade" to a new version of Microsoft software.

Windows 7

I have been running Windows 7 on my home machine for about a week and I am very pleased with it. I use a Mac at work so am shifting back and forth between the two operating systems and there is not much different. What always hoses Windows machines is adding software, but I don't add software anymore and run the same suite of stuff on both machines (Firefox, Thunderbird, Skype, MS Office, and Acrobat Reader). As many have noted, the window management on Windows 7 is far superior to Mac OS X.


As I have been posting for years now, Google search sucks, and vertical search is much better. Bing is very good at vertical search. Who cares if the SEO spam links that are below it are "better" than Google's when users have already found what they are looking for at the top of their screen? When "search" = 1 result, the game is over.

Innovating is Hard When You're Microsoft

Look, everyone loves to hate Microsoft, even when the do actually innovate. Back in 1998 I used a NEC Mobile Pro 770 running Windows CE. It had instant on, 8 hours of battery life, a 640x200 pixel screen, almost fullsize keyboard, Internet Explorer, and a lite version of Microsoft Office. It was a GREAT machine. 10 years later everyone is heralding the advent of NetBooks, when Microsoft was the innovator of the segment.

But Copying is Easy!

What Microsoft is really, really good at is copying, and it does not take much to copy Google Docs, Mac OS X, and Google Search. And it does not take much to give a Microsoft software user a hosted experience that is very similar to their desktop experience. They could still royally screw this up, but I think they are starting wake up in Redmond.

Thursday, October 22, 2009

Feed the Feed - How to Populate the Real Time Web

Now that search engines like Google and Bing are incorporating status updates from Twitter and Facebook, getting your content into peoples' feeds has become more important than ever. Sites like The Huffington Post that are Facebook Connect enabled are seeing tremendous traffic growth from just newsfeeds, and this will be magnified by their links showing up in search results. So how can a Facebook Connect enabled site easily add engagement features that entice users to post links into their newsfeeds?

We at Transpond call populating the real time web "feed the feed", and the latest version of our Write Once Engage Anywhere technology offers Apps that automatically detect if they are embedded in a Facebook Connect enabled site, and prompt the user to share their engagement activity to the web.

For the CBS Emmys feature, Transpond powered the nominees poll, red carpet poll, fashion face-off, videos library, and image galleries. All of these features were tied into Facebook newsfeed updates.

When voting on a gown in the red carpet poll there is a "Publish to Facebook" checkbox placed directly below the Vote buttons.

Clicking on a vote button indicating whether you like or dislike the gown then prompts you to share your thoughts with your Facebook friends using the standard Facebook publish dialog.

The feed story than shows up in your news feed and on your friends' Facebook homepages and mobile apps.

If the user is not logged in via Facebook or the site is not Facebook Connect enabled, embedding a Transpond App prompts the user to share their activity with Facebook, Twitter, and MySpace icons:

You can easily "feed the feed" with your content by adding Transpond Apps to your website and Facebook/MySpace fan pages. Integrating deeply into Facebook Connect enabled sites is just another example of how transforming an App into a native part of its host is a far better solution that generic solutions such as embedding Flash.

Tuesday, October 20, 2009

How Facebook Will Become the Top Ad Network

Facebook Connect is growing like a wildflower through websites, dramatically increasing registrations and logins now that users do not need to remember an account and password, and also driving traffic when users publish website activity to their Facebook newsfeeds and draw clicks from their friends.

Now when a user goes to a Facebook Connect enabled website, the website contacts Facebook to get their Profile photo and name.

Since Facebook knows a lot about people from their Facebook profiles, it lets advertisers place super targeted "Social Ads" on These ads are already drawing over $500M in revenue for Facebook.

The next logical step is for Facebook to let Facebook Connect enabled websites show these same targeted ads and share revenue with the websites. So when users log onto Facebook Connect websites, they see the same highly targeted ads that Facebook shows them when they browse Facebook. And users are more likely to click on these ads on websites that they are browsing than on Facebook where they are engaging in a social experience.

Transitioning ads from one high volume website into a network of websites is the same thing that Google did. First Google offered AdWords that showed ads based on what a user was searching for at, and then it offered AdSense that let websites show the same ads based on what was on the page a user was looking at. Google then augmented its website ad network by purchasing DoubleClick's display ad network.

Facebook's ads are much more targeted than Google's ads. A Google ad at can assume that the user is a male aged 20-35 and then randomly show a Toyato truck ad or a movie ad for the new Farrelly brother movie. A Facebook Connect ad would know that the user (or a lot of the user's friends) are a fan of Something About Mary, that the user had checked out Facebook Connect enabled earlier in the week, and show the Farrelly movie ad. If the user is not logged into Facebook Connect, Facebook will at least know which types of ads are doing better for different demogaphics.

Facebook surprised everyone with its rapid revenue growth to $500 million in Social Ads... its evolution into a Social Ad Network will accelerate revenue into Google territory.

Tuesday, October 13, 2009

Will Twitter Get 250m New Users or Will Facebook Copy its Features

Which of these is more likely to happen:

(1) Twitter adds 250 million new users that use it regularly.

(2) Facebook copies all of Twitter's features.

As I wrote in March, it is not difficult Facebook to copy all of Twitter's features, and they have since shamelessly done so.

Think of your stay-at-home mom in the midwest that is using Facebook to keep in touch with all her friends. She wants to subscribe to businesses to get deals and see what her favorite TV stars have to say. What's she going to do? Try out Twitter with its lack of inline photos and videos and unintuitive nonthreaded comment streams, or become a fan of their Facebook fan pages and have everything cleanly inserted on her Facebook home page? No brainer.

As reported in Techcrunch, Comscore shows Facebook continuing to grow while Twitter is flattening out:

Twitter is great and all, but it needs to fix its UI ASAP if it wants to grow. And no, expecting moms in the midwest to find random 3rd party applications does not count!

Thursday, October 08, 2009

The Accelerating Speed of Business

The way that Internet companies do business has fundamentally changed over the past couple of years. Platforms like Facebook and the iPhone have accelerated from nothing to ubiquity in just a year or two. How can a business keep up?

Well how does Facebook itself operate? Scrum style engineering. Bold new features that are pushed out and then iterated. Marketing is accomplished via blogs and PR. Sales are self-service, and some features such as Apps are free.

Here's a summary of how the new way of operating impacts each functional area of a social Internet company:

LeadershipLaunch and iterateConsensus
EngineeringScrumWaterfall MRD/PRD/spec/etc.
MarketingBlogs and TwitterWhitepapers, emails, and webinars
SalesSelf-service freemiumDirect sales and limited trials

Recently I have been interviewing a lot of executives. The most challenging thing is that a lot of really good, solid people are steeped in the way we all did business 10 years ago. It is hard to figure out who can adapt to this new way of doing business.

This market moves very quickly. There are no visionaries here, but people who know the market well. When it is clear to the people who know the market well that something is about to change, what should a company do?

One way to accomplish this would be to have a series of meetings where the team would seek input from everyone impacted and attempt to achieve a consensus that a new feature should be added. Then marketing would meet with customers and create a Marketing Requirements Document that the team would iterate on and approve. The next step is a Product Requirements Document that the team would again iterate on and approve. Then engineering can create a detailed specification to match the PRD and meet with marketing to ensure that the specification matches the PRD. Up next is user interface design and iteration. Then engineers can implement the feature for the next big push of the website.

As I am sure you have guessed, companies that operate this way are dead or dying. The competition has just shipped 10 features in the same amount of time.

The new way to do this is to quickly create a wireframe of how the new feature should work. A cross-functional product team should in real time edit this spec and remove functionality until the bare minimum for a functional feature is left. At this point there should be a go/no go gut check decision now that there is a concrete wireframe to look at. The wireframe is then quickly skinned and implemented, and shipped with the next maintenance push of the website. The feature is announced via blog and Twitter. If people don't respond to the feature, the company has learned very quickly with minimum wasted time that it needs to move on. If people respond to the feature, the team can then iterate on the feature based on user feedback now that there is traction. Once there are sufficient users and higher level features to add, there is a self-serve upsell opportunity for the feature.

I visited Zynga this morning, and one of their company mottoes is "Zynga Fast". Facebook moves fast. Zynga moves fast. Companies in the social and Internet ecosystems need to move as fast if not faster. Which means people at these companies need to be just as agile. The way we did stuff 10 years ago no longer applies.

Saturday, August 15, 2009

Commercial Open Source Redux

A big congratulations to SpringSrouce, a commercial open source company with an estimated $20M in annual revenue after years of effort sold for 20x trailing revenue. Pundits such as Matt Asay point to this as another success, after pointing out that in the Commercial Open Source Failure piece I wrote for BusinessWeek that I also missed the XenSource and Zimbra exits, two companies with <$10M in revenue that also sold for >20x trailing revenue.

I am in full agreement that these companies all had great technology in interesting spaces, but lets be clear that these are technology acquisitions, not business acquisitions like MySQL and JBoss. I do not think that Citrix and Yahoo! are patting themselves on the back about their acquisitions of XenSource and Zimbra. At least VMWare using SpringSource as an accelerator into cloud computing makes a lot of sense.

As the New York Times pointed out last week, the trend here is that enterprise IT spending has fallen into the gutter and is not going to climb back out. And even with the open source marketing message, commercial open source companies are selling proprietary software with direct sales, marketing at tradeshows, and supporting it with support staff into a market whose growth is almost on par with the rate of inflation.

These companies are fundamentally no different than the enterprise software companies that preceded them, which is the main point I am making and the point that no one in all of these "rebuttals" wants to address. Yes, a few will be successful as businesses, and for the rest there will be technology acquisitions, some exorbitant. But in the end the enterprise software market as we knew it is history and is morphing into cloud computing.

Monday, July 27, 2009

The Commercial Open Source Failure

This post was also published in Rachael King's BusinessWeek column.

So what happened to open source as a business? There was a wave of “commercial open source” companies that were going to change the world, including SugarCRM, Alfresco, Jasper, Pentaho, and ActiveGrid, a company I started in 2003 to bring open source software into businesses. Each of these companies were going to lower costs with the open source business model and displace existing vendors in categories such as CRM, Document Management, Reporting, and Business Intelligence.

It’s been over six years, and no commercial open source companies other than Red Hat, MySQL, and JBoss have had liquidity events. So what happened? Oracle and IBM, which derive the vast majority of their software revenue from proprietary software, have an increasing share of the software market. And there’s a bunch of commercial open source companies still trudging along.

1. The only successful open source companies sell commodities. Linux, MySQL, and JBoss are the only open source companies with significant liquidity vents, and their success is more indicative that operating systems, databases, and application servers are commodities than that open source is a successful business model. No one can succinctly tell you the difference between Linux and IBM’s AIX, or JBOSS and Oracle WebLogic, so why not buy the cheaper one!

2. It costs as much to deliver open source software as proprietary software. Many open source companies exhibit at tradeshows, have salespeople, systems engineers, customer services departments, and on top of that employ the vast majority of developers that are working on their open source project. So their costs are comparable to proprietary software companies that offer free trial versions for lead generation. It’s great that some random guy in Lithuaniacan fix a bug for the commercial open source company, but the headache of maintaining a community and integrating random code patches is just as expensive as fixing reported bugs with your own people. So commercial open source companies have the same cost structure as the enterprise software companies that preceeded them.

3. Selling software is miserable. Selling software to a large business means approal from architecture committees, security audits, achieving approved vendor status, business unit sponsorship, executive sponsorship, etc. Enterprises see IT as a cost center and IT does whatever it can to not have to deal with yet another software vendor, especially a small startup that is likely to go out of business or get bought and then deal with software that’s no longer supported, making the IT buyer look bad.

4. Customers are switching to SaaS intstead of more software. Let’s say you run a division of a company and you need customer relationship management (CRM) or business intelligence. What are you going to do? Call your IT department so they can find software, purchase it, customize it, deploy it, and then roll out to your users, all while charging your through the nose? Or call a service provider like or GoodData and be up and going tomorrow? No brainer.

So when is open source successful? When it’s free and supported by a broad community of developers, not just one company trying to extract revenue. At my own company Transpond, where we help companies reach their audiences and customers on social networks and mobile platforms, we use a ton of open source software including Apache, PHP, Tomcat, MySQL, JQuery, Debian Linux, Quartz, Eclipse, Maven, and more, and we haven’t paid anybody anything other than contributing code to some of these projects. The point of open source was for people to share the costs of developing, debugging, and deploying common infrastructure. That does not mean that every successful open source project can sustain a commercial company, especially when they are delivering complicated applications rather than simple plumbing.

Thursday, July 23, 2009


I remember during the .com how many companies wouldn't say what their products did or how much they cost on their website. The one that particularly comes to mind was E.pithany, whose website was laden with benefits and customer praise, but after spending 15 minutes perusing their website, I couldn't figure out what they did.

Business-as-a-Service is one of the best aspects of Software-as-a-Service: the commerce is frictionless. We at Transpond consume many services ranging from email to source code control to customer service systems, and we have never had to talk to anyone, negotiate a contract, or deal with anything complicated. We look around online, find a vendor that we like, check out their terms of service, use their free trial, and if we like it, we start paying monthly for the service.

What I have found surprising when looking for services is we still run into some that don't answer the three basic questions that every business should answer:
  1. What do you do?
  2. How do I get it?
  3. How much does it cost?

While looking for a SaaS subscription billing service, I checked out one that has received a lot of press lately but did not have any indication of pricing or a guide on how to integrate the technology. I sent them an email stating specifically what we needed and asked for information on pricing and integration. They wanted me to schedule a call with a sales rep! So perhaps they are delivering Software-as-a-Service, but definitely not delivering Business-as-a-Service. Of all companies, why don't these guys let me choose what I want from a menu and then start billing me for it!

At Transpond, we definitely subscribe to the Business-as-a-Service philosophy and answer these three questions quickly and simply:
  1. What do you do?
    Home page headline: Easily Create Apps for Social Networks, Mobile, and Connected TV.
  2. How do I get it?
    Home page and every page: "Get Started" button or link.
  3. How much does it cost?
    Products/Overview: Price list.

While it seems the entire web is Business-as-a-Service now-a-days, we should recognize the pioneers, which include Google, WebEx, PayPal, Intuit, and Salesforce. These companies were all very clear about what they did, how to sign up, and how much it would cost... and they all grew into huge businesses, some without ever publishing a phone number!

Tuesday, July 21, 2009

Real Time or Popular Recent Time?

On one hand the "real time web" is great, but on the otherhand it is overwhelming. I find myself checking in and out of the stream. In some ways, I miss Facebook's old weirdly algorithmic way of summarizing what everyone was up to on my homepage. There is a bit of that sandwiched to the right of the Facebook homepage, and Tweetdeck try to guess at popular tweets, but I find these interface lacking.

TweetMeme and The Hype Machine are interesting takes on what is popular on the overall stream.

What is ironic is that Facebook redesigned to mimic Twitter, but there are popular Twitter sites that are mimicing how Facebook used to work. Something in between the two is likely to be the eventual answer...

Introducing Transpond

As covered last month in the Wall Street Journal, VentureBeat, and other media, iWidgets is now called Transpond and has a new url:

Why the name change? Since we started the company to deliver what we then called "native widgets", the market has developed a name for what we were doing, which is "Apps". Widgets became pigeonholed as blocks of Flash that didn't do much, while the kinds of units our company helps you build for Facebook, MySpace, iGoogle, and now Palm Pre, are definitely called "Apps".

So rather than hold ourselves back with our name, we update the name to Transpond, which in the telecommunication field means to "amplify a message on a different frequency." Well that's what we do, we take your message and amplify it virally onto new platforms!

In addition to our rebranding, we have also released an updated site that makes it supereasy for marketing and web folks to deliver native social network and mobile apps. The screen shot below shows a simple three step wizard that you can use to quickly deliver a fully branded, native App for Facebook, MySpace, iGoogle and Palm Pre.

Over the next few weeks, we will add numerous viral types of applications, reintroduce drag-and-drop styling, and add additional native platforms. Stay tuned!

Thursday, May 14, 2009

Woah - Apps Are Expensive

MediaPost posted an article about a new Forrester research report on the cost to create an app: "Among the initial considerations is cost: Forrester estimates that mobile apps range from $20,000 at the low end to $150,000 for more sophisticated ones. Since most marketers don't have the in-house resources to build apps, the report suggests turning to experienced mobile app developers."

This is the same price range I have seen for custom Facebook and MySpace apps as well. Flat and decreasing web traffic is forcing publishers to syndicate their content into social destinations and mobile platforms, but at a low-end of $20,000 a platform, that’s a lot of money for a publisher to get their content out.

Writing a custom app is like making an e-commerce website n 1995 - you had to use CGI and add your own thread pooling, session management, etc. This is exactly why we started iWidgets, to provide an easy, cost-effective on-ramp to the future. The Forrester report is the first independent confirmation I've seen of what it costs to deliver even the simplest app.

Wednesday, May 06, 2009

Journalism vs. Blogging

I have a couple of journalist friends that have been lamenting the demise of journalism in general and newspapers in particular. Both of my parents were journalists (AP, UPI, NYT, Time-Life), so I was brought up with and can appreciate journalism as we once knew it.

However, journalism itself changed drastically even before blogs and the Internet. As "the news" became a big business, there was this obsession with making the news "objective" by massaging facts to appease the overall audience. Telling people what they didn't want to hear meant they would watch/read less, which meant less advertising revenue. We saw this during the prelude to the Iraq War, and even today with editorial boards going into contortions over what to call "torture." Dan Lyons pointed out how the business press is sycophantic in its reporting of Apple.

The turning point for me was watching CNN last year during the oil crisis. There was a story on the price of oil, with a breakdown of how the $110 for a barrel of oil was distributed between gas stations, refineries, oil companies, taxes, etc. At the end of the piece, there was an interview with an oil industry lobbyist that stated that the price of a barrel of oil had no relevance to the profits of the oil companies. Did CNN question this or state that this was in conflict with the facts that they had just laid out? No. That wouldn't be fair. Maybe they would lose their access to the oil lobby!

When the mainstream news media refuses to question or challenge anything at all so as to not annoy some people, it is no longer journalism in my opinion. If the reporter is looking at a blue car, does he/she have to report that some people think that the car is red, rather than reporting that some people are color blind? Is that really venturing an opinion?

Nowadays, I read blogs for my news. I read tech blogs, design blogs, liberal blogs, conservative blogs. I see which ways different people are spinning a story, and I form my own opinion. Guess what: Sarah Palin has no clue about foreign policy. The bank bailouts have no metrics associated with them. It's called torture if I would be arrested for felony assault for doing it to someone. There are some things that can be ascertained by simply observing, and should be reported as such. It does not matter if it pisses people off, just tell things like they are! There is definitely a place for fact-checked, insightful articles, but putting them in a Wolf Blitzer-style, "don't want to offend anyone" context that does not point out the obvious mutes the story and is no longer relevant to anyone.

And stop wasting trees to send text around! :)

Tuesday, April 21, 2009

Oracle-Sun: Best of Breed Doesn't Matter Anymore

10 years ago the enterprise mantra was open systems and best-of-breed. Customers wanted the best server hardware, best operating system, best database, best application server, best middleware, and expected vendors to make that stack work together. Now there is a trend towards vertically integrated stacks from IBM, HP, and now Oracle with its acquisition of Sun.

Apple vs. Microsoft is playing out on the enterprise level, and vertically integrated stacks with a unified experience are edging out open systems. Why is this happening? It occurs to me that each of the layers of the enterprise stack have been completely commoditized - can anyone tell me the difference between AIX and Solaris, Oracle WebLogic and IBM WebSphere, DB2 and Oracle, etc. in a way that would make sense to a CIO? There is barely a difference anymore for technical people, let alone those signing the checks.

Essentially vertically integrated stacks of mainframes and minicomputers were replaced by layered open systems and are now being replaced by vertically integrated stacks of open systems.

The big question for Oracle is whether it can successfully start to sell server hardware and balance monetization vs. stewardship of open source like MySQL and Java. Imagine if you had a gardening service you really liked, and it was bought by Jiffy Lube and they called you up and said that they were going to "improve operating efficiencies" etc. You might consider finding yourself a new gardener.

Sunday, April 19, 2009

Java Cloud Edition

Last week Google announced Java support for Google App Engine. Contrary to Simon Phipp's complaint that "it's wanton and irresponsible" to only support a subset of Java, I applaud what Google has done. Here’s why.

Java is almost fifteen 15 years old, and is a huge amalgam of classes that are no longer necessary for web development. What Simon seems to be suggesting — that Google include the Swing user interface classes, AWT, and Java2D for a cloud-based offering whose only interface point is a web browser? — makes no sense.

Google is solving a huge business problem with Google App Engine's Java support. One of the biggest challenges facing enterprises is that they have a ton of data, a ton of backlogged applications that their users want that are based on that data, and no way to deliver those applications. I learned the hard way at ActiveGrid that scripting languages was not a good way to solve that problem since it required a new type of server to be installed inside the firewall. A couple of years ago I theorized that the problem could be solved with a "data wiki" type of approach that would combine a simple form editor with tiered access controls, which I still think is a viable opportunity but runs up against the challenge that enterprises do not like to buy or install software anymore, especially especially from startups.

With Google's App Engine, enterprise developers can write straightforward JSP/JDO Java applications that can connect their existing back-end databases with Google's Secure Data Connector. The apps can then be moved to run inside the firewall on a standard Java EE server. I imagine Google may even have an enterprise version of App Engine that enterprises could run inside their firewall, just like the Google Search Appliance.

Kudos to Google. Their App Engine for Java should be called “Java Cloud Edition.” They have solved a big enterprise application development problem, and have also helped to refocus Java for the cloud. Cloud services aren’t a mere trend, it’s an evolution in the software industry. Companies that aren’t cloud-based run at a huge and costly disadvantage. At iWidgets we run a Java-based cloud backend for customers, and we are not using Java classes like Swing. Anybody complaining about some of the legacy Java classes are falling by the wayside isn’t in touch with this industry transition.

I think Sun should take the class subset used in Google App Engine for Java and call it Java Cloud Edition.

Thursday, April 02, 2009

Startup 101: Capital Efficiency

You may have read in BusinessWeek or TechCrunch when iWidgets closed an A round in January, despite the tough economic climate. In addition to having a strong product that solves a big headache for content publishers, a big part of why we got funded is our capital efficiency.

iWidgets spent only $1.5M in 18 months to bring our iWidgets platform to market, launch it with a major customer (CBS), and then sign three more customers.

Our leading competitor spent $10M in the same amount of time, just laid off a large portion of their staff, have no business model, and a product that solves yesterday's problem.

While you can’t build a business without a plan, capital efficiency is the often-overlooked key ingredient for a successful startup. Capital efficiency isn’t just about not wasting money, it’s also about not having any fat. Too often, bloated, fatty companies are insulated from market realities — you can’t feel it when the road starts to get rough. So it is not just in this economic climate that startups should be capital efficient, it is always.

Here are the eight cardinal NO’s of capital efficiency, plus the four things you HAVE TO spend money on.

- No Management

A small team of top-notch engineers does not need a VP of Engineering, it needs a technical lead. An Engineering VP costs a bunch of money, and in a fast product cycle does nothing but slow things down. For most of iWidgets' life, the company was composed of four engineers and me. We had a very clear direction: We wanted to integrate website content deeply into popular destinations. But there were a million features that we could do, and we iterated a few times on which ones to include, until some started to stick in the market. We did not have to ship a complete, moon launch product, and we did not have to have big team meetings to align ourselves. I would say, “I met with these three companies, I see a need for X at each of them, let's do a minimal version of the feature and see if they bite.” We did this a few times before we had a product that people needed. Adding a bunch of chiefs to the mix would have only slowed this process down.

- No Sales

A sales person is going to have a quarter to ramp up and set up Salesforce just the way they want it, is going to require sales tools, and is going to come up with a million reasons as to why an account is moving forward. They are never going to tell you your product sucks or that no one wants it. Why would they? They’d be out of a job. The worst thing is, they are shielding you from the market. The CEO needs to hear the “NO.” The CEO needs to get the blown off by a friend of a friend at a large media company. Why? So that he/she can get a sense of the market. It is the job of the founders to get a few customers using the product. A sales guy is not going to do it for you and is instead just going to waste money.

- No Marketing

VCs are going to tell you that you need a marketing person. In my experience, there are very few marketing people that can function in a pre-product company. Their role is essentially market research at that point, and for that you should just use consultants. But I know of plenty of companies without marketing people that have great marketing. At iWidgets, where we have been in the business and trade press countless times, constantly have inbound calls from major media organizations, have a clearly differentiated product, and a business model that prospects like. Once you have all that, a marketing person will do great, because they can actually execute. But a marketing person is not necessarily going to get you all that, and in fact is likely to get in the way.

- No Tradeshows

I have a policy of no tradeshows. They are a complete waste of money, especially if you have to travel and put together a booth. But policies need to have exceptions. We bought a $5K booth at the Facebook f8 developer's conference because it was cheap, plug-and-play, and a great opportunity for us to get feedback about our product since we had just launched into beta. We also thought that we would meet a lot of Facebook folks, but instead our booth was deluged by Google employees working on OpenSocial and iGoogle.

A huge decision that took a lot of hemming and hawing was launching our 1.0 product at the Demo conference last year. The Demo show was $18,500 + travel: super expensive. Ordinarily I would not even consider spending that kind of money. But we were launching with CBS, and the SVP of Entertainment at CBS had kindly agreed to participate in our launch on stage. Given that, we had to provide a venue suitable for such a senior executive, I don't think he would have done it at a free-for-all like the Techcrunch50. As a result, we got inbounds from every television network within two days. I still look back on that and don't know if I would spend that kind of money again, but what's done is done, and I have no regrets.

- No Office

We didn't have an office for over a year, and would have staff meetings around my dining room table. We had a constantly running Skype chat window where everyone could talk to each other. Once the engineering team got to 5 people we sublet a few desks for engineering from a friend of our CTO. Once we had a signed termsheet for our Series A, we got an office, and instead of an office ended up getting a live/work loft because it was cheaper (and better!).

- No Servers

This is a big one for me: Do not install a single server. Use SaaS for everything, even source code control and bug tracking. You will save a ton of cash and will not need any IT people. We use SaaS for everything, including Amazon EC2/Rightscale for our website and backend, Google for email and wiki, CVSdude for source code control and bugzilla bug tracking, BrowserCam instead of a QA lab of machines, Quickbooks Online for accounting, Zoho for CRM, Packetel for Fax, Egnyte for file sharing, and TriNet for HR/Payroll.

- No Administrative Staff

Do it yourself, you will actually know what is going on in your company. I once briefly worked with a guy at a pre-revenue company that needed a ton of infrastructure, including an admin to take notes and log action items during his staff meetings, and even hired a consultant to fire people for him. Needless to say, a lot of money got spent with very few results.

- No Travel

Skype video. Enough said.

And now that you know the NO’s, here are the YES’s, which are just as important.

- Yes Engineers

Hire the best you can find.

- Yes Lawyers

They feel like a rip-off, but hire lawyers that understand startups. And get them to kick some cash into your seed round which covers what they are charging, so it turns out to be a wash.

- Yes Bookkeeping

Use Quickbooks and a part-time bookkeeper, they are cheap and your accounting will be suitable. P&L statements are an important way to keep track of what you are spending money on, and investors are going to want to see it.

- Yes Wireframing

Wireframe everything before building it, it will save a lot of time and mis-steps, and therefore money. I am a big fan of Axure Pro and we even wireframe simple three step wizards to make sure they are understandable and flow smoothly. It is better to figure out early what works, validate new features with customers before they are implemented, and the engineers appreciate getting a straightforward spec that is not going to be subject to too many changes after the feature is added.

As I’ve pointed out, sometimes you have to break the rules to take advantage of a one-time opportunity, but these are the guidelines I’ve developed over the years and I’d recommend them to anyone starting a business — not just “in these tough economic times.”

Thursday, March 26, 2009

The Problem With Old Media is... It's Old Media

I had the honor of meeting marketing guru Regis McKenna a few weeks ago as he is an advisor to our lead investor, Opus Capital. I like to think that we at iWidgets are on top of our social media game, but I went into the meeting thinking that I could get some advice on positioning iWidgets, how to differentiate in a noisy marketplace — in other words, the usual marketing stuff, but this time from one of the all-time marketing gurus. Little did I know that I would walk out of the meeting with a new perspective on the transition to social media and the web.

It was a great conversation with Regis. I described the iWidgets social syndication platform to him and told him about the traction we’ve been getting in online video with customers like CBS. I outlined our move into the music vertical since a lot of songs also had video, and that all of this video and audio was easily monetizable due to the prevalence and acceptance of short pre- and mid-roll ads.

But while video and music uptake has been awesome, I shared with Regis that one issue we are facing at iWidgets is that we have a lot of inbounds and self-service customers with text-based content, accompanied by photographs, that is increasingly difficult to monetize. Usually text content is monetized by adwords or banner ads, and as I have blogged before, adwords and banner ads have abysmal CPM rates on Facebook.

Regis made a pronouncement on the transition to digital media that is still reverbating with me weeks later. "Text is dead," he said. "Those publishers will have to transition to video and audio or they will languish. Focus on the growth businesses."

In one sentence, Regis explained why newspapers are failing, why banners ads are decreasingly effective, and why video ads are booming. Brilliant!

Monday, March 23, 2009

Video Killed the Banner-Ad Star

This post was also published in AdWeek.

Over the past couple of years, the decreasing effectiveness of banner ads has led to increased hand-wringing. When I give talks nowadays, I start off by asking who clicked on a banner ad that day. No hands go up.

Banner ads were all the rage for many years, culminating in 2007 with Google's $3.5 billion acquisition of DoubleClick. But as more people spend more time online, banner ads are becoming the billboards of the Internet.

Billboards can be very effective. Ubiquitous Coca-Cola billboards build the brand. Billboards placed in highly visible locations like Times Square and Giants Stadium attract a premium. But it's quite the exceptional billboard that actually makes you take action -- calling a listed phone number to learn more about the business, for example -- just like it's quite the exceptional banner ad that gets you to click on it.

And well-placed billboards, such as one on a highway advertising a restaurant in a nearby town, can be effective and lucrative, just like a banner ad on a travel site tied to what you're already searching for can catch your eye. But overall, billboards are a niche market, and pale in comparison to broadcast TV and cable advertising.

The equivalent of TV advertising online is, of course, online video advertising.

For a long time, online content was mostly user-generated video -- the equivalent of cable-access channels, the garbage can of TV advertising. But now that major TV networks are distributing premium content over the Internet, users are increasingly watching premium video content from networks such as CBS and Fox, and Webisodic content from sites like and Not surprisingly, a recent report from eMarketer states that "video ad spending will run counter to overall economic developments, rising by 45 percent in 2009 to reach $850 million."

Now that premium video content is available, it's a natural fit for top advertiser dollars. Premium online video has only just begun its penetration. Previous generations of online content have already flowed from source sites such as YouTube and CNN into thousands of embedded widgets, RSS newsreaders and profiles on social-media sites.

The same phenomenon will happen with premium content, which is just starting to flow from destination sites like and to popular MySpace and Facebook pages. Particularly in an economic downturn, it's important for advertisers to leverage their existing assets such as short-form video ads and maximize consumer attention with ads that viewers will actually watch as they wait for their content to start.

Similarly, online video advertisers will get the most benefit from their investments when their content goes viral by tapping into the social graph on social-media sites. Ads will no longer have to wait for viewers to come to them; they'll find them where they live online.

As Charlene Li, author of Groundswell and founder of Altimeter Group in San Francisco, has said: "Social networks will become like air. They will be anywhere and everywhere we need and want them to be." If you're a content company, that's the distribution platform you want. You can go where the audience is. That's something that a banner ad can't do.

Video killed the radio star in the 1980s and, 20 years later, it's killing the banner-ad star.

Peter Yared is founder and CEO of iWidgets. He can be reached at

Facebook Getting Better, Slightly Worse, Better, Better

Robert Scoble has summarized the Facebook redesign very well:

My former boss, Jim Fawcette, used to say that if you asked a group of Porsche owners what they wanted they’d tell you things like “smoother ride, more trunk space, more leg room, etc.” He’d then say “well, they just designed a Volvo.”

People fundamentally do not like change. People also do not recognize that a product always has compromises. A product that never changes is going to die, and a product without compromises will never ship. I admire that Facebook is willing to experiment and push its technology forward even if it has drag its customers along. There was a similar outcry about the newsfeed to begin with, which is what everyone now is wishing worked the way it did when they originally complained about it!

The new realtime newsfeed is far from perfect. It really needs a thumbs down/thumbs up feature, for example. Just because I don't want to see a friend's constant quiz results does not mean I don't want to see photos they are tagged in or what videos they have rated. It is definitely prompting people to organize their friends into lists so that they can filter the newsfeed down to different circles of friends depending on their mood.

All in all, congrats to Facebook for experimenting and pushing social computing forwards. I am certain they will continue to tweak the experience and also add new features. Pushing the envelope like this is what a startup is all about and fortunately for everyone who uses it (all 180 million of us), Facebook still thinks like a startup.

Wednesday, March 18, 2009

Sun & IBM - A Good Move for All

Well I blogged back in 2004 about what a great acquisition Sun would be for IBM and then blogged recently about how well Sun's open source is doing, making Sun an even better acquisition for Sun.

Sun's market capitalization has been roughly equal to its cash position for years, so it is effectively worth $0. IBM's market capitalization less its cash position is 1x its revenue, so any revenue it gets increases its market cap by $1. So buying Sun at 2x its market cap is an ACCRETIVE acquisition for IBM!

What IBM gets:

- Sun's $600M in open source revenue from Java and MySQL are the crown jewels of this deal. IBM now gets to control the infrastructure language that most of its customers are running. They get to upsell MySQL customers to DB2. This is a dream come true for IBM.

- Sun's $6B in server revenue. IBM can make a SPARC to PowerPC microcode translator and kill Sun's entire server group. The customers can get upgraded machines that run faster than anything on Sun's roadmap. The small x86 server group can be immediately killed and customers transitioned to IBM's x series.

- Sun's $5B in service and professional services revenue and people get migrated directly into IBM Global Services.

- Solaris and StorageTek storage can be integrated into IBM's existing Unix and Storage groups and put into life support as customers get migrated to Linux and modern storage solutions. This is the same thing IBM is doing with AIX and its own legacy storage products.

- Everything else can be killed, including management, grid, etc. These are all nascent markets for Sun, and already strong markets for IBM.

All in all I think this is a good move for Sun and IBM, and Sun's customers. Sun has been slowly and successfully transitioning into an open source software company, and IBM is a great home for both Java and MySQL. IBM is an expert and maintaining and migrating customers from obsolete technologies, which it will be able to do for SPARC, Solaris, and StorageTek.

Goodbye, Sun, you will be missed.

Wednesday, March 11, 2009

Internet-Enabled TV for $10

There is no question that the under-25 demographic does not watch TV. Go through any college dorm and you will see people watching shows on their computers, and very few TVs. For the rest of us, there are a quite a few solutions for watching Internet content on your TV, including Vudu, Roku, and AppleTV. But they all require spending $200+ for your own Internet set top box or adding software like Boxee or Playon to an existing device. These boxes have interfaces ranging from good to bad, but, with the exception of Boxee, none are have really nailed a great user experience.

People paid $200 for DVD players, TiVo's, and DirecTV receivers, but for whatever reason, a lot of people intuitively don't want to pay for an Internet set top box. So what can we do about this? It is really holding up the Internet TV market.

As discussed at itvt's TV of Tomorrow Show this week, one solution is for cable operators like Comcast and Time Warner to throw in an Internet set top box as part of a broadband + cable TV package. This would enable cable operators to continue to offer cable service, but supplement their on demand offerings with broadband on demand and also integrate premium subscriptions such as HBO.

Until then, a lot of households only need to spend $10 to get Internet content on their TV. If you have a flat panel TV and a laptop computer, all you need is a $10 VGA cable and a standard audio cable (make that $30 if you have a Mac since you will need to get a VGA adapter). Connect your laptop's VGA out to your TV's VGA in port and the headphone jack to the audio in on your TV and you are all set. You can then pull up whatever content you want (including Hulu) on your web browser, and then hit full screen.

To jump on the Internet TV bandwagon, you don't have to buy anything other than a cable, learn anything new, or have your content kicked off your device since it is connect to a TV like the recent Hulu/Boxee scuff. There is a ton more content available for free on demand on the web than there is on TV, including on-demand cable. Soon enough even subscription content like HBO will be available.

An exciting aspect of viewing shows on computers is that there is much more potential for interactivity, particularly social interactivity when people can easily rate, share, and opine on shows and scenes.

Thursday, March 05, 2009

Share Beats Search: More Hits from Facebook than Google

A monumental shift in web traffic happened over the holiday season. Sites ranging from gossip such as to live streaming such as suddenly began getting more of their traffic coming from Facebook than from Google.

As reported in HitWise, last week had 8.7% of its visits from Facebook vs. 7.62% from Google. is a very popular site that recently scored its best traffic day ever in late February with 13.9 million page views.

Content sites have spent billions of dollars — $12.2 billion in 2008 according to a recent eMarkerter report — on search engine optimization and search engine marketing in order to get traffic from Google and the other search sites. Yet organic traffic from Facebook is beating hits from Google.

Why? The answer is pretty simple. It’s the same driving force that’s behind so much of social media. You are much more likely to click on a link that your friend recommends than you are to trust the arbitrary data that Google churns out in response to your search.

I have written often that social networks like Facebook and MySpace are the best route for content publishers to monetize their content. But even I am surprised at how rapidly this transition has occurred.

With only nascent investment in social media, publishers are already seeing better traffic from Facebook than from Google. Soon the SEM/SEO spend will start to follow the eyeballs and transition from Google to social media.

Smart social media campaigns are a much more efficient use of marketing dollars since it takes the potential that obviously exists in a friend’s recommendation and turns it into something lively and unique — breaking the mold of boring search ads and easy-to-ignore banners and entering the world of social syndication.

--- This morning NewTeeVee wrote about the same thing, I brought the HitWise report up at their Beet.TV roundtable earlier this week.

Saturday, February 28, 2009

Newsfeed User Engagement - It Really Works!

I have been talking and blogging about the power of newsfeed engagement for a long time, but even I was impressed last week during a demo to a large publisher. Usually I use a fake Facebook profile for demos, but this time I used my own since I was already logged in as myself.

The first demo was our American Idol Social Poll, followed by our Inauguration Video Poll. Both of these polls posted my actions to my newsfeed.

Within 15 minutes, friends had engaged with both newsfeed items! One friend commented on who she thought would win American Idol, and another friend reminisced about hearing JFK's inaugural speech as a child. I have to admit, even I was blown away by the immediate friend engagement.

It used to be that user engagement features such as polls would engage a user for a few minutes and they would move on to the next thing. Now, brands like CNN and have the opportunity to help a user express themselves, and also engage a user's friends which leads to higher engagement-based CPMs and additional traffic. Self-expression for the user + more revenue for the publisher = win-win for both the user and the publisher!

Tuesday, February 24, 2009

The Rising Sun

I have been ruminating on Sun for a while, ever since Savio Rodriques' blog on Sun's software earnings last month. Last quarter, Sun's Java licensing generated $67M and MySQL generated $81M. Not even taking into account the 50% growth rates that these products are seeing, that is a $600M annual run rate.

As Matt Asay pointed out, Red Hat and Sun had virtually the same market cap for a bit, although Red Hat has since dropped. If you account for the cash that Sun has, Red Hat is actually worth a couple of billion more. Red Hat had $500M in revenue last year.

Contrary to Matt, I don't think these numbers are as exciting for Red Hat as they are for Sun. What is interesting is that if you look at the numbers only, Sun is now a software company. If Sun consisted only of its Java and MySQL groups, it would be a $600M/year commercial open source company that is worth, at the same multiple as Red Hat, $2.5B. When you take into account Sun's cash position, that is more than Sun is worth right now!

So yes, I was quoted in Forbes saying that Sun should jettison its Java group, and I have to take a mea culpa on that one now that I see that Sun is making over $300M/yr off of Java.

Sun is now a modern, open source software business, and a layer up above Red Hat at that. The hardware at this point is a channel for the software and will soon be rentable as a cloud offering. I imagine that Sun will continue to climb up the software stack and roll up SugarCRM, Alfresco, Jasper, and others.

Kudos to Jonathan Schwartz. It took a long time to redirect the lumbering ship, but he has done it. I think that soon enough the market will recognize this new, rising Sun as a counterbalance to Oracle's proprietary solutions. Congratulations!

Sunday, February 15, 2009

"Social Polling" for American Idol on

We just launched an American Idol "social poll" on, an extremely high volume site. The poll is the first largescale deployment of iWidgets' new polling feature and shows off:

  • A sophisticated poll with 36 American Idol finalists to choose from

  • Realtime vote polling using AJAX

  • Badges that voters can add to their Facebook and MySpace profiles to show off their choice and get their friends to vote.

  • The poll is an iWidgets widget that is embedded into as an embedded iFrame and is virtually indistinguishable from The ability to fit into existing sites is an excellent benefit of HTML/JavaScript widgets. The poll lets users vote for their favorite Idol from the final 36 idols:

    Once the user votes, the poll shows how many votes each American Idol finalist has received as well as a banner promoting that the user show off their favorite Idol and get their friends to vote:

    Clicking on the Facebook / MySpace / embed icons on the banner takes you to your social network to install a profile badge. The first step here is a new "targeted invite" feature that finds your friends that are already interested in American Idol, which people are much more likely to invoke:

    A native social network widget is then added to your profile. Such profile badges can also include fresh video clips that can be monetized with pre-roll ads:

    There is of course to obligatory "narrow" view so that Facebook users can add the widget to their highly visible wall/profile tab:

    And a newsfeed update so that the users' friends take the poll and drive additional traffic to

    iWidgets' social polling is a great way to add a fun, seamlessly integrated widget to your website that engages your users to add your content to their social profiles and gets their friends to visit your website.

    Tuesday, January 20, 2009

    As iGoogle and Yahoo! Get Socialized, Will Facebook Get Portalized?

    Techcrunch is writing about how iGoogle gadgets are getting social features and VentureBeat is writing about how Google's profile feature and Yahoo!'s FriendFeed feature. Clearly the two largest portals are adding social features: they already have plenty of users, tons of widgets, and are the default browser home page for millions of users — why not add these new features to keep them there even longer?

    Portals built their business on the idea that there was money to be made making the web easier to manage for the average user by aggregating data and providing a defined starting point. With the growth of RSS feeds and widget technology, portals have become even more sophisticated, but the addition of social features — the ability to create a profile and start building a “social graph” — gives users one more reason to keep coming back.

    But why isn’t Facebook adding portal features to its homepage? Sure, the big long friend activity feed on Facebook is interesting, but as you scroll down there is a bunch of white space to the right. It seems like and obvious competitive move for Facebook to allow people to add weather forecasts, stock quotes, movie listings, their RSS feeds, etc. Facebook already has the widget framework, the API, and the developer community that can create these widgets. This way Facebook has the opportunity to become the default home page for people, and can fight Google and Yahoo! on their turf - while it defends its own.

    Friday, January 16, 2009

    SpringWidgets, Flektor, Sprout Dead - The End of the Flash Widget?

    I have long maintained that Flash widgets are a nonstarter on all the growth areas of the net - social networks like Facebook and MySpace which use FBML and OpenSocial, portals like iGoogle which uses Google Gadgets, and mobile platforms like the iPhone and G1 that use native APIs. In the past few days, Fox Interactive announced that it is shutting down Flektor and SpringWidgets, which is an obvious move since these properties are not compatible with MySpace's OpenSocial API, and SproutBuilder is shutting down its free service in order to target ad agencies, which focus mainly on banner ads rather than full-fledged widgets. The only area where Flash widgets are competitive is embedded in websites and blogs, and Clearspring, Widgetbox and Gigya all have excellent solutions for distributing and tracking such widgets "in the wild".

    At iWidgets we are strongly committed to delivering the next generation of deeply integrated widgets for the growth destinations where content sites want to be like Facebook, MySpace, iGoogle, and iPhone. Each of these platforms provides great APIs for destination-specific features like viral channels, caching, and native components. We will definitely maintain our free widget solution, as it is very cheap for us to host via Amazon's EC2 and provides us with invaluable advertising and cross-promotion space, and our paid offering is pay-for-performance rather than hosting charges. Our customers have validated that native widgets, advertising-based free widgets, and pay-for-performance premium widgets is the way to go, and it is our job to continue to add features and refine the user experience so that more and more content sites can grow through off-domain impressions.

    Tuesday, January 13, 2009

    The Widget Stampede at CES - Yahoo! Connected TV + Palm Pre

    The Widget Stampede Continues at CES: Yahoo! Connected TV + Palm Pre

    The two top stories at CES were both about new widget platforms: Yahoo! Connected TV and the Palm Pre. The capability to distribute your content as widgets onto more and more devices is growing and as more of these devices become available, it’s only going to get more complicated.

    The Palm Pre is a gorgeous new phone with multitasking applications and an actual keyboard. It looks like it is a generation ahead of the iPhone and the Android G1. What is really exciting about the Palm Pre is that its widgets — Apple calls them “apps” — are built using derivatives of standard web API's like HTML and CSS, so you don't have to learn weird API's like the iPhone or Android in order to create a widget for it. It is a much lower bar, though it is still a bit of a pain since it is not a standard web page. I predict the Pre will give the iPhone a run for its money not only in terms of numbers of people who prefer keyboards, but also the number of widgets available.

    Yahoo! Connected TV is a really cool technology where integrated into the actual TV are the services people today are using set-top boxes like the AppleTV for, such as streaming video, photos, and music. With the Connected TV widget feature, a user can add widgets from a content producer right on their television’s screen, just like they add a widget to their iGoogle home page or Facebook profile. Such TV widgets include real-time stats on your fantasy league as you are watching a football game and recipe lookups as you watch Food TV. Yahoo! Connected TVs is also based on a customized HTML derivative, not an exclusive API.

    I think we can make out two trends here:

    (1) Widgets platforms are multiplying at a fast pace, offering huge opportunities for content producers to engage their audiences wherever they are — their iGoogle homepage, on their favorite social network, on the go with their phone, and now in front of their TV.

    (2) Most of these widget platforms use a derivative of HTML with hooks into native features like newsfeed updates and multiple form factors, Facebook's FBML and Google's OpenSocial being the most pre-eminent examples. It is a huge effort to create a version for each destination, a problem iWidgets has solved with our WidgetWORA technology.

    Adding more handsets and now TV's to the widget mix will make 2009 a very interesting year indeed.

    Saturday, January 10, 2009

    Management Page 2.0

    When I check out a startups' website, one of my first stops is the management page so I can see who is involved with the company. Even the most innovative startups offer up staid, out-of-date photos, and there is no way to quickly figure out if you have mutual friends, since you have to search for the name at LinkedIn and Facebook and there are often multiple people with the same name. For our management page at iWidgets, we took the easy way out (why pay for photographers?) and show Facebook badges with our current pictures and status updates as well as a link to the LinkedIn profiles. The pictures are always fresh and also include live status updates, and people can clickthrough and see if they have common friends.

    It would be great if LinkedIn had a little "bio" widget you could add to web pages that would show your picture, employment history with detail hovers, and even a list of common connections for the person viewing the bio if they are logged into LinkedIn.