01 Jul 2015

Twitter Must Thrive

Twitter is an important institution that has the admirable goal of giving "everyone the power to create and share ideas and information instantly, without barriers." Here are OTTN's ideas to make sure that it thrives.

OTT Nation is a fan of Twitter the company because:

  1. its goal to give everyone in the world the power to create and share ideas and information instantaneously is an admirable one; and

  2. the company has gone a long way towards making Twitter the product into a “global town square” of sorts.

Twitter the product has changed the way that public discussions occur and how information is disseminated to the public. The ways that people use Twitter are many:

  • to self-curate and outsource curation of news, opinion and analysis,
  • keep track of, and interact with, public figures and institutions,
  • identify and participate in conversations surrounding trending topics,
  • discuss topics of particular interest with like-minded individuals,
  • mobilize social activism,
  • network,
  • show their comedic chops,
  • and so much more.

Here’s how the Library of Congress, which is archiving every public Tweet ever created, has described the importance of Twitter:

Twitter is part of the historical record of communication, news reporting, and social trends . . . . It is a direct record of important events such as the 2008 U.S. presidential election or the “Green Revolution” in Iran. It also serves as a news feed with minute-by-minute headlines from major news sources such as Reuters, The Wall Street Journal and The New York Times. At the same time, it is a platform for citizen journalism with many significant events being first reported by eyewitnesses.

Library of Congress Blog, April 28, 2010

OTTN completely agrees with the Library of Congress about Twitter’s importance. And while Twitter has quite a long way to go before it gives everyone the power to create and share ideas and information instantaneously, it has as good a chance as any to do so in a way that is a net positive to society. Furthermore, unlike some other tech behemoth contenders, Twitter seems to give (or at least appear to give) real consideration to important societal values like privacy and dignity – which quite obviously can be impinged upon by technology making the sharing of ideas and information frictionless and instantaneous.

So it’s disturbing to see Twitter the company stumble in ways that threaten to derail its progress. When Twitter’s Q1 earnings were inadvertently released early on April 28, the stock tanked, not only because of the earnings release flub, but also because Twitter’s revenue and user growth did not meet investors’ expectations. Indeed, it appears public market investors have grown increasingly skeptical that Twitter can grow its revenue commensurate with its market capitalization, much less achieve its goal of reaching every single person on the planet.

This negative sentiment is a very bad thing, as it puts undue pressure on management to take actions that may not necessarily be good for the platform and its current and prospective users. For example, Felix Salmon, an influential financial journalist, has called for Twitter management to sell the company to a bigger player in the “social” space. And more than one commentator has suggested a wholesale house cleaning at the management and board level, despite turnover being quite a regular thing at Twitter HQ. (Former CEO Dick Costolo appears to have been a casualty of this commentary, with co-founder Jack Dorsey now stepping into the interim CEO role.) OTTN won’t opine on these particular recommendations, except to say that they seem a bit extreme – and really not that constructive.

Still, OTTN has decided to join the “Here’s what Twitter should do” chorus and put out some ideas and recommendations that might help solve some of the issues threatening to undercut the company’s significance going forward. These ideas and recommendations overlap in many respects with the company’s articulated objectives of:

  1. strengthening Twitter’s core;

  2. removing barriers to consumption; and

  3. building new applications and services to increase Twitter as a utility across the world.

However, there are some additions to, and major deviations from, the company-stated goals as well.

So let’s begin.

Strengthening the Core through Design

When Twitter refers to strengthening the core, they are talking about growing the number of active Twitter users and maintaining and perhaps increasing the engagement of those users. Continuously introducing design changes that enhance the product experience is a key way to maintain and increase user engagement for consumer services like Twitter.

Twitter clearly cares about design. They have made quite a bit of progress towards making the service more usable for new users and more functional and delightful for active users. For example, Twitter has focused a lot on how to make the on-boarding process better so that new users understand how to derive value from the service. They’ve also greatly expanded the richness of content in Tweets. No longer is it just 140 characters, text and links. Users can add emojis, images (including animated GIFs) and videos, share Vines, and include links that become rich media cards.

But there’s always more to do, and there are several design fixes that OTTN would suggest to further strengthen the core.

Fixing the “Firehose”

First off, Twitter has to introduce more design fixes to solve the “firehose” problem. Typically, when people refer to the Twitter “firehose,” they are talking about monetizing the vast amounts of data that Twitter is continuously receiving from its user base. But there is also the micro-level firehose that Twitter users have to deal with every day.

Evan Williams, co-founder of Twitter, has acknowledged that there is “too much data and too much information” coming to users from Twitter, and that the company must help its users avoid being overwhelmed by the Tweet stream. Williams specifically calls for playing around with the reverse-chronological list of Tweets.

Getting away from – or at the very least, enhancing – the reverse-chron list is a good idea. The reverse-chron list fits well with the spirit of Twitter being real-time in nature, but the reality is that in order to follow real-time conversations and happenings, users often have to go back in time to get context, and filter/compartmentalize the stream of Tweets coming in from those they follow.

Pushing out conversation threads and subsequently tweaking how conversations are displayed are great examples of how Twitter has introduced context to the timeline. Another way that Twitter could enhance the timeline, with context in mind, would be to make it easier to set bookmarks in lists which users can jump back to when resuming their use of Twitter. Your author has a pseudonymous account where he interacts with the Twitter Finance community via a list stream. He periodically has to scroll all the way down to the last Tweet he remembers in the list in order to make sure he hasn’t missed some vital info from earlier in the day that would inform the chatter going on at the moment. The end result is that your author simply checks Twitter less than he might otherwise because of the annoyance of that long scroll.

Twitter’s list functionality must be improved, generally. Lists are fantastic because they allow users to self-curate the firehose into digestible streams based on their various interests. But as of now, creating and getting to lists is quite painful.

Lists functionality shouldn’t be buried the way it is in the desktop and mobile apps, as this makes it neither easy nor obvious how to go about creating lists. In fact, many new users don’t even realize that they can create lists. Just as there are key menu icons for Home, Notifications, Messages, and Me, there should be a Lists icon (perhaps, a dropdown) that allows users to create lists and quickly get to and from their lists. Adding accounts to lists is also very difficult, as the functionality is hidden under the “More user actions” gear button. There should be a dedicated button for adding an account to one’s lists, just as there is one for following an account. If the company is afraid that too many users will add accounts to lists without following them, perhaps the proposed list button should only show up when an account has been followed. Adding followed accounts in bulk to lists should also be an option.

Lists are also surprisingly devoid of algorithmic curation. Twitter lists have tabs for Tweets, Members, and Subscribers. But there is no tab for suggested Tweets to view, which would actually be quite useful for most users. Since algorithmic curation is something that the Twitter team has been working on, one would think that this feature for lists could be added in short order.

Twitter could even enhance algorithmic curation in lists by giving users the ability to associate pre-defined topics or categories with lists. These categories/topics could be based on the interest categories and sub-topics for which the company allows businesses to target advertising on Twitter. Allowing users to categorize lists would come make Twitter’s search algorithms smarter and enhance the browsing experience on Twitter. (More on search and browsing below.)

Likes, Bookmarks, and Favorites, Oh My

Another design change that must be made is to the “favoriting” functionality in Twitter.

Right now, Twitter users must use favoriting to either “like” or bookmark a Tweet. But the purpose behind “liking” a Tweet is obviously quite different from that of bookmarking or archiving a Tweet. Likes are a way of acknowledging Tweets that are clever, useful, inspiring, etc. Bookmarking is a way for users to save certain Tweets and their linked contents for later use. Combining the two types of functionality into “favoriting” reduces the usefulness of Twitter and actually confuses new users.

With regards to bookmarking/archiving Tweets, some users actually connect the service to their Evernote account via IFTTT so that they can have a searchable archive of Tweets. (Unfortunately, these users have to do this because Twitter’s search functionality is not that robust yet – more on that later.) These users tell IFTTT to send all Tweets they have favorited to Evernote, and so they only favorite Tweets that they want to make searchable. The “liking” functionality drops out for these users.

Twitter users would benefit immensely if they could split the favoriting functionality into liking and bookmarking. And Twitter itself would benefit from being able to distinguish between the two types of actions when determining the relevance of a Tweet.

Solving the “Short-Form Problem”

Twitter is very good at giving individuals the ability to publicly connect and engage with others in real-time based on interest, whether or not these individuals know each other in real life. However, it is pretty well acknowledged that Twitter is not necessarily the best place for complex, well-thought-out public discussions and debates. This is because the 140 character limit and simple messaging format – which OTTN acknowledges are fundamental reasons why the core service is conversational and, therefore, unique – inherently limit the complexity of ideas expressed.

It’s what OTTN has dubbed Twitter’s “short-form problem.”

One way that Twitter has tried to solve the short-form problem is by improving the core service’s direct messaging functionality. Private group messaging went live in January of this year, and Twitter revealed to developers last month that the 140 character limit would shortly be removed for direct messages. Both of these improvements are very welcome for Twitter users, but they do not really solve Twitter’s short-form problem.

First, group messaging is always private, and neither the public nor Twitter can benefit from valuable long-form discussions on particular matters of interest that occur behind a login. Second, even if group messaging discussions could be made public, the messaging service/chat format is an inferior one for encouraging well-thought-out discussions that involve large numbers of people. It’s simply not feasible for a large number of individuals to either contribute to or follow a complex long-form discussion in a simple chat session.

With the chat format, there is no threading, which makes it difficult to respond to a particular line of thought that has popped up. There is no ability to easily quote others or indicate which message is being referred to in a response. Participants cannot clarify what they are saying through the use of formatting in the same way one might find in, say, a well-designed web forum. The linear nature of messaging encourages real-time responses, rather than deliberated ones. Finally, discussions that exist in a chat session tend to be rather ephemeral in nature (though not in the Snapchat, self-destruct sense, of course). Rarely does one go back to a chat discussion in the same way that one might go back to a discussion that occurs in a web forum. That may have something to do with the fact that web forums have an organizational structure centered around threads on particular topics of interest.

If Twitter wants to be able to facilitate public long-form discussions, it will have to do more than improve direct messaging. Twitter will have to provide some level of functionality to its users that replicates posting on a message board or publishing on a blog. But not on Twitter itself, as conversation on the core service is supposed to be short-form in nature.

To that end, Twitter should consider tightly integrating with new long-form discussion venues, perhaps serving as a jump-off point for entering into discussions on those venues. For example, Twitter should consider tightly integrating with Medium, the publishing platform startup founded by Twitter co-founder Evan Williams.

Here’s how OTTN would like to see a Medium integration work.

Essentially, a Twitter user who wants to start a discussion with a long-form post would be able to embed a Medium story into a Tweet the same way that one can embed photos. There would be a Medium button, just as there is a camera button, in the “What’s happening?” box. Clicking that button would bring up a list of the user’s Medium stories (including drafts), as well as a Write a Story button. A single click on a published story would embed it into the Tweet. Otherwise, clicking a draft story or the Write a Story button would open up either Medium itself or an on-Twitter editor linked to Medium, and the draft or newly created story would be queued up for tweeting upon publishing.

A Tweet with a Medium story embedded into it would be fully expandable so that the entire story could be read on Twitter, as well as “hearted” and bookmarked. The reader could then respond to the author in three ways:

  1. with a Tweet reply or a Tweet embedding the original;

  2. with a Medium response, which could be either short-form or long-form, but which would only show up on Twitter as a reply to the original Tweet if the response was visible as per Medium’s own rules (see “What responses do I see?”); or

  3. with a Medium note, which would only show up on Twitter if recommended by the story author. Visible Medium responses and notes would float up to the top of Twitter conversations.

Users could also use an original content Tweet or linked-content Tweet (including one that has a card embed) to create a jump-off point for discussion on Medium. In addition to the reply, retweet, and favorite icons in a Tweet, there would be a Medium icon that, when clicked, would create a new Medium story with the Tweet embedded into it (or referenced in some way). Once the story is published, it would be posted on Twitter as a Tweet with Medium embed, and the Medium count – similar to the retweet count – for the original Tweet would go up. The new Tweet with Medium embed would stand alone but could also show up as a reply to the original Tweet if the author of that Tweet “hearted” the story.

In addition to coming up with integrations like the one described above, Twitter may have to release new apps if it wants to solve the short-form problem, but more on that later.

Contemporizing Twitter

Finally, while it pains OTTN to say this, the Twitter user experience seems to be a bit outdated in terms of visual design. In a world that’s moving towards material design and more immersive experiences on mobile, introducing some three dimensionality and physics into the Twitter experience could be a really welcome refresh for existing and prospective users.

For example, one idea would be to create an immersive view for lists, where Tweets can be swiped through in a “cover flow” or “Tinder-like” style. That would make going through a stream of Tweets far more enjoyable, and could also make promoted Tweets pop out to users in a way that is much more likely to create engagement.

The elephant in the room for Twitter is that its search capabilities are less than robust. If one of the raisons d’être of Twitter is to allow individuals to connect to the “pulse of the planet”, then being able to easily search the Twitterverse for what’s happening right now must be a core function of the service. And it must be available to both logged-in and logged-out users.

Twitter understands this very well and is working with some of the biggest names in the tech world to bring more exposure to Tweets via off-Twitter search. The company recently announced a deal that puts Tweets back into Google search, allowing Google users – most of whom probably do not have Twitter accounts – to easily discover relevant and, presumably, recent Tweets. Twitter also appears to be working with Apple to “surface great Twitter content and accounts directly in Spotlight Search on iOS and OS X.”

These third party deals are good ones. But Twitter has to do much more to improve on-Twitter search functionality. And once that robust on-Twitter search functionality is pushed out, the company must make sure that as many people as possible understand that Twitter is the place to start when trying to get real-time information about what’s going on in the world.

A personal anecdote may be of use. In April and May of this year, two deadly earthquakes hit Nepal, with the death toll at over 8500 at the time of this writing. When the second earthquake occurred, your author’s mother was informed about the earthquake from a relative who felt the tremor in Northern India. And the first place she turned to to see what was going on with the earthquake was . . . CNN. But CNN wasn’t yet reporting on the second earthquake. Nor were any of the Hindi-language news channels that she had access to.

At no time did your author’s mother even think to turn to Twitter to discover news on the aftershock. When your author tried to probe into her understanding of what Twitter was, she basically referred to Twitter as the place where anybody can just broadcast their feelings to the world and follow celebrities. Imagine if she instead thought of Twitter as a place where you can search for anything that is going on right now and discover relevant and contemporaneous impressions, news, audio, video (both recorded and real-time), and more?

Again, it is quite clear that Twitter is working on solving on-Twitter search – and has been for a long time. Indeed, Twitter just rolled out a new search interface to logged-in users. The new UI is a great start, as it allows users to filter by top Tweets, live Tweets, photos, and more.

However, OTTN believes that the on-Twitter search experience has to be much more immersive to get both logged-in and logged-out users to actually use that functionality. For the desktop, this means presenting a search results interface that is much less reliant on the “old school” ribbon-like stream, and that has more visual pop when filtering by rank, contemporaneity, media type, and so on. This also probably means getting rid of the unrelated Trends section in the search results – after all, the user is searching for a specific topic – and greatly minimizing the “Who to Follow” suggestions, which are an annoying distraction. The search results, including promoted Tweets, should be the primary visual focus.

The desktop search portal should also be more than just a minimalistic search box plus a text list of trends. Instead, it would be much more useful for the search box to be accompanied by a live tile interface that has pre-populated trending searches and results. Not only would this benefit users who don’t quite know where to start with their search, but it would also make users really feel that they are connecting to the pulse of the planet. Twitter’s mobile search, which shows trending searches plus short descriptions for each trend, is closer to what the desktop search needs to look like.

While one response to the above recommendations might be that most Twitter users use the mobile search interface, which does not have many of these issues, it is important to get the desktop experience right because:

  1. logged-out users probably don’t use the Twitter app; and

  2. the desktop experience will inform the bigger screen and augmented reality experiences for Twitter search.

Finally, Twitter’s search functionality has to take into account that, for logged-in users, the service is about more than just what is happening right now. Twitter is also a repository of user activity that has occurred in the past. Logged-in users must be able to easily search through their own Tweets and Tweets they have bookmarked in order to get full use of the service. Users should never have to jerry-rig a searchable archive by using third party services like IFTTT and Evernote.

In addition to enhancing its search functionality, Twitter needs to roll out an interface that allows users to more easily browse through Twitter content. OTTN suggests that Twitter add channels and trending/popular content to a Discover or Explore tab that both logged-in and logged-out users can use.

This functionality would be modeled to some degree after YouTube’s logged-out interface. (It’s no accident, by the way, that many of YouTube’s channels are named using hashtags.) Channels would include pre-defined categories or topics – Music, Sports, Comedy, News, etc. – as well as Live and Local content. There would also be separate buttons to bring up trending and popular content.

Twitter is already doing this with Vine. The Vine search experience is branded as the “Explore” tab. The Explore tab has a search box, “Popular Now” and “On the Rise” buttons, channels, and trending tags. Presumably, traditional Twitter will look more like Vine after a few more iterations of the product; this would be a great thing.

A recent report also suggest that Twitter is working on a new event-centric browsing experience code-named Project Lightning. Project Lightning will provide users, whether logged-in or logged-out, a human-curated set of content centered around specific live events occurring at the moment. The experience will be immersive, taking over the full screen, and will even be available for syndication off Twitter (on the web and in apps). Covered events might be pre-scheduled, such as awards shows or sports, or they might be breaking events/moments.

This is fantastic news, and in OTTN’s opinion, Project Lightning can’t come soon enough. Project Lightning is demonstrative of what Twitter has always had the potential to be – a place where everyone can connect to the pulse of the planet.

Extra, Extra – Read All About It!

It’s no secret that a substantial number of active Twitter users use the service primarily for the curated consumption of information, and not as a vehicle for sharing their own ideas and impressions of what is going on in the world. For these users, Twitter and “the news” go hand in hand, with news consumption facilitated through direct following of news outlets or via referral by other users.

Yet Twitter has done surprisingly little to become the predominant consumption vehicle for users who gets their news and analysis from multiple online sources. And this despite having one of the biggest advantages amongst all the social networks with respect to news distribution: a user community that engages based on interest.

It’s not as if Twitter hasn’t thought about its place as a news distribution service. It’s just that Twitter seems to be lagging behind competitors such as Facebook and Snapchat in building out its capabilities. Here are several ideas for how Twitter could become the best at news distribution, which would not only benefit existing Twitter users, but also potentially draw a substantial number of new users onto the Twitter platform.

Following Facebook: Instant Articles

The most obvious recommendation to be made is for Twitter to implement functionality similar to Facebook’s new Instant Articles product, which has caused so much consternation among the media old guard.

Here’s how Facebook describes the product:

Instant Articles makes the reading experience as much as ten times faster than standard mobile web articles. Along with a faster experience, Instant Articles introduces a suite of interactive features that allow publishers to bring their stories to life in new ways. Zoom in and explore high-resolution photos by tilting your phone. Watch auto-play videos come alive as you scroll through stories. Explore interactive maps, listen to audio captions, and even like and comment on individual parts of an article in-line.

It does appear that Instant Articles will make the news reading experience substantially better on Facebook. However, publishers are right to be wary of Facebook’s increasing clout in the news distribution space and its ability to tailor users’ News Feeds to favor whichever articles and publications will create the most engagement.

OTTN posits that a similar Twitter product would not only be great for users, but would almost certainly be a net positive for publishers, large and small. The big difference for publishers between the Facebook Instant Articles product and OTTN’s proposed Twitter version stems from the fact that Twitter is not as reliant as Facebook on a single news feed that must be algorithmically curated.

While Facebook is a social graph-based experience, Twitter is an interest-based experience, with logged-in users self-curating their Twitter stream, by and large. Although algorithmic curation will certainly be a part of what Twitter offers going forward, presumably users will always have the option to focus on the stream that they themselves have created. This user-generated stream includes Tweets from publications they have specifically decided to follow, as well as from users who share the same interests and, therefore, share relevant info and news. And Twitter users can also create multiple streams (i.e., lists) based on interest. Thus, publishers do not have to be as worried about Twitter having undue influence on what content users get to see.

Furthermore, publishers can benefit greatly from Twitter’s interest graph, which should be far more powerful than Facebook’s social graph when it comes to serving up relevant ads. For publishers who are not that savvy at serving up their own ads (especially on mobile), a Twitter Instant Articles-like product could provide a far more effective way to monetize their content.

Testing the Waters with Micropayments

One of the more annoying things on Twitter is to click on a shared article link only to find that the article is behind a paywall or login. This annoyance could actually spell opportunity for Twitter, as the company may have as good a shot as anyone to make micropayments a viable solution for online news. Specifically, OTTN could see Twitter cards + a pre-paid wallet of micropayment credits being a decent solution for breaking through paywalls for high value content.

Twitter’s interest-based nature is the reason why the service could succeed with micropayments where others have failed. Critically, many if not most links to paywall content on Twitter have associated data that lets users better determine whether or not the linked article has value.

The default actions on Twitter – tweeting, retweeting, and favoriting – all demonstrate the value of linked/embedded news content, particularly when made by trusted accounts. Users also add data to Tweets that help followers determine the value of an article: “must read” labels, quotes, infographics, and even paragraph screenshots. And if a conversation occurs surrounding the article, that’s even better for determining whether or not to pay up to read the article.

While there is plenty of understandable criticism surrounding the idea of micropayments for the news, it might make sense for Twitter to at least experiment in this area. Because if Twitter can get micropayments right, then the company truly could become an indispensable part of the news ecosystem.

Replacing RSS

Since the early days, it appears that Twitter has had an aversion to becoming a replacement for RSS feed readers. This certainly made sense when Twitter was a younger company, given that the product’s most unique quality was its real-time nature. It was determined, rightly, that breaking news and facilitating conversations around that news was what Twitter would be best at.

But focusing only on real-time news has kept Twitter from acquiring a large number of users that read news stories and analyses that are not time sensitive in nature. After all, to get a pulse on the world, one does not necessarily have to get information in real time – sometimes only timeliness is needed. Indeed, many online news readers (including your author) use products like Flipboard and Feedly for getting their news.

A product like Feedly, for example, serves as a self-curated inbox for news that helps the user compartmentalize and avoid information overload. The user creates different collections of feeds for each interest he has, and he can easily keep track of whether or not he has read through a given article in a collection.

Twitter should leverage its strength in news distribution to capture these RSS users. Expanding its reach in this way will allow the company to enhance the online news reading experience by getting rid of duplicative services, and by making it much easier to start conversations and share information with others who share the same interests.

However, it is important to recognize that Twitter the product is not currently designed to be able to replace the RSS feed reader user experience. So Twitter the company would have to either enhance the current design of the traditional service or create a new application and service that integrates with traditional Twitter.

There is some indication that Twitter is in agreement with OTTN on this front. Recent reports have suggested that Twitter has been in talks to acquire Flipboard. While buying Flipboard or a similar app could be a good way to help replace the RSS feed reader experience, reading online news is such an important daily habit for Internet users that it is almost inevitable that the major OS providers will offer their own news applications.

Indeed, Apple just announced a new News app that takes aim both at Flipboard and Facebook. As with Facebook’s Instant Articles product, publishers will keep 100 percent of the ads they themselves sell within the News app, but Apple will take a revenue share for unsold ad inventory that Apple helps fill.

Presumably, Google will be working on functionality very similar to Apple’s new app. It would make a lot of sense for Twitter to work closely with Apple, Google, and other OS providers to integrate its service with those news applications and perhaps receive a cut of the ad proceeds.

Publisher Integrations for Personalized Off-Twitter Experiences

In addition to enhancing the on-Twitter news experience, there is an opportunity for Twitter to enhance the off-property news experience. In fact, Twitter could be the perfect partner for generating personalized experiences on news outlets’ branded properties.

Imagine a Twitter user who also happens to be an Economist subscriber. The Economist produces a huge amount of quality content that gets incorporated into the weekly magazine. But it can be daunting, not to mention mentally exhausting, to get to the Economist at the end of the week, decide which articles to read, and then finally read through them.

What if, during the week, an Economist subscriber could use the publication’s Twitter feed to preview and select which articles would be presented at the end of the week in the Economist app? The subscriber might decide to read articles as they become available, or wait to see which articles other users retweet, favorite, and discuss, and then queue some of those articles up for a personalized Economist experience.

Giving users the ability to easily create a personalized, yet branded experience would increase publishers’ ability to monetize their content by serving more targeted ads to users. This would be a win-win both for Twitter users and publishers – and for Twitter the company, of course.

Accelerating the Apps Build-Out

Twitter has articulated a strategy of releasing new applications and services that will increase Twitter’s utility around the world. This strategy is key to fulfilling the company’s goal to reach at least a billion users both off- and on-Twitter and its promise to Wall Street to generate revenue from that reach.

The company’s current focus for new applications appears to be on content creation and sharing tools that will improve the quality and breadth of on-Twitter content, and increase the syndication (i.e., off-Twitter distribution) potential of Twitter-controlled content. Both Vine and Periscope are examples of this approach (though both are content consumption vehicles as well).

Releasing better content creation apps makes all the sense in the world. After all, strengthening the core service’s content, and expanding Twitter’s content reach across the web, mobile, and TV, ought to lead to greater monetization potential for its ad business. The release of such apps also dovetails nicely with Twitter’s mission statement.

However, a focus on content creation is not enough. It is pretty well known that, when it comes to online activity, the vast majority of content creation – including conversations surrounding primary content – comes from a small minority of users. The 1% rule lives, even in a world that is more socially networked and mobile. Given this reality, releasing new content creation apps will not be enough to maintain and grow Twitter’s logged-in user base.

Growing the logged-in user base is vital because logged-in users are significantly more valuable than logged-out users from a monetization perspective (except perhaps in the context of search). But, more importantly, growing the number of logged-in users is part and parcel of achieving Twitter’s informal goal of creating a platform that connects everyone to their world.

That’s because, in order for Twitter to really connect individuals to their world, using Twitter services has to be a personalized experience. Moreover, the more logged-in users that connect to Twitter services, the more comprehensive the world actually becomes for all users. Indeed, reaching as many logged-in users as possible is a key incentive for getting content creators to create and distribute content on Twitter. In other words, network effects matter, and Twitter could be the ultimate multi-sided platform, but only if the consumer end user connection is a strong one.

The fact that most logged-in users would be consuming, rather than creating, content does not decrease the importance of having as many logged-in users as possible. In a world where substantial metadata is created and collected, the curated consumption of content itself creates valuable content about what’s going on in the world.

Given the importance of growing the logged-in user base, Twitter should attempt to mitigate certain characteristics of the core service that limit the size of that user base. This will require building or tightly integrating with apps that do not necessarily rely on what the company views as the platform’s key point of differentiation: its being “simultaneously public, real-time, conversational and widely distributed.”

OTTN is well aware that this is a substantial deviation from Twitter’s strong tenet that the “four pillars” should be behind every Twitter product. However, if the company truly wants to become an “information, sharing and distribution platform,” deviating from the “four pillars” mantra is unavoidable.

This is not to say that Twitter should be trying to do anything and everything to connect people to their world. But when it comes to interest-based connections or searches where the timeliness and trendiness of content matters, the company should try to get involved.

Below are some areas of focus that OTTN would recommend.

Content Consumption-Focused Apps

Twitter should be creating more new apps for logged-in users that are primarily about consuming specific types of Twitter content. The company should also be partnering with third party app developers to create logged-in, consumption-focused experiences that combine/co-mingle these developers’ own unique content with Twitter content either by:

  1. pulling specific types of Twitter data using a special API; or

  2. requiring users to create or maintain connections through the Twitter service.

The reason why the company needs to release separate consumption-focused apps is because traditional Twitter is set up to be real-time, conversational, and public – hence its initially being sold as the “global town square.” Consequently, Twitter’s design cannot take into account the various ways in which users consume different types of content.

Time, place, and form matters when it comes to content consumption. Sometimes, real-time in the moment consumption is neither desirable nor feasible. Instead, timeliness or trendiness, combined with ease of digestion, is what matters. A user’s content consumption also may not be intended to be in public. And the form in which content is displayed is often the controlling factor for whether or not a user will bother to consume the content.

There are many people who do not use Twitter because the type of content they are interested in – content that is available on Twitter – is not optimally consumed through the core service. This is because of the time, place, and form issues that arise from the very nature of the core service. That’s why the company must build new consumption-focused apps or partner with third parties to create new consumption-focused experiences.

Vine is a great example of a Twitter app that is optimized for consuming a certain type of content: the short video. The Vine app is designed to be less about real-time viewing and conversation than about flipping through trending short videos and sharing and commenting on them. Finding Vines by category is also much easier than finding Tweets by category. And sharing of Vines can either be on public or private social networks.

We’ve already described another consumption-focused app that would be useful for current and prospective Twitter users: a news app that replaces the RSS feed reader. Such an app could, in theory, be built by Twitter. But given the OS providers’ interest in making news apps, it might make sense instead for Twitter to integrate very tightly with those news apps. For example, Twitter could work with Apple to make it possible to send articles that a user discovers on Twitter to the News app for later reading.

Here are some other ideas for consumption-focused apps that Twitter itself could release:

  • Local Events and Meetups:   An app for logged-in users to privately browse, search, and sign up for public local events and meetups. “Event cards” would be created by Twitter users – individuals, organizations, event venues, etc. – using their own websites or a special Twitter website, and these cards would become live in the new app once posted as public Tweets. Events could be shared by app users to their public and private social networks. App users could also follow event creators, either publicly or privately. Such an app could become a vital event discovery and notification tool, especially if integrated with users’ calendar tools of choice. And Twitter could go a step further and integrate Buy buttons into cards for events requiring tickets.

  • Local Commerce:   An app dedicated to connecting logged-in users with local businesses. Many, if not most, Twitter users do not follow local businesses because they don’t want their timelines clogged up with too much marketing material, and they’re not typically conversing with businesses. Even those who do tend to follow businesses only have a limited number that they will follow. OTTN proposes a new app that allows users to search for local businesses, follow them either privately or publicly, and receive Tweets from those businesses filed into different categories based on hashtag: #Coupon, #Special, #Announcement, #Event, etc. Users could set app notifications by filtering based on Tweet-type or white-listed account. Users could also communicate privately with the local businesses they follow through direct messaging.

  • Public Notifications:   An app acting as an inbox for all Tweets from certified public officials and institutions. Logged-in users of the app could set notifications based on the types of Tweets that come in. For example, a user could follow all certified public institutions for his city and set notifications to appear whenever those accounts tweet out public emergency warnings, road closures, election notices, etc. This app would essentially allow logged-in users to connect to the pulse of their governments.

As for third party consumption-focused apps, there are at least two types of situations where Twitter should consider working closely with – or even partnering with – third party developers, rather than simply letting them incorporate Twitter content through the normal APIs and SDK. Both involve personalized, consumption-focused experiences that require logged-in usage of Twitter services and that combine or co-mingle unique third party data with Twitter data.

First, some third party app providers may want to provide personalized, consumption-focused experiences that use Twitter data that is not easily pulled using the normal public Search and Streaming APIs. It might make sense to give these developers special access to Twitter data if their own content would allow them to run unique, personalized searches on behalf of logged-in users. Essentially, these third party apps would be using their insights about users to do authenticated Twitter searches on their behalf and provide the results for consumption along with their own content. Partnering with such third parties would help the company to better understand its logged-in users’ interests and provide more venues for targeted ad monetization.

Second, some third party app providers may be able to create new consumption-focused experiences that rely on logged-in users creating or maintaining Twitter connections. Such apps would pull Twitter content from a logged-in user’s existing list of followed accounts or from a curated list of accounts for the user to follow – or both. If these third party providers also have unique content that could draw new users into the Twitter ecosystem, Twitter should consider partnering with them.

An example of a third party app that could fit both types of situations is a stand up comedy app that has unique content from comedians, as well as information about the types of comedy its users enjoy. If the comedy app provider were a Twitter partner, it could deliver user-personalized comedic content from Twitter in the form of Tweets, Vines, and Periscopes to supplement its original content. That personalized content could be pulled either through a special Twitter search or based on the comedians that the user follows on Twitter. And then Twitter could help that comedy app provider monetize its content through use of the MoPub platform or Twitter’s own branded advertising.

Ultimately, working closely with third party developers on consumption-focused apps will allow Twitter to enhance the utility of its services far more quickly than the company can by itself.

Apps for Interest-Based Groups

Public and private virtual spaces for people to communicate about, and organize around, their shared interests have flourished since the creation of the Internet. In the past, such spaces have included electronic mailing lists, usenet newsgroups, IRC channels, web forums/message boards, etc. Newer spaces include “social” websites and apps that are either purpose-specific (e.g., Imgur) or more general purpose (e.g., Reddit). Even ubiquitous social networking services like Facebook have gotten into the game by providing group functionality in their main apps and in separate apps.

Given the multitude of interest-based community spaces that already exist, and the fact that Twitter is already connected to many of them with Twitter Kit and the ubiquitous Tweet button, one might have the view that Twitter should not be trying to reinvent the wheel. However, there at least two types of interest-based group apps that Twitter could release that might add value and increase Twitter’s logged-in user base.

First, Twitter might consider creating an organization-focused app with an emphasis on events and meetups. Unlike the consumption-oriented events and meetups app described in the previous section, which would derive its data from public Tweets, the app envisioned here would allow for either public or private organization around shared interests by like-minded people. Essentially, Twitter would create an app that would compete with Facebook Groups and Meetup. However, the app would be completely separate from Twitter, and users would log in using Twitter’s Digits product, not using a Twitter account – though they could, of course, link a Twitter account if desired.

Like Facebook Groups and Meetup, the app would allow group members to create events, have group discussions, direct message other members, conduct polls, upload files and photos, collect membership dues, etc. For public groups, a public Twitter account would automatically be created, and events, posts, and polls could easily be shared to Twitter by the group admins. Easy Tweet (and other) embeds in discussions and direct messages would be a core part of the app’s functionality.

Unlike Facebook or Meetup, the app would keep a user’s groups and interests completely private by default. Users could also create entirely different profiles (or personas) per group, choosing which information they wish to disclose – pictures, contact info, location information, etc. Pseudonymity would be much less problematic because, if a group admin thought that a member was being a troublemaker, the member could essentially be kicked out forever, unless he could somehow get use of another mobile phone number. Still, the groups most likely to benefit from the app would be those that remain small and require real identification for membership.

There would be obvious benefits to Twitter if it could gain traction with a group app like the one described above. First, it would jump start the adoption of Digits. Second, it would likely increase the number of logged-in users for the core service. Third, it would create more content for the core service and increase the viewership of Twitter content.

Whether Twitter could pull off a decent adoption rate for such an app is unclear. The main reason why Facebook has sway with interest-based groups is because it lowers the transaction costs of joining and interacting with groups. Facebook is much more ubiquitous than Twitter, and people who check Facebook fairly regularly can easily find, share, and check up on groups. So Facebook has an advantage in this respect. Nevertheless, a successful effort by Twitter could increase its ubiquity and, therefore, might be worth pursuing.

The second type of app that Twitter might consider releasing is one that allows interest-based groups to easily create public discussion spaces focused on long-form thought and organized by topic. Essentially, Twitter would update the concept of the public message board/web forum, making it easy for groups with a presence on the web and mobile to create discussion forums that are accessible cross platform.

What OTTN envisions is an app that works like a hybrid of Medium and Slack, the enterprise messaging app that has grown to over a million daily active users at the time of this writing. App users would be able to create public groups, similar to Slack teams, that could be focused on anything under the sun.

A group would have multiple admin-created topic categories similar to Slack channels, but that would further expand to reveal threads created by group members. Members would be able to post to threads in a way geared towards long-form writing, with the ability to add rich formatting and notes, and attach or embed documents, pictures, videos, and Tweets, among other things. (In this regard, the app would be more like Medium and not like Slack, which is geared towards a simple messaging/chat format.) Every post would incorporate a Tweet button, as well as heart and bookmark buttons. Lurkers could view group content, but only admitted members would be able to post, heart, and bookmark.

Each group space would be accessible on the web, either at a Twitter-provided sub-domain, or at the group’s own domain. Spaces would be theme-able so that groups with an online presence can keep the look and feel of their spaces consistent. On mobile, group spaces would be accessible on an app released by Twitter. Authentication would be handled using Twitter’s Digits product, with the ability for app users to create different personas for each group they actively participate in.

Once again, the benefits to Twitter of creating the above-described app/service are fairly obvious. Whether Twitter could gain traction is another question. It is also entirely possibly that companies like Slack are already working on this type of app. If so, then it would make a lot of sense for Twitter to closely partner with such companies in order to tightly integrate their products with the Twitter core service.

Both of the types of group apps described above would advance Twitter’s goal of connecting its logged-in users to their world, as well as provide tools that would be superior in many ways to existing options.

Real-Time Infrastructure Services

The fact that Twitter is as reliable as it is across the globe is nothing short of amazing. A 2013 blog post from Twitter’s engineering team demonstrates how hard it is to make things work in real-time at global scale. Thus, it is no surprise that the company lists its ability to “develop a reliable, scalable, secure, high-performance technology infrastructure that can efficiently handle increased usage globally” as a major risk factor in its public filings.

As Twitter continues to fortify and build out its infrastructure, the company should consider opening it up to third parties in a way that doesn’t require the creation of Tweets. Specifically, Twitter should create a real-time infrastructure service that allows third party software developers and hardware manufacturers to stream, store, and sync data in real-time across multiple devices (or connected things, really). Essentially, Twitter would create its own version of services like Firebase, which was recently acquired by Google, and PubNub.

OTTN is well aware that Twitter’s relationship with third party developers has been strained over the last few years. In the past, third parties have built apps and services using Twitter APIs that were subsequently undercut by the company, either through the purchase or release of competing apps and services, or through changes to its terms of use. Whether or not these actions by Twitter were unfair is debatable, as they generally reflected the inevitable conflict that tends to occur between platform providers and developers. They were also the result of the company trying to figure out a business model, which depending on your viewpoint, was a compelling reason for it to do what it did.

Still, Twitter recognizes the importance of third party developers, the mistrust among them that it has to counter, and the necessity of creating win-win solutions for monetization. The release of the Fabric mobile development platform last year was a great step in that direction that hopefully will encourage developers to come back to Twitter. And with the purchase of MoPub, Twitter has finally developed a game plan for monetizing off Twitter that should also benefit its developer partners, as MoPub is essentially Twitter’s AdSense.

OTTN believes that third party developers would flock to a Firebase-like service offered by Twitter because this would also represent a win-win solution for Twitter and developers. With the envisioned service, the data generated by software developers and hardware manufacturers would be kept separately from Twitter-controlled data. This would be unlike Twitter’s dormant Annotations proposal, which some have lamented would have been a boon for developers.

Twitter’s real-time infrastructure service would be monetized the traditional way – through payment for use. However, Twitter could also make it easy for anyone using its real-time infrastructure service to take the data they control (and that they are allowed to use, of course) and put it into the Twitter firehose for monetization. Essentially, GNIP, Twitter’s social data services arm, could become a reseller of real-time data (not necessarily “social” in nature) being generated by these third parties.

Twitter could even partner with other infrastructure service providers – for example, Amazon, IBM, or Microsoft – to offer its services in conjunction with their own. For example, Twitter could make it easy for a customer to store the data generated using its service with Amazon, or to dump the data generated into IBM’s Watson services for processing.

More and more data will be going across the Internet for real-time applications, and Twitter should participate in this trend by monetizing its technology infrastructure.

Up Periscope

It’s obviously too soon to tell how big Periscope, Twitter’s live video streaming app and service, might become. However, Periscope has many really interesting things going for it that could make it a key vehicle in revolutionizing how live video experiences are created and consumed.

First, it makes the creation of live video “broadcasts” incredibly easy; using the app is a piece of cake. Second, although Periscope presents a one to many experience, it also provides for real-time interaction and feedback – currently, in the form of messages and hearts – to the broadcaster. Third, because Periscope is an app on the phone, the most personal of devices, Periscope broadcasts have the potential to be more intimate, more unexpected or serendipitous, and less staged than typical live broadcasts. Fourth, Periscope is helping to popularize the notion of ephemeral video, at least on the consumption end, which could present an interesting value proposition in a world with unlimited video storage capability. Fifth, by being connected to Twitter, Periscope has a decent discovery and, possibly, curation mechanism built into the product.

All of those attributes make Periscope a product that has many potential uses. Periscope could become a vital tool for the delivery of breaking news. Key moments in people’s lives will be able to be easily broadcast to the world or to a select few who cannot be there in person to share the moment. Historical moments and events will now have a multitude of “cameramen” with different viewpoints on what’s going on. Live arts performances and sports events will be available in ways that they never were before.

And so on. The possibilities are endless, and OTTN is sure that Twitter and the Periscope team have huge plans for the app and service. Nevertheless, we will provide some input on what we think ought to be considered about Periscope going forward.

Discovery will be a huge part of the value added by Twitter to Periscope, and design changes may need to be made to enhance broadcast discoverability. For example, Periscope might consider requiring hashtags in addition to titles for broadcasts, with a default tag of #Periscope being added to each broadcast. It might also be worthwhile to come up with topic categories that can be added to broadcasts by viewers.

The availability of Periscope broadcasts on Twitter will also be a major value add to Twitter itself. Therefore, the company’s plans to enhance on-Twitter search and browsing must incorporate Periscope broadcast discovery in a big way. Twitter should also consider cutting out broadcast discovery from the search deals it does with third parties (such as Google) because it shouldn’t be giving away such a valuable advantage that could potentially turn Twitter into the ultimate real-time search engine. And when it comes to browsing for Periscope broadcasts, Twitter should consider making a desktop and TV experience in addition to the mobile experience. It is understandable if Periscope wants to use browsing as a sweetener for downloading the app in the user growth phase, but eventually logged-in and logged-out browsing of broadcasts on more than just mobile will have to be an option.

When it comes to lengthening the amount of time that Periscope broadcasts are available for replay, Periscope should tread very carefully. Part of the reason Periscope is special is because of its ephemeral nature. Viewers are seeing through the eyes of someone else right now or, with the 24-hour replays that are available, on a very short delay. Indeed, the fact that these broadcasts are fleeting may increase their value, and one of the promises of Periscope is to open up a long tail of live event experiences that can be experienced remotely.

But the minute videos become archived forever by default, one will have to ask whether Periscope has just become YouTube Capture with a live streaming component to it. That may not be such a bad thing, since not archiving videos by default could make it quite difficult to monetize Periscope (through ads shown at the beginning of videos, for example). But Periscope should proceed with caution and be hypersensitive to audience reactions indicating that the app and service is becoming less magical.

OTTN also believes that the Periscope brand should probably remain coupled to the mobile device. There might be a temptation to turn Periscope into a live broadcast service that takes in video from more than just phones – for example, from high-def camcorders, webcams, etc. But Periscope probably shouldn’t go that route. Instead, to the extent that Twitter wishes to directly facilitate live broadcasts from non-mobile devices – as opposed to just being a vehicle for sharing and discovering them – it should do so using the Twitter property or some other property that would be like Twitter’s version of YouTube. Periscope should remain a window into the world through someone else’s eyes, and that means keeping the property associated with the mobile device.

Finally, Periscope will need to take an active role in protecting content rights, particularly with respect to key Twitter partners. This will require developing well-designed tools for content owners to police their rights and partnerships with parties that Twitter may not have worked with in the past, which could include Internet service providers, venues, and artists/performers. Taking this active role may provide Periscope a moat of sorts by giving it a vital role in the supply chain for live content.

For example, imagine a Golden State Warriors game being played at the new arena being planned for them in San Francisco. That arena will likely have free Wi-Fi provided to spectators by a big Internet service provider (ISP). Periscope could partner with the NBA, the venue, and the ISP to make sure that only certain types of live streaming content get out of the arena – and that they only get out via Periscope.

Or imagine a weeknight at a local music venue where several artists are performing their sets in front of a crowd. While the size of the show’s audience would normally be limited by the capacity of the venue and the crowd turnout, with Periscope those performances could be live broadcasted on the fly by viewers who have friends who might be interested in watching. What if Periscope gave venues and artists the ability to register a location and event to make sure that any videos live broadcasted out to the world were monetized in some way – either through a watch fee or through commercials? Periscope could also help syndicate that performance out to other video distribution channels with it getting a cut of the ad sharing fees.

The bottom line is that Periscope could be a very big business opportunity for Twitter, and hopefully the company is thinking about it the right way.

Miscellaneous Monetization

Below are some miscellaneous ideas on how Twitter can increase monetization.

  • Selling Financial Analysis in Real-Time:   For its direct response ad program, Twitter should consider focusing on high value content that is worth purchasing in real-time. The Twitter Finance community might be one to target for direct response ads, as time is often of the essence when it comes to taking action. Twitter could capitalize on this by being a distributor for paid financial analysis from reputable sources.

  • Twitter Tix:   Twitter should put together a special commerce program focused exclusively on selling tickets and that tries to reach as many ticket sellers as possible. It appears that the majority of Buy Now experiences for tickets are from big partners like the NBA. But there is an entire long tail of tickets that could be sold through Twitter by venues, artists, and event sponsors. In fact, the long tail of tickets could ultimately be where Twitter has a key advantage relative to bigger players such as Ticketmaster and Live Nation. Twitter could make it easy for people in the long tail to sell tickets either directly or through Promoted Tweets.

  • The Human Sensor Network and the Firehose:   Twitter has the infrastructure to become very big in the crowd-sourcing business, which has proven to be viable with the growth of mobile devices. For example, companies like Waze have adeptly used crowd-sourced data to create services of great value. Twitter should come up with creative ways to provide standardized real-time input from its users to data consumers, essentially offering value added services on top of the firehose. For example, in the local government realm, Twitter could make it trivial for citizens to report repairs to a city by providing a standardized channel between logged-in users and government. Adding this type of functionality to Twitter could turn it into the ultimate human sensor network.

The Privacy Protective Social Conglomerate?

Finally, OTTN would suggest that Twitter has an opportunity to become a “social conglomerate” that is far more protective of the individual privacy and dignity of its users than its larger competitors (i.e., Facebook and Google). As Apple is showing, paying attention to user privacy can be made into a competitive advantage, and Twitter should strongly consider taking the same route.

The core Twitter service has always been less problematic for privacy than other social services. It is well known by those who use Twitter that it is public in nature. After all, Twitter was originally known as the “global town square.” Therefore, users’ expectations of privacy on Twitter have been much more realistic when compared to, say, their expectations of privacy on Facebook.

At the same time, Twitter has tried harder than many other social networking services to make sure that users are informed, and given more choice, about what information is being gathered and revealed through the service. Twitter’s privacy settings are much easier to navigate and customize than with other services. User opt-in is standard when it comes to adding location data to Tweets, getting direct messages from the public, and sharing contact info through Digits for “friend finding.” Other privacy-related options are not opt-in but can easily be opted out of through Twitter’s fairly straightforward settings section.

Do Not Track is supported by Twitter. Pseudonymous participation has always been allowed on Twitter. The company has been fairly assertive when it comes to protecting user data from arbitrary government requests and fighting for the right to disclose when government requests have been made. Even where the company has failed to protect the privacy/dignity of its users – for example, with respect to the harassment of female users – the company has admitted its failures and tried to change its policies accordingly.

Twitter clearly has “privacy cred,” but it could do a lot more. For example, it could introduce filtering options for users to only allow public messaging on Twitter and Periscope when there is a verified identity behind an account. This would lessen the amount of pseudonymous harassment on the service. When it comes to Twitter’s interest graph, the company could be even more upfront with its users about how their information is used and what limits are or are not in place to keep third parties from using this data. It could add an “ephemeral location” option to its services, so that users get relevant location-based information and offers but do not have location data persistently connected to their accounts and Tweets.

For MoPub, Twitter’s mobile advertising platform, the company could be a lot clearer to users and the public about how this product will be integrated into its other offerings, including the Twitter firehose. For example, if the company ever gets into app indexing and user activity annotation, as Google is doing (not to suggest that Twitter ought to do this), then it should make abundantly clear to the public when apps are utilizing this technique to send information about their users to Twitter. This should especially be the case now that the company has rolled out its Digits identity system, which makes Twitter more and more like Google and Facebook than ever before in terms of its potential reach across the web and mobile.

Twitter’s past as an organization that appears to care about human values such as freedom of expression, privacy, and dignity is what makes OTTN optimistic (hopefully, not irrationally so) about Twitter’s potential to be the privacy protective social conglomerate. Indeed, that is probably the key reason why we wrote this piece and why we want to see Twitter thrive.

We shall see how things turn out.