Tag: tech

Twitter Says Its Health Is Improving While a Second Opinion Paints a More Grim Picture

If Twitter’s latest financial reports are to be believed, the company’s prognosis is good.

The latest reports released show $909 million in revenue and a $225 million net profit, a surge in profitability that is being attributed in part to a 16% decrease in abuse reports. This measure is an essential one to gauging Twitter’s “platform health” metric. The company has pledged to focus on this measure in the new year, cracking down on abusive accounts and pledging to “reduce the burden on victims of abuse” to report their experiences, as well as “taking action before abuse is reported.”

To that last goal, however, a University of Iowa study has identified major deficits in Twitter’s ability to effectively take action in a timely manner. Computer science professor Zubair Shafiq, along with graduate student Shehroze Farooqi, built a tool to automate and identify the causes of problematic tweets, Engadget reported. Although many of Twitter’s public efforts to improve platform etiquette and public discourse has centered around removing accounts created by bad-faith actors, Shafiq and Farooqi found vulnerabilities in a different method of account abuse: scamming and spamming done through third parties using the platform’s API.

How big of a problem did they uncover? By one count the pair took, 167,000 apps accessing the API have been used to “spread disinformation, spam, and malware.”

In one of the study’s most startling findings, Shafiq and Farooqi were able to identify an account’s potential to be abusive on a large scale, from its first seven tweets. By comparison, Twitter’s standard protocol looks into an account’s patterns of abuse after it has tweeted 100 times. “[A]ll sorts of nefarious activity remain undetected by Twitter’s fraud-detection algorithms, sometimes for months, and they do a lot of damage before Twitter eventually finds and removes them,” Shafiq said specifically of the abuse that was being amplified by spam accounts.

As you might expect, Twitter takes issue with the findings. “Research based solely on publicly available information about accounts and tweets on Twitter often cannot paint an accurate or complete picture of the steps we take to enforce our developer policies,” a spokesperson shared with WIRED when confronted with the findings. And while the company has looked deeper into the apps with access to its API and is purging those using it with malicious intent, the University of Iowa pair insists it isn’t enough. The process takes too long, is too shortsighted, and victims of abuse are left with little protection.

For many of these victims, abandoning their accounts could be a more attractive option than waiting for Twitter to help defuse the situation. This departure may be why Twitter often shares massaged stats like monetizable daily active users (mDAU), which reportedly rose from 124 million to 127 million from Q3 to Q4 of 2018. Comparatively, monthly active users, a more representative number of who’s using Twitter on a regular basis, dropped from 336 million to 321 million in the same time frame.

The platform gives itself a clean bill of health; an independent observer says it’s sicker than we might expect. Who’s correct? While the platform has come a long way from its “bottom” in 2016, when advertisers were steering far away from Twitter’s reported toxicity, it still has some healing to do. With any luck, this latest wake-up call will galvanize the platform to action in continued pursuit of a “healthier” online space.

Join 100,000+ fellow marketers who advance their skills and knowledge by subscribing to our weekly newsletter.


The post Twitter Says Its Health Is Improving While a Second Opinion Paints a More Grim Picture appeared first on Social Media Week.


Long Overlooked IGTV is Coming To Your Instagram Feed…Whether You Like It or Not

No longer content to wait for you to navigate to it, IGTV is coming to you. Starting this week, minute-long clips of Instagram’s live broadcast content will now be placed on your main Instagram feed.

This move was undoubtedly made to combat the relative apathy that the platform’s feature has thus far been greeted with. Launched last June, IGTV allows users to film and share up to 10 minutes of content (creators can share up to an hour of content there). Despite access being moved adjacent to the exponentially more popular Stories feature, usage still has not picked up. Apparently left with no other choice, Instagram has opted to simply not let you avoid it.

The challenge isn’t one of users being willing to watch mobile video: Instagrammers watched 60% more video as of mid-June versus the prior year, according to Business Insider. Rather, it seems to be one of enticement. While it’s possible that users could navigate to IGTV from the Explore tab or while swiping through Stories, few have. By placing these previews on the feed, Instagram is dangling a promise of good content—or so we’d hope—in a space where users are regularly interacting.

But this approach has its risks. The main feed is already chock full, between content from people users choose to follow, sponsored content and ads that users have less control over, and suggestions of Stories to watch. As users lose control over what appears in their feed, the attention that has been paid thus far could wane. As Engadget’s Nathan Ingraham put it, these previews are “yet another piece of content invading the Instagram feed that no one is really asking for.” We’ll have to stay tuned to see if more prominent placement will shore up its user numbers, or if its lack of viewers continues to hamper its growth.

Join 100,000+ fellow marketers who advance their skills and knowledge by subscribing to our weekly newsletter.


The post Long Overlooked IGTV is Coming To Your Instagram Feed…Whether You Like It or Not appeared first on Social Media Week.


Spotify Eyes Its First Content-Based Acquisition in Gimlet Media

Spotify has said several times in recent months that the next frontier they aim to tackle is the world of podcasting; this week, they appear to be literally putting their money where their mouth is.

The streaming service is, per Recode, in advanced talks to acquire Brooklyn-based podcasting company Gimlet Media for nearly 200 million dollars. Responsible for popular offerings like Reply All (about the complicated relationship between the internet and its users) and Crimetown (about the impact organized crime has on the cities that incubate it), as well as the recently adapted Homecoming, Gimlet also creates sponsored content for companies like Gatorade. It last raised funds in 2017, and was valued at 70 million dollars.

For Spotify, the move has a number of incentives. First, it allows the demand of podcasts to be met on its platform in new and unprecedented ways. Following the decision to allow all podcast creators to put their offerings on the platform earlier this year, listening in that genre of audio grew 175% in 2018. At CES, the company’s Head of Studios and Video Courtney Holt noted a desire to make the platform even more attractive for fans of the form—while courting new users who currently use it for music. “People who consume podcasts on Spotify are consuming more of Spotify, including music,” she said, while also touting the natural fit of another audio medium for the platform. “We’re a great audio platform, and we really wanted to focus on providing the best audio experience. We found podcasting to be a great complement to our music service.”

While Spotify has excelled in conquering the world of streaming audio, its attempts to establish a foothold in video have sputtered; here too, Gimlet may be able to help. Their first foray into the arena came with Amazon Prime’s Homecoming, a Sam Esmail-produced adaptation of their show of the same name. Starring Julia Roberts and Stephan James, the show garnered critical acclaim—and could provide an infusion of credibility for Spotify in the space.

With this blockbuster acquisition, will likely also come an influx of spending in the podcasting advertising market. In 2017, podcasting generated an estimated 315 million dollars in ad revenue, but that number is growing quickly…and it will be more attractive to do so if access to podcast offerings becomes more ubiquitous through platforms like Spotify.

Join 100,000+ fellow marketers who advance their skills and knowledge by subscribing to our weekly newsletter.


The post Spotify Eyes Its First Content-Based Acquisition in Gimlet Media appeared first on Social Media Week.


How To Create Engaging Social Videos With Splasheo

Video content is in high demand for social, largely for its ability to help tell a story in a relatable and evocative way. And by this point, we’re sure you know how we feel about stories. But, it’s not always easy to get right. Preferred aspect ratios vary across platforms, editing can be challenging to do well and at the pace that keeps us competitive, and closed captioning is needed for video watched without sound or by hearing-impaired viewers. Splasheo has managed to take all those challenges, and address them in a slick and intuitive way.

What Size Should It Be, Again?

Creating videos for social on Splasheo starts by selecting a targeted aspect ratio, optimized for your desired platform. You can create square, horizontal, or vertical videos, expressly designed to work on mobile—but not at the expense of the desktop experience. By working with templates, you can ensure the video will fit neatly into the feed of the targeted platform…minimizing distraction and allowing you to highlight your brand.

Editing Made Easy

Many of us love the idea of creating mobile content, but dread the editing process that is essential for high-quality results. Splasheo seems to have solved that issue, making it easy for you to add your company’s logo, embed your brand’s colors or a call to action, and promising to “take care of the rest” once you’ve selected a template and uploaded a file. With a 48 hour turnaround, they’ve got the expertise and capacity to do the editing that would likely take your team far longer.

Worry-Free Captioning and Transcripts

85% of Facebook videos are watched without sound, which means you need another way to tell your story if you want to ensure your content fully reaches every viewer who comes across it. For most of us, captioning can address this reality…but it is challenging and time-consuming to do. This may be the most attractive feature of Splasheo. The videos they create on your behalf are transcribed, and with captions “burned into” them. This way, the captions look the same no matter where your video ends up. Further, you have access to transcripts, should viewers or prospective clients ever need them.

Splasheo has clearly identified the biggest gaps in the video creation process, and has responded with a service that fills them in nicely. The comprehensive nature with which they handle these challenges for you, make it well worth its price tag.

Join 100,000+ fellow marketers who advance their skills and knowledge by subscribing to our weekly newsletter.


The post How To Create Engaging Social Videos With Splasheo appeared first on Social Media Week.


3 Ways to Bring Out the Best Stories on Your Brand’s Podcast

The 2019 global theme for Social Media Week is “STORIES,” and we’d be remiss if we neglected the role that podcasting is playing in telling our most interesting, necessary, and entertaining stories.

But a microphone, an Internet connection, and a host or set of hosts does not a good story make. It takes considerable thought, intention, and planning to ensure that your recording isn’t done in vain—or worse, that your listener’s time and energy isn’t spent in vain. After all, as we aim to elevate the role of stories, we believe that “with great influence, comes great responsibility.” How can you ensure that your time and energy produce the most engaging and entertaining stories possible?

Begin With the End in Mind

This tip admittedly belongs to Stephen Covey, but it’s an essential one for any content that you want to tell a story. As you consider a show topic or interview subject, think first: what do you want people to know about this thing (or person) by the time they finish listening? Our brains follow stories easily because they embrace narrative structure, present information in a digestible and easy-to-follow fashion, and answer the questions they pose.

As you conceive of a podcast episode, or of a podcast season (if you plan to connect episodes or tell a long-form story), these tenets have to stay front of mind. They will affect the questions you pose and ideas you share, they’ll affect how episodes and seasons are edited, and they’ll affect how you and your brand or organization are perceived as a result. Television showrunners will develop a “bible” to define the parameters of the world they’re creating; podcast creators and producers can do the same to help structure the stories they’re about to tell.

Think Like a Rookie

Anyone in a position of authority with information – an author, a keynote speaker, a podcast host – is in a position to teach someone, something. As such, it can be helpful to look at your task ahead through the mind of someone with less experience.

What would a “rookie” want or need to know from what you’re about to tackle? If you’re interviewing a luminary in a field or industry, what could they have to offer someone aspiring to their heights? What would you have wanted or needed to know when starting out, and how can the stories you tell address those gaps?

Don’t feel burdened to have all the answers, however. References to books, TED talks, or even other podcasts are fair game and can send curious listeners off on tangents that can fortify the information you do share. Feel free to refer to sources often, citing them in show notes for those not in the position to catch them at first listen.

Think of it as curating a syllabus for your listener; even as you aim to tell a complete story, consider sharing recommendations for new stories to embark on as they absorb yours.

Be Interested to Be Interesting

As a frequent interviewer myself, I always feel an inner twinge of glee when a query is met with the phrase, “that’s a good question.” Believe it or not, asking good questions doesn’t always come naturally—I’ve been to enough panels and book readings to confirm that notion, trust me.

The best questions come from a place of deep curiosity. Being truly curious about the topic at hand, whether you’re approaching it with an interview subject or not, makes for good listening. While you should honor what your audience wants to know on a topic, the excitement will shine through when you also honor what you want to know about it. Finding out what questions have already been answered, and framing your queries to supplement or further explore those details, is far more interesting than having a compulsory set of questions—or worse, posing questions with Googleable answers.

If you do employ a structure where the same questions or segments are employed, be mindful of getting too robotic as you engage in them. Provide yourself time and space to explore the topic or interview subject as fully as possible. Remember, you’re not a substitute teacher with an agenda to get through, you’re a facilitator of learning: able to provide some insight, but ultimately preparing listeners to learn and grow on their own. And we often don’t remember the subs who came to the classroom and held us to routine, but we do remember the educators who sparked the interests that sent us on new learning journeys.

Want to hear how we do it? Social Media Week’s Lead2Scale podcast aims to provide you with practical insights and emerging trends in social media, through interviews with authorities and trailblazers in the field. Give it a listen wherever you find your podcasts! Join 100,000+ fellow marketers who advance their skills and knowledge by subscribing to our weekly newsletter.

The post 3 Ways to Bring Out the Best Stories on Your Brand’s Podcast appeared first on Social Media Week.


YouTube Upends Its Mobile Experience with Swiping Navigation

YouTube is swiping right on…well, swiping.

This week, users who update the app will notice new functionality to navigate through watched and recommended videos: swiping. “Previously, YouTube viewers had to active tap to move between videos, both on the site and on the app,” Popular Science noted; but “now, app users can swipe left to see the video they were just watching, and swipe to move to the next recommendation in the queue.”

Ease of use was certainly a motivator in this fundamental shift in the app’s navigation, and the decision to swipe was likely an easy one to make. Given how the motion is familiar to users of dating sites like Tinder and Bumble, as well as with “stories” enabled platforms like Instagram and Snapchat, YouTube seems like a logical additional place to employ the motion.

Incorporating swiping navigation, as well as the relatively new “double tap” to skip forward or back on a video, represents a necessarily more sophisticated understanding of mobile interfaces. 70% of YouTube viewing takes place on mobile now, TechCrunch reports, and the company is responding in kind with a number of improvements to their mobile interface.

Swiping and double tapping join functions like screen time controls, dark mode, in-app video sharing and messaging, all designed to give users more agency over how they experience YouTube on their mobile devices.

But it’s possible the move was one designed to provide a little less agency as well: in addition to exposing viewers to videos, creators, and stories they may not have otherwise found, “YouTube is most likely hoping that it could lead to longer hours spent on the platform, since it feels more natural swiping on a touchscreen display than having to scroll down and tap on the next video to play,” Engadget speculates.

iOS users can expect to see the feature rolled out this week; Android testing has been taking place for the feature since August, but there’s no word on when the feature will deploy for those users.

Join 100,000+ fellow marketers who advance their skills and knowledge by subscribing to our weekly newsletter.

The post YouTube Upends Its Mobile Experience with Swiping Navigation appeared first on Social Media Week.


How Social Media Giants Leverage Big Data And ML To Serve Users Better

The growth in social network popularity continues posthaste. As of 2018, the number of social media users exceeded 3 billion, and it doesn’t seem the situation is going to change overnight.

To get people hooked and deliver wow user experiences, Facebook, YouTube, LinkedIn, and other big players apply the cutting edge of technology, with big data solutions being the go-to option. Underpinned by artificial intelligence (AI) and machine learning (ML), these solutions let social media thoroughly analyze large amounts of user data, derive actionable insights, and, in turn, deliver hyper-personalized offerings.

And this is just one example of how machine learning solutions can be implemented in the social network environment. Read further to find out how giants like Instagram, Twitter, and Reddit are taking this advanced tech up another notch.

Instagram: In a fight against trolling

Coming in sixth on the list of most popular social networks worldwide, Instagram aims to make the platform as tolerable as possible. For this purpose, they capitalize on DeepText, Facebook’s “learning-based text understanding engine that can comprehend, with near-human accuracy, the textual content of several thousand posts per second.”

Before going live, the system was trained on at least two million comments and categorized them into segments like “bullying, racism, or sexual harassment.” Now, users just have to turn on automatic and manual filters in their account settings if they want to activate offensive comment functionality.

Image source: geek.com

To determine tone and intention, i.e. give the target word or phrase an appropriate interpretation and distinguish between abusive language and constructive criticism (across cultures and languages), Instagram’s AI also carefully studies the contextual meaning of surrounding words.

Besides, DeepText assists Instagram in detecting spam. Empowered by huge data assets and human input, the system identifies fake accounts and cleans up their spam comments on posts and live videos. This feature is currently available in nine languages, but the social media behemoth is working toward expanding this list.

To improve its AI system’s accuracy and avoid becoming an over-sanitized platform, Instagram continues gathering and analyzing new data sets.

Twitter: A step toward engaging users

Twitter, another social media giant, banks on ML to make the grade in image cropping. By using data from eye trackers, Twitter trains its neural networks to predict the areas users might want to look at — which are usually faces, text, animals, and other salient image regions.

As neural networks for saliency prediction tend to be too slow and cumbersome to make smart auto-cropping in real time, Twitter splits the process by using two techniques. The first one, knowledge distillation, is employed to train a smaller network to imitate the more powerful one and make a prediction based on a set of images and third-party salient data. The second technique, Fisher pruning, is used to delete features or parameters that are in some sense redundant, while lowering the computational cost.

Such a smart combination allows Twitter to obtain much more runtime-efficient architectures for saliency prediction and to crop images as soon as they’re uploaded — 10x faster than in a vanilla approach. This makes the uploaded photos more engaging and positively impacts the overall user experience.

Below is an example of how Twitter’s shift from a face detection to a saliency prediction algorithm redefined image cropping.

Image source: blog.twitter.com

Reddit: In a bid to improve website search

For Reddit — a vivid hub of internet news, pics, stories, memes, and videos — advanced search is of top priority. So it stands to reason the social media giant implements the best of tech to increase its searching capabilities and provide users with a custom-fit stream of high-quality content.

Aimed to make its search relevant, fast, and easy to scale with the platform’s growth, Reddit employs Lucidworks’ AI-based platform called Fusion. This helps the company successfully tackle the challenge of updating their indexing pipeline — by pulling together data from several sources into one cohesive canonical view. Also, Reddit not only indexes new post creations, but also updates their relevance signals in real time — based on votes, comments, etc.

The partnership with Lucidworks has given Reddit impressive results:
1. There was a 33% increase in posts indexed.
2. The reindex of all the website content slashed from 11 to 5 hours.
3. The error rate was down by two orders of magnitude, with 99% of search results served in under 500ms.
4. The number of machines needed to run search dropped from 200 to 30.

On top of that, Reddit excelled in boosting user experience and keeping operational costs down. Here’s how the tech stack of the revitalized search platform looks like now:

Image source: redditblog.com

A final word

From crafting personalized offers to fighting spam to enhancing search, machine learning delivers business value to an array of social media platforms. Facebook, Instagram, Twitter, and others have already found the ML-enabled solution to reap these benefits. Have you?

Join 100,000+ fellow marketers who advance their skills and knowledge by subscribing to our weekly newsletter.

The post How Social Media Giants Leverage Big Data And ML To Serve Users Better appeared first on Social Media Week.


PODCAST: The Rise of Messaging Bots

This week’s episode of Social Media Week’s Leads2Scale podcast features Kieley Taylor, Managing Partner, Global Head of Social at GroupM, as well as Mike Blake-Crawford, Strategy Director at Social Chain.

GroupM and Social Chain are in many ways very different, but in this episode, Kieley and Mike both picked up on a trend that has become too big to ignore — messaging bots.

Featured in this episode, they discussed:

  • The potential in messaging bots to reduce the cost of customer care
  • How messaging bot technology can create a huge opportunity for brands to develop meaningful interaction with customers

Listen and subscribe via the following platforms: Anchor, Apple Podcasts, Breaker, Google Podcasts, Pocket Casts, RadioPublic, Spotify, Castbox, Overcast, and Stitcher.

If you have suggestions for who we should interview or what topics you would like us to cover, please reach out to us at leads2scale@socialmediaweek.org.

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post PODCAST: The Rise of Messaging Bots appeared first on Social Media Week.


AI Researchers Warn of Urgent Need of Algorithm Supervision

It has been yet another drastic year for Artificial Intelligence. The public has applauded its achievements, but also couldn’t help but frown upon its faults.

Facebook has gone under fire for facilitating a genocide in Myanmar; Google was accused of building a censored version of their search engine for the Chinese government; not to mention the Cambridge Analytica data scandal which got under every Facebook user’s skin. The list goes on.

It seems like the public is in a blind spot, where they don’t know enough about AI, but are the victims of anything AI has done wrong. This is the focus of the annual report just launched by AI Now, a research group made up of employees from tech companies including Microsoft and Google.

The report elaborates on what the researchers termed as “the accountability gap,” meaning there are many social challenges of AI and its algorithms that are urgent to be addressed but remain unregulated because the public lacks the tools or knowledge to hold tech giants accountable. It also puts forward recommendations on the steps to be taken to address these problems.

The case studies showed in this report are nerve-racking. According to the report, throughout the year, it has been actual people who suffered from the fails of experimental AI systems. In March, AI-powered cars killed drivers and pedestrians; in May, an AI voice recognition system developed in the UK falsely detected immigration frauds, canceled thousands of visas and deported people; in July, IBM’s Watson system gave “unsafe and incorrect treatment recommendations.” It’s horrifying to think that a lot more cases remain unreported.

Usually, AI software is introduced into public domains with the purpose of cutting costs and increasing efficiency. However, results from these actions are systems making decisions that can neither be explained, nor appealed. “A lot of their claims about benefit and utility are not backed by publicly accessible scientific evidence,” AI Now’s co-founder Meredith Whittaker told The Verge.

The report puts forward ten recommendations on securing a healthier future of AI, among them are the need to have more precise supervision systems, and matching marketing promises to reality.

First, sector-specific agencies need to be put in place to oversee, audit and monitor tech companies that are developing new systems. The report claims that a nationwide, standardized AI monitor model will not meet the specific requirements needed for detailed regulation, especially since domains like health, education, criminal justice, etc. all have their own frameworks, hazards and nuances. For marketers, when implementing AI systems to different clients, this should also be kept in mind so that technologies will not be turned against us.

Second, marketing promises should be accurate when it comes to AI products and services. This especially applies to consumer protection agencies to ensure “truth-in-advertising” laws to AI products. The reports warns AI vendors to be held to even higher standards for what they can promise, especially since the scientific evidence that are supposed to be backing these promises is still inadequate.

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post AI Researchers Warn of Urgent Need of Algorithm Supervision appeared first on Social Media Week.


Facebook Watch Adds Cult Classics ‘Firefly’ and ‘Buffy the Vampire Slayer’ to Attract Millennials Viewers

Last Friday, Facebook Watch made an effort to boost its video streaming service, by dropping every season of “Buffy the Vampire Slayer,” “Angel” and “Firefly” on the platform.

These three Whedonverse shows were added to the streaming service as a result of a deal between Facebook and 20th Century Fox Television.

Facebook launched Facebook Watch, its on-demand video service, back in August 2017 and become available internationally to all users a year later, but has yet to gain popularity among the majority of its users, especially among already existed competitors like YouTube and Netflix.

With these three new shows, Facebook Watch is also encouraging viewers to watch them with its Watch Party feature, which allows users to watch videos while discussing in real-time with friends.

“What we’ve been focused on Watch is building a people-centric video platform, creating a social viewing experience where you can connect with other people who love the shows, and even the creatives who worked on them,” Matthew Henick, Head of Content Planning and Strategy for Media Partnerships for Facebook, said in a conversation with Variety.

Watch Party is a feature launched to Groups in July 2018 and has now been made available to all Pages and users, meaning anyone can start a Watch Party with friends directly from the videos they are watching, or by sharing on their Timeline. Since Watch Party’s launch, more than 12 million Watch Parties have been streamed, according to Variety.

What makes it even more fascinating is in Facebook Watch’s choice of streaming these particular three shows, which dp have a cult following two decades ago but don’t seem to be the favorites of Millennials and Gen Z. It thus seems to be Watch’s effort to focus on attracting older Millennials.

“We are excited to bring iconic pop culture favorites like these series for their avid fan communities to experience them in new ways, and for new fans to discover these awesome 90s classics,” said Facebook’s VP of Video, Fidji Simo, in her Facebook post.

Although, as CNBC reported, it can also be the hint of Facebook’s new strategy in reaction to teens leaving the main Facebook platform, with a new focus on viewers in their 30s and older.

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post Facebook Watch Adds Cult Classics ‘Firefly’ and ‘Buffy the Vampire Slayer’ to Attract Millennials Viewers appeared first on Social Media Week.


With Amazon Answers, Users Can Help Alexa With the Tough Questions

Anyone who’s ever played “Stump Alexa” already knows: while Alexa is extremely helpful, her information retrieving powers do have limits. A new information-sourcing initiative through Amazon is aiming to change that.

After an extensive test in-house that added more than 100,000 answers to the voice assistant’s database, Amazon announced this week that it’s opening up its Alexa Answers crowdsourced answer program to select users, via email invitation. “In addition to advanced technology—such as machine learning and natural language understanding—and our many trustworthy information sources, we’re involving the Alexa customer community to help us answer questions Alexa can’t quite answer yet,” they shared in their company blog this week.

Addressing this gap in utility is part and parcel of this evolving method for search. The South Florida Business Journal notes that this new challenge is grounded, in part, by how queries are framed to a voice assistant versus how they’re framed through a search engine. “Asking Alexa questions will make you realize that your voice searches usually include more than five words and start with ‘how’ or ‘what.’ It will also help you understand the growing need for pages on your website that have ‘conversational content’ that, when searched, can deliver understandable answers to voice queries.” Alexa Answers will aim to make answers available to Alexa that are framed in this easily accessible manner.

It’s worth noting that Amazon users are currently able to inform one another in a number of different ways. In their blog, Amazon reminds readers “for nearly 20 years, we’ve allowed customers to offer their input on Amazon.com products through customer reviews and community-based answers.” The fidelity of this input has been addressed through features like “Verified Purchase” reviews. For the time being, this important process is even more tightly controlled; at this point, editors are being invited via email, and the pool of Alexa users selected to inform the voice assistant is small. When these crowdsourced answers are shared, they will be prefaced by a disclaimer of sorts indicating them as having come from an Alexa user.

Although this approach has been tested heavily with sites like Quora and Wikipedia, there are still some risks to having humans provide information to AI. Recently, Google had to disable portions of its SmartCompose and SmartReply software after its challenges with gender and gender pronouns. Two years ago, Microsoft’s Tay had to be taken offline after Twitter users supplied it information that made it racist.

The email-invite nature of this project may prevent some of that, but Amazon would do well to pay close attention to the data supplied, and adjust as needed to avoid embarrassing or damaging gaffes. Nevertheless, a very real danger that exists for Alexa Answers is a danger that exists whenever data is supplied to AI tools, and The Next Web’s Vishwam Sankaran put it best in reference to the Google story: “AI is only as fair as the data it learns from.”

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post With Amazon Answers, Users Can Help Alexa With the Tough Questions appeared first on Social Media Week.


Google Sunsets Allo, Struggles to Find Its Niche in Chat

Another Google messaging solution has bitten the dust.

Last week, Google announced via its company blog that its long under-utilized chat app, Allo, will be sunsetting operations in March 2019. “We’ve decided to stop supporting Allo to focus on Messages,” VP for Consumer Communications Products Matt Kleiner wrote.

Allo originally garnered interest for its use of AI-assisted messaging; this farewell announcement acknowledged, “we’ve learned a lot from Allo, particularly what’s possible when you incorporate machine learning features, like the Google Assistant, into messaging.” Kleiner’s message was reflective in its nature, noting how Allo’s brief time in operation had improved the Messaging product by providing insight into needed and popular features. SmartReply, GIFs, and desktop support all now exist as features of Messaging because of their initial success in Allo. But the product has otherwise struggled to gain a following since its launch in 2016. Google has struggled mightily in the messaging market; the discontinuation of its latest and flashiest effort signals that they’ve not yet solved this challenge.

Elsewhere in the Google messaging space, the video chat product Duo—originally launched in tandem with Allo—will live on, bolstered most recently with the ability to leave video messages. The Hangouts product will live on as well, but will evolve into a business-focused platform with its Chat and Meet functionalities. And yet, even with these apps, the question remains: given several unsuccessful attempts, where does messaging fit in Google’s goals now?

The shuttering of Allo leaves Google with a larger challenge: finding a messaging platform and protocol that can compete alongside messaging powerhouses like WhatsApp and iMessage. TechCrunch reports that its latest big bet is on RCS, or Rich Communication Services, but this bet will be challenged by carriers who will need to develop messaging services around the tech (thus far, only Verizon has agreed). Further, RCS is an unencrypted protocol, a feature that runs directly counter to market trends. With such major obstacles, it’s clear there is still more work to be done before Google establishes a clear foothold in the messaging space.

In the meantime, those using Allo will have access to the platform through March of 2019, and will have the option to download their message archive before it’s shuttered for good.

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post Google Sunsets Allo, Struggles to Find Its Niche in Chat appeared first on Social Media Week.


Instagram’s ‘Close Friends’ Feature for Stories Emphasizes Private Sharing

Had you recently noticed a green ring, instead of a pink one, around your Instagram friends’ Stories updates? Last Friday, Instagram announced yet another update for Stories, named “Close Friends,” which allows users to share Stories with a specific closed group of people of their choosing.

To add people to your close friends’ list, just go to your profile and tap on “Close Friends.” Once done, when you pose Stories, you can choose to post them either to all followers or just this specific friends group only. Nobody other than you can see your close friends list. They also cannot request to be added.

Instagram’s move is evidence of the trend that many other social media brands are following — investing in more private and personal conversations. Stats show that 85 percent of the messages shared on Instagram are distributed to the same three friends, according to Social Media Today. This further proves that users value Instagram for its function of private sharing.

On its announcement blog post, Instagram says that it hopes to make the platform a more comfortable place for the community.

“Instagram Stories has become the place to express yourself and share everyday moments, but our community has grown and sometimes what you want to share isn’t for everyone. With Close Friends, you have the flexibility to share more personal moments with a smaller group that you choose.”

In addition, The Verge reported back in June 2017, that Instagram had been experimenting with a potential new feature named “Friends Lists,” which, similarly, would allow users to share posts with a limited group of friends.

So what does this mean for marketers? If the same model would be available for business accounts, this feature can be a great tool for brands to share customized content to specific groups of audiences or groups that the brand expect could generate more profit.

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post Instagram’s ‘Close Friends’ Feature for Stories Emphasizes Private Sharing appeared first on Social Media Week.


PODCAST: Deep Social Listening w/ Michael (MJ) Paradiso, VP, EMEA, at NetBase

This week’s episode of Social Media Week’s Leads2Scale podcast features Michael MJ Paradiso, VP of Sales, EMEA, at NetBase.

Michael joined NetBase a year ago, and has more than eight years of experience in B2B sales, team management, and sales leadership. Previously, he has worked for Meltwater, P&G, as well as General Electric.

Featured in this episode, they discussed:

  • What does “deep social listening” mean and how that has helped their clients
  • How NetBase utilizes all sorts of data to drive social media strategies
  • How the company relies on innovation to brace for future challenges in social

Listen and subscribe via the following platforms: Anchor, Apple Podcasts, Breaker, Google Podcasts, Pocket Casts, RadioPublic, Spotify, Castbox, Overcast, and Stitcher.

If you have suggestions for who we should interview or what topics you would like us to cover, please reach out to us at leads2scale@socialmediaweek.org.

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post PODCAST: Deep Social Listening w/ Michael (MJ) Paradiso, VP, EMEA, at NetBase appeared first on Social Media Week.


Stymied by AI, Google Removes Biased Smart Compose Suggestions

Earlier this year, a Google research scientist sent a message with the assistance of Gmail’s Smart Compose software: “I am meeting with an investor next week.”

The software responded, with no additional prompts: “Do you want to meet him?”

Smart Compose, at its best, is supposed to simplify the composition of emails by auto-completing sentences and sentiments expressed as the writer creates a new message. Smart Reply, a companion feature, offers a few possible responses to messages you receive. Both are powered by Natural Language Generation, an AI technology that takes in language relationships and patterns from emails, literature, web copy, and other information it’s fed to provide a type of predictive text. But as the example above shows, these processes are not immune to bias by any means. Now, while the features will stay on, neither will suggest pronouns for your messages—no matter what cues are presented in the text.

This assumption that an investor is male is based in large part on the data that its AI technology processes, but can present challenges when this data affirms bias. Even if finance, technology and engineering fields are predominantly male, the AI developed to communicate in these fields shouldn’t make assumptions about who is being spoken to or spoken of. The Next Web encapsulated the issue simply: “AI is only as fair as the data it learns from.” And recognizing the inherent unfairness from how the tool has learned, Google took the step of blocking pronoun suggestions outright.

“Not all ‘screw-ups’ are equal,” Gmail product manager Paul Lambert has said in regards to the announced change. “Gender is a big, big thing to get wrong.”

Reuters reports that eliminating this sort of suggestion wasn’t the company’s first choice to resolve the challenge. “The SmartCompose team of about 15 engineers and designers tried several workarounds,” they reported, “but none proved bias-free or worthwhile.” And yet, an elimination strategy has become relatively common when AI has failed or displayed bias.

In 2015, a photo-identification app incorrectly identified photos of a Black couple as gorillas; the company’s solution was to block the app from identifying gorillas in any form. And in 2012, women who searched topics like Computers and Engineering or Parenting were finding themselves identified as middle-aged men in their Ad Preferences. So although it is appreciated that Google understands the damage that incorrect pronouns can cause, simply disabling the functionality feels like an incomplete and short-sighted solution to a problem with far deeper roots.

Work continues, both at Google and at other companies, to identify potential points of bias and to combat them through more thoughtful machine learning. Google’s AI ethics team has aimed to start with the most hyperbolic examples. “A spam and abuse team pokes at systems, trying to find ‘juicy’ gaffes by thinking as hackers or journalists might,” Reuters reported. But the ease with which AI can deliver biased or potentially offensive predictive text proves we have a long way to go before these technologies can be deployed in an unassisted manner.

“The end goal is a fully machine-generated system where it magically knows what to write,” says Automated Insights’ John Hegele, whose company auto-generates news items from statistics. But this latest gaffe from Google proves there’s still a great deal of room for learning—for AI, but also (and perhaps especially) for the engineers that coordinate machine learning. “There’s been a ton of advances but we’re not there yet.”

The post Stymied by AI, Google Removes Biased Smart Compose Suggestions appeared first on Social Media Week.


BrainGate2 Helps the Paralyzed Access the Internet in New Ways

BrainGate came to prominence in 2017 when it allowed a quadriplegic man to move his arm through thought alone. Now, BrainGate2 is reportedly simplifying how paralyzed individuals navigate the internet, and their findings could affect how you present yourself and your brand online.

BrainGate is an array of strategically placed micro-electrodes that has been used to decode neural signals associated with movement. Previous uses of this technology have been localized to facilitate limb movement. But now, in a recent trial, three individuals were able to use non-modified tablets to use apps and explore the Web without adaptations.

The trio of participants, according to Engadget, “had electrode grids implanted over part of their motor cortex — the area of the brain that helps control movement — which picked up neural activity indicating they were thinking about moving a cursor on the screen.” Those neural activity patterns were sent to a tablet wirelessly paired with a virtual mouse. This process allowed subjects to stream music, text others, and browse the Web, even facilitating a real-time live chat between two of the study’s participants. The study’s findings brings new hope for intuitive web navigation for “people with spinal cord injury, brainstem stroke, and ALS,” among others, according to Braingate.

What’s important to note is that participants were able to navigate interfaces on the tested apps and utilities without having access to all the functions that a tablet typically demonstrates; click-and-drag and multitouch, for example, couldn’t be easily completed by these tools. For app developers, as well as companies considering hosting options for their content, knowledge of what would be easiest to navigate by modified systems is an important consideration. What’s more, this trial was conducted with Android devices, and the accessibility could be extended further by enabling the accessibility options that came standard with the device.

Although the report didn’t specify precisely what apps or utilities were used as the trio navigated the internet, any tools that employ accessible design to develop and position their content—and, by extension, any brands or organizations that prioritize accessible design—will extend the reach of their awareness and adoption. A few tips for designing or hosting content with accessibility in mind:

  • Accessible design, often, ends up being better for everyone. The avatars we develop content or platforms for often capture a narrow segment of who may actually use it. While we shouldn’t abandon avatars as a means to identify target populations, we should cast a wider net. Smart Design’s John Anderson recommends thinking about the most advanced and unencumbered user, as well as the user who may be highly challenged by our original impulse…and then advocating for a solution that serves both populations equally well.
  • Vary the formats in which you present information. Last month in New York, designer Marie van Driessche urged people to think of how many ways people could need to consume information, and then preemptively provide options. The solution, according to 99U? “Make sure your product or website includes multiple options for how to engage. For instance, when posting a video, include transcripts, video captions, and additional video of a person signing. That way, users can select the option that’s best for them.”
  • Prepare for your “fix” to evolve. van Driessche further shared, “Just as each human capabilities change over time—as they learn sign language, or lose mobility in their fingers—so too must accessible design evolve with them.” Paralysis can be temporary; similarly, those with degenerative diseases may need more assistance over time. Building products and utilities that can adapt with individuals over time will build the best loyalty as conditions evolve.

Although Braingate’s hardware is still far too cumbersome to be used in a widespread capacity, its miniaturization, and subsequent wider adoption is on the way. The best developers, designers, and brands, will be ready when the time comes.

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post BrainGate2 Helps the Paralyzed Access the Internet in New Ways appeared first on Social Media Week.


LinkedIn Protects Email Addresses in Latest Export Update

If your brand or company relies strongly on LinkedIn connections to contact prospective leads via email, your efforts have hit a roadblock in recent weeks…whether you know it or not.

A TechCrunch reader alerted the publication to this quiet change, in which exported data archives no longer feature user email addresses. The change can be traced back to a new default setting for user email. The platform asks, “Allow your connections to download your email in their data export? If no, your personal email address will not be included when your connections export data from LinkedIn.” The site defaults this toggle response to “no,” protecting the individual user’s contact information. This is good news for privacy-minded users, but presents challenges for businesses, recruiters, and other entities who depended on LinkedIn as a source for email addresses.

At a time when our personal data feels all too vulnerable, this should be a newsworthy announcement for the professional networking platform—especially given recent high profile breaches from competitors like Facebook and Google. However, two sets of circumstances may have impacted LinkedIn’s decision to keep this newest change secret. First, the company is embroiled in its own scandal. News surfaced of 18 million non-member email addresses used to target prospective LinkedIn users with Facebook ads, in an audit period ending just a day before GDPR took effect. This revelation, unearthed by Ireland’s Data Protection Commission and “amicably resolved” between the company and DPC, represents a data usage violation that paints LinkedIn in a bad light.

LinkedIn has since ceased the practice, and responded to the story with a statement from TechCrunch that stated “the strong processes and procedures we have in place were not followed and for that we are sorry. We’ve taken appropriate action, and have improved the way we work to ensure that this will not happen again.” One could wonder if the heightened protection of their own users’ addresses isn’t a reactive move to their own past transgressions.

Second, and closely related to the first revelation: few users are aware that their email addresses could have been accessed by their connections in that fashion. TechCrunch’s Josh Constine points out, “perhaps LinkedIn didn’t want to bring attention to the fact that it was allowing your email address to be slurped up by anyone you’re connected with, given the current media climate of intense scrutiny regarding privacy in social tech.” To bring such a longstanding practice up now, after news they’ve made others vulnerable, could prove even more detrimental from a PR perspective. And yet LinkedIn may find itself in an unwinnable position: by hiding this change to user privacy, they risk drawing the ire of folks who relied on this archive capability to build their customer bases or recruitment pools.

As it currently stands, the default setting for users is to protect this vital contact information from those seeking to collate it from their networks. To make your address available for such data collection (though TechCrunch correctly notes the LinkedIn help center provides little information for why one would want to do this), one need only toggle the setting by navigating to Settings & Privacy -> Privacy -> Who Can See My Email Address?

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post LinkedIn Protects Email Addresses in Latest Export Update appeared first on Social Media Week.


Google Maps Takes on Facebook Pages with In-App B2C Interaction

For years, the go-to space for local business communication has been Facebook Pages, and the company boasts 65 million pages to that effect as a result. But Google Maps is starting to stake their claim on this parket, as evidenced by the latest features added to the Google My Business platform.

This month, a pair of new features were added to the platform that specifically enhances the Google Maps experience for local businesses. The first, Post, goes beyond allowing a business to upload photos or menus, to add time-sensitive promotional offers. According to Business Insider, “The new Post button enables all businesses to upload photos to their profiles and create special offers and events for consumers.” Previously, business owners had no way to offer promotional details into the app; now, prospective customers don’t have to navigate away from Google Maps to see these deals.

Another feature that users now won’t have to navigate away for, is new functionality for direct messaging between businesses and consumers. These conversations previously had to be held outside the app (and currently remains the case with Facebook Pages); it is Google’s hope that allowing these communications in-app will boost their standing in their already supremely popular product. 67% of mobile navigation app users count Google Maps as their go-to app, while Facebook Pages integrates with the less commonly utilized Apple Maps.

These aren’t the first moves Google Maps has made to create a more social experience in their app. This summer. The Explore tab was added to display top events in a user’s area, as well as restaurant rankings and recommendations. October saw the debut of the Follow function for Android, where users could opt to “follow” their favorite businesses in the app. Further, their recent integration with Spotify makes a connection to music on your commute seamless as well. The name of Google Maps’ current game is “seamless,” and time will tell how these adjustments help the app evolve into a seamlessly connective experience for businesses and users alike.

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post Google Maps Takes on Facebook Pages with In-App B2C Interaction appeared first on Social Media Week.


Spotify Debuts Its First Analytics Tool For Music Publishers

Streaming music service Spotify just debuted Spotify Publishing Analytics, its first analytics tool built specifically for music publishers.

According to Spotify, this is their effort to take care of the needs of music publishers, a group that hasn’t been given the necessary analytics tool.

“We know that there are usually many more people involved in the creation of your favorite song than just the artist whose face appears on the billboard,” according to Spotify’s press release.

“One of our core missions at Spotify is to enable creators the opportunity to live off their art,” said Jules Parker, Head of Publishing Relations & Services, EMEA and APAC, Spotify. “The publishing community is integral in supporting the songwriters that create the music we love. With more information, publishers are empowered to make the most of the opportunities the global reach of Spotify provides, and the more information we can share with each other, the more opportunity we can help create for songwriters.”

The publishing analytics tool allows music publishers to track daily streaming stats, including playlist performance, and view data across all the songwriters on their roster. In addition, it offers insights into streams by song, songwriter and recording — how songs are doing on playlists; a way to see all the different versions of a song; and a way to export metadata to internal systems for further archival and analysis.

Spotify says that this new tool will be the first built specifically for publishers from a music streaming service, although, according to Musically.com, companies like Socan, Kobalt and Songtrust have already been providing publishers analytic tools, though not as a streaming service.

This move shows Spotify’s recognition of the fact that collecting data nowadays is more important than ever, especially for publishers, according to Patrick Joest, EVP Global Content Partnerships & Synch, BMG.

“Armed with this level of streaming data, directly from Spotify, music publishers can gain insights into new opportunities for their songwriters, more efficiently collect royalties on their behalf, and moe effectively market their works,” said Joest.

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post Spotify Debuts Its First Analytics Tool For Music Publishers appeared first on Social Media Week.


SpiNNaker’s New Supercomputer Uses AI To Assist With Tough Accents

“I’m sorry, I didn’t understand that.”

This is a common and frustrating refrain from voice recognition tools, and it can be all the more challenging or demoralizing if the speaker has an accent. But a recently switched on supercomputer is aiming, among other things, to reduce the frequency of these misunderstandings over time.

The SpiNNaker supercomputer, activated this past week, mimics the human brain by sending packets of information to a number of destinations at the same time. This mechanism differs from how a standard computer works, sending one packet of information at a time to a single receiver. In their words, the chip-powered computer “breaks the rules followed by traditional supercomputers that rely on deterministic, repeatable communications and reliable computation.”

By mimicking the human brain, the hope is that this computer will be able to assemble received information and learn over time. One of the most immediate and most urgently needed applications of such technology: learning to adapt to a speaker’s voice.

Just as humans can learn to understand individuals with accents over time—certain words will always sound a certain way, and their sound will differ from how it may have originally been programmed—means that virtual assistants like Siri, Alexa, and Cortana could learn and get better at hearing their owners over time. In turn, this means these tools will be effective for more and more prospective users over time. Developers aiming to connect their products to voice assistants can rest easy knowing that the tool will understand their users with increasing ease.

The technology being piloted is twenty years in the making, and yet has a good deal longer to go before it’s consistently effective. For those accustomed to grumbling and then repeating themselves louder or with more enunciation, SpiNNaker’s work can’t go quickly enough. But thus far, it shows promise in creating a voice-assisted world open to people the world over…no matter how they sound.

Learn the latest trends, insights and best practices from the brightest minds in media and technology. Sign up for SMW Insider to watch full-length sessions from official Social Media Week conferences live and on-demand.

The post SpiNNaker’s New Supercomputer Uses AI To Assist With Tough Accents appeared first on Social Media Week.