Facebook has created a new way for businesses and developers to target ads to users who have taken an action on Facebook or in an app within a more specific time range.

Using the “action spec” targeting capability, which allows advertisers to reach users by the actions they take in Open Graph apps and on Facebook directly, an advertiser could designate a time range shorter than 14 days. The default is still 14 days, but previously there was no option to set a shorter window.

This change gives advertisers more accuracy in reaching the most relevant users for a particular message. For instance, a local business might want to reach users who checked into their store within the past week. A developer might want to target users who installed their app in the past three days. A page owner might want to retarget users who viewed one of their page tabs the day before.

Action spec targeting is still a beta feature limited to advertisers working with Preferred Marketing Developers with the Ads API, so it isn’t widely known or understood. However, the feature offers developers unique opportunities to reach users who have taken specific in-app actions, including in their competitor’s apps. It also gives brands ways to segment and target their fans by the actions they take on their page or a competitor’s page.

Action spec targeting is also interesting in that advertisers can define a “negative action spec,” meaning users who have not taken a particular action. For example, a developer could reach users who have played a game, but not made an in-app purchase in the last 10 days. Advertisers can also reach friends of users who have taken a particular action.

More technical information about action spec targeting, including targeting by app activity in a time range is available here. A list of Facebook actions that advertisers can target against using action specs is available here. Businesses that are interested in this type of targeting should seek out an ads-qualified PMD.

Article courtesy of Inside Facebook

Lambda Labs, an early stage startup out of San Francisco, is preparing to release a facial recognition API for developers working on Google Glass apps. The API will be available to interested developers within a week, company co-founder Stephen Balaban says. The move comes on the heels of a Congressional inquiry into Google’s new wearable technology, still very much in the prototype phase.

The startup’s facial recognition API, launched into beta last year, is already used by 1,000 developers, including several major international firms. It now sees over 5 million API calls per month, and is growing at 15 percent month-over-month. Balaban also says that the company has been cash flow positive since November.

Now that same API has been tailored specifically for Google Glass Apps to enable both facial and object recognition.

Applied to Glass, the technology will enable apps such as “remember this face,” “find your friends in a crowd,”  ”networking event interest matching,” “intelligent contact books,” and more, Balaban explains. (You can see what apps developers are tweeting about here.)

As potentially amazing / horrifying as that technology sounds, any apps using the technology couldn’t do so in real-time – that is, you couldn’t just walk around automatically recognizing people you see through Glass. The way Google’s Mirror API works right now is that you first have to snap a photo, send it to the developer’s servers, then get the notification back. The lag time on that would be several seconds at least, and would depend on how fast you could take a photo and share it. A forthcoming Glass software development kit (SDK), though, may change that.

“There is nothing in the Glass Terms of Service that explicitly prevents us from doing this. However, there is a risk that Google may change the ToS in an attempt to stop us from providing this functionality,” Balaban says. ”This is the first face recognition toolkit for Glass, so we’re just not sure how Google, or the privacy caucus, will react.”

The privacy caucus he’s referring to has to do with the congressional inquiry from earlier this month where eight members of Congress reached out to Google CEO Larry Page with over half a dozen questions about Glass’ capabilities and the potential impacts to user privacy. The Bi-Partisan Privacy Caucus, a group led by Texas Republican Joe Barton, wanted to know if Glass would collect data from users without their consent, whether or not Google would consider privacy before approving third-party apps, and a host of other things.

One of those questions was whether or not Glass would have support for facial recognition. That’s something Steve Lee, Glass director of product management has already answered. In a statement offered to The New York Times, he replied, “We’ve consistently said that we won’t add new face recognition features to our services unless we have strong privacy protections in place.”

That’s not a solid “no,” of course. It’s more of a “no, for now.” Glass is simply too new of a technology to begin limiting what it will or will not do, at least in such definitive terms.

Facial recognition, however, doesn’t appear to be specifically prohibited in Google’s API policies, which inform Glass developers what they can and can’t do in their applications. That means, for now at least, Lambda’s facial recognition API for Glass developers would be permitted.

The only cause that would impact its use, according to Google’s policies, is one that says Glass is “not intended for use in connection with applications and services that might be subject to industry-specific privacy regulations.”

Obviously, lawmakers could still enact such a policy, if they chose to do so.

“Assuming Google and Joe Barton’s Privacy Caucus don’t attempt to stop us, [the API] will be available to everybody within the week,” Balaban says.

Google, it should be noted, has long since had the technology to build apps capable of facial recognition itself, but has always tread very carefully to not incite a privacy backlash.

In 2011, there were reports that Google was developing a mobile app that would allow users to snap pictures of people’s faces to access their personal information. That app never arrived, but facial recognition has since made an appearance within Google’s photo-sharing service Google+ Photos (previously Picasa), where users can now opt in to have their face recognized. This makes finding “pictures and videos of you easy,” explains the company’s documentation on the technology.

Perhaps one day, users will be able to “opt in” to having Glass apps identify their faces, too?

Time – and Congress’s reaction – will tell.

Image credit: Glass concept from Jack Morgan on Dribble; Additional reporting: Frederic Lardinois

Article courtesy of TechCrunch

Former Badoo COO and ex-Googler Ben Ling has joined Khosla Ventures, according to sources. Ling, who has held senior operations roles for a number of big companies in the mobile and Internet space, has been added to a growing team at Vinod Khosla’s venture firm. (Update: Khosla Ventures just confirmed Ling has joined, starting this week.)

Ling most recently served as COO of Badoo, where he was hired to oversee product, engineering, partnerships, and business operations at the company. Ling joined Badoo in May 2012, but he left after only about six months.

Prior to Badoo, Ling spent a number of years at Google, where his last role included overseeing images, videos, books, news, and finance as its senior director of search products and local business products. At YouTube, Ling was senior director of partnerships and platform, where he was responsible for music, movies, sports and news, as well as mobile, TV and API partnerships. In his first role at Google, he oversaw the company’s e-commerce products, including Google Checkout and Product Search.

At Facebook, he served as the director of Facebook Platform, working on developer relations at the fledgling social network. There he helped build Facebook Connect, which is the social network’s hook into a number of third-party websites.

In addition to his senior operations roles, Ling has also been an active angel investor and advisor to startups. Recent investments include Fab.com, Palantir, Square, PracticeFusion, and Quora. Ling has also held an advisory position at Pinterest and Pulse.

In a conversation with TechCrunch, Ling told us that most of his experience as an entrepreneur has been in operations roles, helping companies like Google and Facebook scale their businesses both in terms of users and bringing in revenues. Meanwhile, over the last four years, he’s been working as an angel investor, identifying great products and teams and helping them to scale.

“In terms of thinking about my next steps, and thinking about what I could do next, I wanted to maximize the impact I could have by helping entrepreneurs to grow their businesses,” Ling said. He said that the opportunity at Khosla Ventures enables him to do that, since there are a lot of “parallels and similarities” to what he was doing as an angel investor — evaluating teams and talent, helping companies scale and with their hiring strategies.

Ling said that we can probably expect to see him make investments in a lot of the same areas that he’s worked in previously — that includes companies focused on e-commerce or shopping, developer networks, search, finance, and local. He’s also had some experience with three-sided ecosystem businesses, such as the Facebook platform’s network of users, developers and advertisers.

Of course, Ling isn’t the only operations specialist to join Khosla Ventures. His hire comes just a few months after former Square COO Keith Rabois joined the firm.

Part of the reason Khosla Ventures has been interested in bringing on entrepreneurs with that type of experience is that they can provide guidance for startups who need operational experience. Rabois, who is now 10 weeks in at Khosla Ventures, said, “Our organizing philosophy is to provide assistance to entrepreneurs and help them to build the most interesting companies they can… It’s about understanding what entrepreneurs need is less capital and more formative advice. Every time we evaluate a company we ask whether we can help this entrepreneur build something special.”

Both Ling and Rabois say that they’ve been impressed with the quality of entrepreneurs and the quality of the Khosla Ventures team as it works with those startups. With that in mind, Rabois said he’s really excited about building several big businesses as part of the Khosla Ventures team.

Article courtesy of TechCrunch

Imonomy, an Israeli startup which makes software that analyses webpages and automatically inserts relevant, copyright-free images to accompany the content, has closed a $400,000 seed round from a group of angel investors. Investors include Inon Axel, former CEO of Kasamba (acquired by LivePerson for $40m), Liron Rose, cofounder of AfterDownload (acquired by ironSource for $28m), and Itai Levitan and Tal Shaked, partners at AfterDownload.

Imonomy said it will be using the new seed funding for product development and initial marketing and sales activities.

The startup was founded in the middle of last year by Oren Dror and Amit Halawa who previously held senior R&D and engineering positions at Yedda, an online Q&A service that was acquired by TechCrunch’s parent company AOL, back in 2007.

Imonomy targets its software at smaller-sized web publishers who have a pool of online content but don’t necessarily have the means to spice it up with illustrations — either lacking the production staff to spend the time hunting down royalty free images or the licensing money to pay to display copyrighted images. Imonomy says its semantic software is being used by more than 500 medium-sized websites (with up to 10 million monthly impressions) at present, including AOL Answers and Articles Base.

Imonomy’s software scans web content to figure out relevant images to serve up from its database of copyright-free images, and also determines the optimal place to position them on the page to improve user engagement. Inserted images support hover over links to other articles and also displaying ads, giving publishers (and Imonomy) a way to monetise the added eyecandy. It’s effectively a more aesthetic version of inline text link ads.

“The idea behind Imonomy is that publishers of content-heavy web sites need to tools to help their sites be visibly attractive,” the startup tells TechCrunch. “High quality copyright-free images are hard to come by and the time and effort required to locate such pictures is a hassle. Imonomy created its content enrichment and monetization system to automate this process in order to help publishers save time, improve user engagement and create monetization opportunities.”

Here’s how it describes its system on its website:

Our database contains millions of images that cover every possible topic. Our system scans your webpage, finds the best fitting image and automatically insert them into the published page. Thus making content more interesting and informative. Our technology brings our customers greater user engagement and lower bounce rates, which has been in proven to result in significantly increased revenues. imonomy also creates intelligent links between pages, which encourages visitors to easily navigate to additional relevant content. The visual semantic engine can be implemented easily on any website, and we also provide a free API that expands the functionality and the systems abilities.

Article courtesy of TechCrunch

Twitter today made the latest push in its bid to cozy up to Madison Avenue and the world of big-budget advertising, by tapping more into the kind of mainstream mediums where advertisers like to spend their money. In New York, during Internet Week, the company announced Twitter Amplify, a way of bringing real-time video into the site, with initial partners including the broadcasters BBC America, FOX, Fuse and The Weather Channel. It is part of the company’s bigger push that its calling Twitter4Brands, which first kicked off almost year ago exactly, also at an event in the Big Apple.

The instream broadcasting clips will be very closely tied to ads. This is something that it has already been doing with partnerships with, for example, the NBA, where a video also features a link to an ad:

Speaking at the New York event, CEO Dick Costolo talked about how the company has made advertising a more “frictionless” experience because of its emphasis of real-time updates. It’s clear that adding more broadcasting-like experiences into Twitter will further that concept.

The company threw in some fun ad-land perks: a Q&A session with Glee actress Jane Lynch and a Tweeting vending machine churning out swag.

Twitter has been making increasingly strong moves this year to get its platform to be more ad-friendly (and revenue-friendly). That kicked off in February with the launch of an advertising API so that larger advertisers can better manage their campaigns on Twitter; an improved advertising analytics dashboard; and Google AdWords-style keyword targeting (TC coverage here, here and here). Just earlier this week the company also unveiled the official launch of Lead Generation Cards, something Twitter had been testing for a while already, which lets advertisers include actions like requests for more information that users can get automatically by clicking a button in an advertising tweet. (You can see how this last one also sets the stage for Twitter making the leap into commerce, with one-click purchasing.)

While Twitter has not provided any official public guidance on how much it expects to make in advertising this year or in the future, there has been a lot of speculation about the number because many expect Twitter to go public, with a likely date in late 2013 or 2014, according to observers. A report from eMarketer in March noted that it was raising forecasts for the company to $583 million in 2013 and $950 million in ad sales in 2014, 60% coming from mobile.

The stats that Twitter’s president of global revenue, Adam Bain, provided last year shows just how much the company has grown over the last year. Bain noted at the time that Twitter had 140M+ active users; now that figure is estimated to be closer to 300 million.

Bain also had noted that 55% of users access Twitter on mobile, with 40% growth quarter over quarter, and that among Twitter’s active users, only some 60% actually tweet, but all of them “listen.” And in a sign that Twitter was always going to figure out a better way of leveraging ads on the platform, even a year ago, some 79% of people on the site were already following brands.

More to come.

Image: Jim Prosser

Article courtesy of TechCrunch

Cloud encoding vendor Encoding.com launched Vid.ly a couple of years ago to provide video creators with a way to publish a single universal video URL and then have that content accessible on any device. Now it’s providing a way to monetize those videos, thanks to an integration with ad delivery platorm FreeWheel.

The idea behind Vid.ly is that Encoding.com does all the hard work of encoding it into as many video formats and renditions as necessary, then serving up the appropriate copy of the video depending on which device was accessing it. In addition to transcoding, it also provided all of the storage, video player technology, device detection, streaming, and analytics needed by video creators. Customers could simply connect with the Vid.ly API and have a single universal URL created for them.

All of that’s great, especially for brands and agencies and marketers who wish to make their videos playable for all audiences on every PC, mobile phone, or tablet. But what Vid.ly didn’t provide (until now) was a way to monetize all of those videos. Hence, the partnership and integration with FreeWheel.

By integrating with FreeWheel’s ad-serving platform, Vid.ly will be able to provide all the same convenience and reach to publishers, but it will also enable them to monetize those videos across all those devices. By connecting with Encoding.com’s user interface or API, when a video is requested, Vid.ly will pass along user info to the FreeWheel ad server and pass along targeted ads along with the video. Pre-rolls, mid-rolls and post-rolls, as well as banner overlays, will all be supported.

Encoding.com has raised $4.5 million since being founded in 2008. While Vid.ly is a growing piece of its business, the company is still primarily focused on providing cloud encoding services to a growing number of publishers moving their content online.

Article courtesy of TechCrunch

GiftCards.com, a Pittsburgh-based company that has been around for more than a decade and has sold 5 million gift cards, agreed to buy San Francisco startup Giftly to grow out a mobile platform.

The terms of the deal weren’t disclosed, but Giftly had raised about $2.8 million from investors including Baseline Ventures, SoftTech VC, Floodgate, Thrive Capital, and Techstars’ David Tisch.

Giftly’s acquisition follows a number of other ones. Karma was picked up very early by Facebook although it may not produce meaningful revenue for some time for the social network, according to its earnings results earlier this year. Another gifting startup, Giftiki, which pooled together people’s money to get gifts, was acquired by Launchrock.

Giftly built a platform that avoided the hassle of individually dealing with merchants and point-of-sale systems. They came out with a native mobile app last fall that made it easier to send presents to friends and family.

The company’s platform didn’t put any limitations on what kinds of presents you could send because the company had a web of relationships with banks and credit card processors. When a recipient would go to redeem their gift, they would pay out of their own pocket, but Giftly would reimburse them that amount through their credit card.

GiftCards.com said Giftly will be rolled into their operations, but will maintain offices in San Francisco.

“We will continue to build out Giftly,” said Giftly’s CEO Timothy Bentley. “Our backend infrastructure will be used for their next generation products. We’ll continue to expand
the ways our technology and services are available to developers, through our API, and merchants, through our merchant services.”

The company is also looking to raise a first venture round, even though it’s been around for more than 10 years. That round will go toward completing the acquisition of Giftly. GiftCards.com has been around since 1999; they sell personalized, pre-designed and discount gift cards.

Article courtesy of TechCrunch

Runscope launched at the Glue conference today with $1.1 million in seed funding from True Ventures and Andreessen Horowitz for its tools that monitor API traffic and address the problems with broken APIs. Also participating were Lerer Ventures, and a group of prominent angel investors.

The Runscope tools come as software is scaling everywhere. And with software comes APIs, which are just a natural way to connect services. When web services ruled back in the IT glory age, software integrations were complex and expensive. Today REST-based APIs make it easy to connect apps. The problem is in the complexity of the distributed nature of building them.

Distributed apps have their own code. It may work across on-premise servers and a cloud provider’s network and servers. It will also use the API providers’ set of code and servers of their own.

It’s from this premise that Runscope built its tools. Designed initially for test and development, the tools will be offered for use in production in the coming months.

“The first thing we want to do is to start to give visibility into the conversations an app is having with other services,” said John Sheehan, Runscope’s CEO. This means building a tool that watches the API traffic and makes problems really visible.

“The real true promise is an app that has multiple distributed pieces but acts as one coherent application,” Sheehan said. “We are starting to realize it but the tools are geared toward building old style applications. For example, performance monitoring tools expect thart code issues are happening on your own servers.”

Runscope is built on Amazon Web Services EC2 and programmed in Python. The CTO, Frank Statton, was lead engineer at Twilio. The system is fault tolerant and was built by Ryan Park who comes from Pinterest. Sheehan worked at IFTTT before co-founding the company with Statton. Prior to IFTTT, Sheehan also worked at Twilio as the company’s developer evangelist.

The business model will in part stem from Runscope’s runtime, which generates data about API traffic. That data can be used for business intelligence. There is also the resilience the service can offer for customers and the assurance that mission-critical apps and their APIs can be watched and issues resolved before they become a major issue.

New Relic looks at the issues that Rusncope addresses from a performance standpoint, especially with a mobile SDK. API management providers such as Apigee and Layer7 are also playing in the space.

Article courtesy of TechCrunch

Truecaller, the Sweden-based creater of a crowdsourced phone directory app and online white pages service, has opened its API to a select group of “handpicked” developers. Truecaller said its directory now contains some 960 million phone numbers, either contributed by individuals or harvested through partnerships with  other directory services. The API covers only the numbers Truecaller has in its own datacase, i.e. not partner numbers, meaning it covers around 600 million digits.

Truecaller’s numbers are global in scope, and include landline, mobile and pre-pay digits — the latter category giving it an edge over other directory services, it argues. Being as phone numbers amount to highly sensitive data in the wrong hands, Truecaller is being careful about who is getting access to its API — hence no open API. Telemarketing companies are specifically barred from getting their wires in. Being the company that helped spammers is clearly not the kind of publicity Truecaller is hoping for here.

One scenario where it envisages its API being a benefit to others but also without causing irritation to phone number owners is for call centres to identify who is calling before starting a call. Truecaller’s API allows for reverse number lookup, meaning developers can attach a name to a known number. It also returns a ‘Spam score’ to indicate if a number is a likely spammer (e.g. telesales or robocalls) and — at the other end of the spectrum — a ‘True score’ to indicate how important the number is. This score is “the measurement of how popular a phone number is with our users over time”.

Name search is not included in the API but remains solely a feature of Truecaller’s mobile app. Truecaller is charging developers to use some of the features of its API, so this is clearly part of its monetisation strategy. Its free API includes only how popular a phone number is. Pricing for the more fully featured APIs starts at $299 per month, rising to $4,999.

Truecaller said cloud e-signing company Scrive has been trying its API — as a way to validate the identity behind a phone number.

Asked about the types of customers it is envisaging for the API, Truecaller CEO Alan Mamedi told TechCrunch: “We’ve had more than a thousand applicants till now even as the API was unannounced. However, we evaluate all of them internally and in all cases test their application before given access. For the time being, the developers and companies that have been given access to our API are developing for B2B services.

“I believe the Truecaller API will benefit various companies such as major airlines like Delta Airlines to improve their customer support and experience (greeting by name, decrease waiting times by connecting incoming name to ticket information), but also identify well networked and loyalty members based on their True score.”

Last September Truecaller raised a $1.3 million Series A from Open Ocean, with the aim of expanding its footprint in its key markets of North America, Asia and the Middle East.

Article courtesy of TechCrunch

Basho Co-Founder Antony Falco has raised $3 million for Orchestrate.io, a database API similar to Twilio in its capability to ease the complexity of adding features to mobile and web applications. True Ventures led this initial round joined by Frontline Ventures and Resonant Venture Partners.

Falco, who left Basho a few months ago, said Orchestrate.io solves the problems that developers face when building feature-rich applications. Often it means adding multiple databases for geo-spatial, time series or any number of other features.

The database problem has been ongoing. It in part stems from the limits of scale with relational databases. Over the years, companies like Amazon and Google reached their own ceilings and were forced to develop new kinds of databases for high-volume queries. The result is a lot of time spent babysitting databases so the applications run well.

Orchestrate.io acts as a service on a service, abstracting the database layer. Twilio successfully simplified the way developers accessed services, such as SMS and voice. Falco sees a service that also allows developers to add features by pulling the data through an API . “The comparison with Twilio and Sendgrid is not around the problem we solve but the pattern,” Falco said in an email interview. “We are taking a complex and burdensome task — running lots of databases — and putting it behind an API that programmers can use to more quickly build apps. Twilio and Sendgrid both do a similar thing, vastly simplifying the complex, for telecom and email infrastructure, respectively.

Orchestrate.io uses in-memory technology for its service. “Memory — storing indexes and hot data in memory — will be critical to performance,” Falco said. “There are three tiers – the active data and indexes in memory, disk storage for durability and data less often accessed, and as data ages and becomes inactive, a cheaper tier of fault-tolerant storage. The more we serve reads out of the memory, the better our performance will be and, without a lot of latency, users will be able to execute relatively rich queries that might require three or four queries, made sequentially, to separate databases.”

Orchestrate.io is using open source databases to build the service. “We aren’t going to build databases,” Falco said. “The databases themselves can change; we are not tied to any one database. Riak (a Basho service) is of course ideal for this use case — for forming part of the foundation of this service. But other than that, we aren’t really tied to any one thing.”

The company will use multiple data centers for its service to help get the data as close as possible to the application and the user. That makes sense considering the potential performance issues that may come when a large enough group of users are using a service that is just in one place.

For example, an application may be installed in Amazon Web Services East region, and there might be a large number of users in London. Orchestrate will have a large enough data center footprint across different providers to accommodate users no matter their locations.

The interesting story for me is about the future of the database. The real gold is in the data, but it is like a pool of oil without a way to access it. Databases access the data, organize and make it available for query. It’s inefficient. And that’s just when a developer is dealing with one database. Add a few as the features build out and the developer faces a Rube Goldberg system. It’s about getting the work done, not herding cats in a data center.

Article courtesy of TechCrunch