Wednesday, May 25, 2011

How Twitter 2.0 Will Make Money

This post was also published in VentureBeat.

This is the second of two articles on likely changes to Twitter. The first article focused on consumer-facing changes to Twitter and this one focuses on monetization.

Twitter has experienced tremendous traffic growth, and more importantly has permeated into the collective consciousness with constant Twitter quotes by news organizations and entertainers. Twitter’s valuation has gone up as people expect it to monetize its traction with both consumers and brands. Here are a few of the features that would seriously inflect the Twitter revenue curve.

Bring on the banners

Twitter has struggled to scale advertising revenue, because like most social sites it tries to inflict strange ad units on both its users and advertisers. From the users’ perspective, things like promoted tweets or Facebook’s social ads are intrusive to their content experience. From the advertisers’ perspective, they can’t use their normal ad units and track results the normal way.

This is easily addressed by supporting normal ad units such as right rail skyscraper banners. As Myspace proved early in the social years, banner advertising works on social networks, especially social networks where a lot of people are following celebrities and media. In addition, to be frank, a lot of the former Myspace userbase is using Twitter to follow celebrities and sports figures, and these are the types of users that respond to banners. It will be a long time before Facebook chief operating officer Sheryl Sandberg can convince Mark Zuckerberg to go after banner ad buys less than a year-and-a-half after he pulled Microsoft’s static units from Facebook, so Twitter will have an advantage going after large-scale media spend.

Special care should be taken to ensure that when banners are introduced, they are placed similarly to other top sites. For example, the infamous “dickbar” mobile banner that caused user uproar was placed at the top of the Twitter app. Most mobile banners such as AdMob units are placed at the bottom of an app. Adding banners will always bring on complaints, but adding banners that look like everyone else’s banners will achieve begrudging acceptance and will make the complainers look like trolls.

To add a social flare, the units can include tweets relevant to the brand or whatever is being promoted. Put the social into the ad unit, not the ad unit into the social. Acquiring the self-service dynamic ad creator Flite, which lets advertisers create and deliver ad units similar to Aol’s Project Devil units, would make dynamic ad units that integrate tweets available to the as Flash units and mobile as HTML5 units.

Twitter should let Facebook experiment with social ad units, and focus on soaking up a bunch of brand banner ad spend.

Monetize (rather than cannibalize) third-party apps

Over the past few months Twitter has been going after third party Twitter apps by blocking some apps’ access to its API and arbitrarily changing authentication systems. Attempting to regain control of the on-ramp with which power users use Twitter is admirable, and the TweetDeck acquisition was a very smart move in this direction.

However, a rich developer ecosystem is a rare treasure and should be nurtured. Rather than go after rollups like UberMedia and smaller Twitter clients, it would be better to exchange volume API access in exchange for placement of the banner ads (you know, the ones I just described) with a revenue share. Pissing off developers is never smart, consider how quickly iOS developers jumped onto Android after all of Apple’s various shenanigans. This is especially important considering that Facebook is nurturing a rich ecosystem of third-party apps, ad, and analytics vendors. That said, it should be expected that Twitter will add obvious features like email notifications, and Twitter should notify developers of changes like this ahead of time to circumvent upset feelings.

Developers love Twitter, and Twitter should be helping them do fun stuff and make money, too. There are a bunch of small, self-serve ad networks that could be acquired to accomplish this, as well as open source implementations such as OpenX.

Check-ins are tiresome, but location deals are useful

Twitter is one of the most used mobile apps, and it could open a great revenue stream by enabling merchants to deliver real-time deals. Check-ins are increasingly lame, but people do want to know if there are deals around. A lot of mobile units have low conversion rates, but a built-in interstitial with a map view would lead to a lot of conversions. Options for this include doing a deal with Groupon Now like Loopt’s recent implementation and Foursquare’s rumored integration, or acquiring a small mobile deals company like BeThere Deals. Replace the #dickbar with the #dealbar!

Sentiment analytics and service tools for brands

I work at Webtrends, which sells social, mobile and web analytics solutions to enterprises, and am continually amazed at the voracious appetite marketers have for analytics, and for the new breed of analytics around social interactions. No CMO on the planet is going to question a marketing budget item labeled Twitter Analytics. A lot of sentiment analysis tools such as CoTweet, ScoutLabs and Radian 6 have already been acquired. However a new breed of more powerful tools such as Crimson Hexagon are ripe for Twitter to acquire to enable social media teams to see, route and respond to Twitter conversations about their brand. Much like Facebook Pages include support for broadcasting to Twitter, it is critical to also include support for Facebook and blog conversations so that brands can have a single, Twitter driven view of what is going on with their brand.

Twitter should continue to sell access to its firehose of data, but delay access for a few hours to external entities other than a few premium partners in order to increase the value of its own offerings.

Sell lucrative sharing, trending, and tipping data

Companies like ClearSpring and RadiumOne are mining what content is being shared by consumers and selling the data at a huge premium to advertisers who want to know what topics are trending and how to better target users. Twitter has one of the best views of this type of data and should sign lucrative deals with publishers and ad networks to help them better optimize their businesses. More importantly, as I wrote back in March, Twitter’s data provides a very unique viewpoint as to how content goes viral and what types of posts, influencer retweets, and incremental advertising could tip a campaign towards a viral spread. Social scoring companies like Klout have very limited algorithms, but this type of infrastructure could be an offshoot of acquisitions such as Topsy and BackType in addition to the more immediate benefits described above. Mining the Twitter data will only become increasingly valuable as marketing campaigns become increasingly automated.

Adding enterprise features to Twitter will be much easier than consumer facing features. Enterprise features have far fewer users, can have structured release programs such as limited early access releases, and generally have a pretty forgiving audience in terms of time to add features and the occasional downtime. There is huge opportunity in for Twitter to sell to businesses; Twitter’s voluminous consumer traction, brand mindshare, and treasure trove of data offer significant monetization opportunities in the advertising and analytics segments. Acquiring companies like those described above would accelerate Twitter into every brand’s media spend.

Tuesday, May 24, 2011

Here Comes Twitter 2.0

This post was also published in VentureBeat.

This is the first of two articles on likely changes to Twitter. This article focuses on changes to Twitter’s consumer-facing side and the second article focuses on Twitter monetization.

Twitter reportedly acquired TweetDeck today, and that’s likely to be the first of many changes. There is a broad consensus that Twitter had stalled out in terms of product innovation, , which even creator Jack Dorsey noted upon his return to Twitter as head of product. With Dorsey’s return, we should expect more changes, and very quickly.

BusinessInsider recently suggested that Twitter only has 21 million active users, and a lot more people who read of those users’ tweets. While at first blush this number may look really bad, it is actually an indicator that Twitter has matured. Twitter is a microblogging platform, and has followed the same trajectory as blogs. At first, everyone blogged. Now only a few blogs really matter.

Consumers want Twitter to be their source of information. Lady Gaga on Facebook has 35 million fans, and on Twitter she has 10 million followers. The numbers don’t tell the whole story: The 10 million Twitter followers actually care what Lady Gaga has to say every day, while a large portion of the Lady Gaga Facebook fans simply added Lady Gaga to their profile as one of their musical preferences, which Facebook later automagically turned into becoming a fan of the Lady Gaga Facebook Page.

As Twitter becomes the go-to source of information for news, celebrities and more, it’s differentiated from Facebook, which is very focused on what your friends are posting. With this shift in mind, below are some features Twitter could quickly add via targeted acquisitions that could bring back the company’s mojo, scale users, and add multiple revenue streams.

Add a news view

Twitter’s home page should become a news style view of what is going on. As LinkedIn’s recent news site showed, there is ravenous appetite for socially filtered news presented in a news format that includes pictures and descriptions. I myself have experimented with PostPost, the real-time Facebook news site, to rapid uptake.

Twitter has become the central avenue for news organizations and celebrities to distribute news. In addition, it is the prime hub for curators and users to retweet the news, essentially adding a vote for the interest level in the news story. Facebook is still primarily a site to interconnect with friends rather than disseminate news. However, Twitter needs to act fast since Facebook is currently driving more traffic to news organizations that Twitter. It’s time for Twitter to double down and dominate its core category.

With a Twitter news view, newbies and infrequent users would start with a broad view of what is currently interesting in the news, and as they follow sources and curators they would be able to personalize their experience and drastically increase Twitter usage. Adding a news feature could be accelerated by acquiring Twitter newspaper makers such as or TweetedTimes, with a blend of TweetMeme to rank the articles.

Offer news streams for events and locations

Last month I was at the Coachella music festival. On the first day, there was one person on Color, 66 Foursquare check-ins, and thousands of #Coachella tweets that included linked photos and videos. Twitter has a fantastic opportunity to offer a dynamic, visual view of what is going on around events and places, ranging from music festivals to political upheavals. Facebook is the go-to site to find out what’s going on with your friends, and Twitter should be the go-to site to find out what’s going on around you and in the world. Since news organizations are always quoting tweets, people are now predisposed to Twitter as a leading source of information. Acquiring companies like Memolane or Storify that have spent time on how to present posts in a structured manner could be helpful. In addition, acquiring What the Trend would add the ability to show trends of tightly focused geographic and demographic slices.

Show photos, videos and other media inline

The Twitter user interface has been stalled almost since its inception. Although users can now click to fetch linked media, it is not an ideal experience, especially since users are now accustomed to seeing photos and videos inline with the Facebook newsfeed. Photos, videos, documents and such should be presented inline directly below a tweet. Twitter should announce support for Facebook’s OpenGraph protocol and support open graph tags for web pages that return metadata such as titles, descriptions, and thumbnails. This feature could be accelerated by acquiring, a site that returns metadata such as thumbnails for arbitrary URLs.

Inline comment threads

Following a conversation stream in Twitter is maddening. Although there is a new feature to see a tweet in context, it only shows a few surrounding tweets that are not necessarily relevant. This is a critical feature that needs to be fixed, and can’t be fixed via an acquisition. Every tweet that is a response to another tweet needs to be linked to its parent tweet, and every tweet needs to be able to find its child tweets.

Facebook addresses this issue by simply grouping all comments under the original post. Twitter should also focus on solving this simple problem first, and then going after the more complicated nested commenting problem. So if I reply to a tweet, and then you reply to my tweet, both tweets are filed under the original source tweet as one long stream. A huge step forward from how it works right now, and the nesting of comments can be addressed later.

Improve search

Although Twitter acquired Summize in 2008 to power its search engine, the search results could use a lot of improvement. Twitter search doesn’t dereference URLs, so you can’t search for references to a particular article. In addition, it doesn’t offer expected features for people accustomed to Google search such as spelling corrections (“did you mean…”). Building fast and effective search is extremely challenging. Options should include partnering with Google like Facebook partners with Bing. Google would be overjoyed to get some social juice, and Twitter would benefit by being able to offload search. Another option is to acquire Topsy, the leading Twitter search engine which is fast, realtime, and effective.

More Twitter widgets for websites

Facebook is blanketing the web with its social plugins ranging from Likes to Comments, and Twitter needs to stake its claim on websites, especially news sites. An equivalent of Facebook’s recommendations plugin that shows what content on a site is actively being tweeted would be relatively easy to create. In addition, context about what people are saying about an article could be added by acquiring BackType, which offers a widget that shows all of the tweets and comments about a particular URL.

With the service’s notorious downtime, the Twitter engineering team has struggled just to keep the lights on and it will be awhile until they can transition to innovating rapidly. Although acquisition integrations are hard, they will be the best way for Twitter to innovate quickly. If the acquired companies are not compatible with existing infrastructure it is critical to incent them to do a quick rewrite as part of the acquisition. In addition, Twitter should be comfortable shipping stuff that people will complain about – Facebook does it all the time. One idea is to start throwing features over the transom under a “TwitterLabs” moniker.

Good thing that Twitter recently hired Mike Brown, Facebook’s former acquisitions guy, and didn’t buy into TechCrunch’s accusations that he committed insider trading. Brown is super tied into the startup community and is well-respected. Looking at the laundry list above, he is going to be a pretty busy guy. (Editor’s note: Brown is also an investor in VentureBeat.)

Tomorrow, I’ll dive into Twitter’s monetization opportunities.

Thursday, May 19, 2011

LinkedIn IPO: It’s Only a Bubble when Grandma can Buy In

This post was also published in VentureBeat.

LinkedIn’s massive 2x pop this morning when it went public reflects that everyone, including the underwriters, seriously underestimated the demand for the shares. There was a time, especially during the .com, when underwriters such as Morgan Stanley would attempt to artificially go out the door with a low price in order to create a pop, but companies going public nowadays are much more sophisticated and discerning about leaving so much money on the table.

So why is there so much pent up demand? It is not just, as Jim Kramer claims, the limited number of shares. Hot, second wave Internet companies are taking a long time to go public and the stock is only available to a select few. Only accredited investors could invest in hot Silicon Valley companies through secondary exchanges like SecondMarket. On top of that, early stage investors and management are much more patient due to the liquidity gained through secondary markets and the willingness of late stage investors like DST to buy out existing stockholders.

I have lived through two bubbles, the .com bubble and the housing bubble, and have concluded that bubbles are essentially giant pyramid schemes. And Grandma is at the bottom of the pyramid. It is a bad sign when unsophisticated Grandma starts investing her money in 1999 .com startups that she doesn’t understand and in 2008 investment property in Phoenix although she knows nothing about investment real estate. Yes, everyone else has made so much money and it is going up and up, and her financial advisors are telling her to take advantage of the opportunity. But after the unsophisticated investors like Grandma jump in, there is no one left for her to sell her stock to.

Grandma is the last stop of the gravy train.

We have felt the early signs of the bubble here in Silicon Valley, but the price of companies has been self-contained to the Bay Area. My company Webtrends decided to move to a different area of San Francisco from uberhip South Park because rents got too high. Engineers are demanding skyhigh salaries.

The key metric to watch is multiple on revenue. LinkedIn is valued at 36 times its 2010 revenues, which is a very high multiple. Contrast that to San Francisco-based RPX Corporation, which acquires patents on behalf of large companies to fend off patent trolls. RPX went public ten days ago and had a slight pop, and is now valued at $1.3 billion, ten times its 2010 revenues. Ten times revenue is generally considered a fantastic multiple, and well deserved for a company that is profitable and growing fast in a novel space such as patent protection.

Grandma is never going to hear about RPX. When the multiple on revenue spread between the LinkedIns of the world and the RPX’s of the world gets too crazy, you know the grandmas are all getting in and it is time to get out. In the mean time, numerous attractive young women are floating around South Park, a sight unseen since 1999, so enjoy it will it lasts!

Monday, May 16, 2011

Can Social Commerce Catch Up with Email?

This post was also published in VentureBeat.

Although many are heralding the advent of “social commerce”, the biggest e-commerce wave of the last couple of years has in fact used the most basic Internet technology. Enormously successful deals companies such as Groupon, LivingSocial, and Gilt Groupe send simple e-mails. Users click through those emails to a single landing page whose only goal is to get users to hit “Buy”.

There is no innovative technology at play here, rather a simple email op-in that asks consumers what regions or products they are interested in when they join the mailing list. Although Groupon likes to tout its tipping point algorithm where a certain number of users have to purchase in order for a deal to happen, there is such volume that all deals go through, every time. Net net, nothing social going on here.

The social revolution has made brands think they can invert how people think. Sure brand ambassadors are great, but they are few and few between, and lots of people on Facebook “like” Lady Gaga. But people do not want to broadcast what they bought, they want deals brought to them.

Last summer, Facebook chief operating officer Sheryl Sandberg claimed that email was dead, but she must have seen the light when it came to the recently launched Facebook Deals service. Consumers are now so programmed to receive deals via email that Facebook had to also send its deals via email, not through Facebook’s countless viral channels such as the newsfeed, messaging inbox, and notifications.

There is now such a heavy volume of activity on Facebook newsfeeds that no one ever goes back in time to see what happened, so branded posts in your newsfeed quickly gets buried. Conversely, most people still scan through all of their email and can see that Groupon sent them a deal yesterday.

Facebook’s other commerce initiatives have all faltered. The MarketPlace classified ad feature is outsourced to Oodle and has not had much traction. Adding Facebook storefronts to brand pages is currently the rage, but in my experience people do not like deep engagement on a Facebook page, and prefer a quick in-and-out, much like they interact with friends. So Facebook storefronts are more useful for promoting a few items rather than incorporating an entire store. The one shining star is the fact that Facebook got its gaming masses to adopt Facebook Credits as a mechanism for paying for virtual goods in games, mainly by strong arming game developers to support Credits.

Traditional commerce success is very important to Facebook. Most of its ads are sold to small local businesses, game companies like Zynga, and deals companies like Groupon and LivingSocial looking to acquire new users. In order to fuel its growth, Facebook needs to be able to close large large ad buys with national brands and retailers.

However, Facebook keeps insisting on atypical “social” ad units, such as sponsored stories. Things like retargeting already creep people out, when they see a shoe or a car following them around the Internet. And people decidedly don’t like when they are the object of retargeting like with Facebook’s sponsored posts feature, for example “Your friend Joe likes Home Depot! You should to.” Most Joes really don’t like that. Let’s face it, most people don’t like ads, and making ads social by including people in them is just creepy.

Even the omnipresent Facebook Like button is not helping retailers very much. Most products have too short a shelf life to accumulate many likes. In addition, very few people are clicking Like on a product since it is broadcast to all of their friends, and instead are are using the “e-mail to a friend” feature so they can ask a select set of people if they like an item. Email is really making Facebook miserable when it comes to e-commerce. No wonder Forrester and Goldman Sachs are questioning whether Facebook can make it in retail.

So what is working? Justin Kistner, a young product manager at my company Webtrends, came up with the idea of using Facebook Ads as the equivalent of an email campaign. We all initially thought he was insane, but in fact we found that it works enormously well.

Here’s how it works: Get a consumer to Like a retailer on Facebook through a Facebook ad campaign where the consumer gets a coupon in exchange for Liking the retailer, which is the equivalent of a traditional banner-ad-to-email opt-in campaign. The retailer can then buy low-cost ad units that target people who have Liked their page, and also those peoples’ friends who are likely to have an affinity for the retailer as well. In the campaigns we have run, these units get a 7x higher conversion rate than other Facebook ad units. Those ads then click through to targeted, simple landing pages on the retailer’s Facebook Page or website that convert into a sale.

Replicating the proven email opt-in, email send, and simple landing page experience to Facebook as a Facebook Page Like, cheap ad buy targeted to the page’s fans, and simple Facebook landing page works remarkably well. The Facebook ad is essentially an additional newsfeed element that is placed off to the right. Adding a one-click Facebook Credits “Pay” option, support for normal ad units such as skyscraper units on the right rail (like Facebook used to display from Microsoft), and deep analytics that track what happens after users click on ads would increase conversions even further.

Is it social? No. Does it poise Facebook as a great place for retail ads and as a payment platform? For sure.