This report tracks the evolution of content formats on social media and the greater internet. It considers how the concurrent development of more powerful computers and increasingly expansive, faster networks allowed photos to overtake text, then video to overtake photos as the primary format for social media interactions. The report traces how these developments were facilitated by the transition from dial-up to broadband, and the miniaturization of computing via the development of smartphones. It concludes with a brief description of pioneering augmented and virtual reality platforms, to hypothesize that, as computers and networks get even faster, content across the internet will gravitate to these even more immersive formats. This forecasts the future of social media networking, and potentially, the impact it will have on business and campaign planning.
As computers have become more powerful with the expansion of data infrastructure, more immersive media formats have dominated the Internet. Social media, much like the greater web, remained a text-heavy environment from its inception 1997, through 2007. After 2007, social media platforms began to facilitate and encourage the use of images over text. Then, as the speed of computers and networks got even faster, platforms like Instagram and Snapchat, began to emphasize video content; this trend took off in 2015. Today we’re seeing technology companies making sizable investments in augmented reality and the metaverse. In short, the development of modern day social media can be divided into the text, photo, and video eras with the focus on the metaverse representing the next chapter.
In this report, we trace the evolution of these increasingly complex content formats, and present the associated software engineering developments that supported and facilitated these evolutions. The purpose of the inquiry was to understand the trajectory of social media development – in content and technology – so as to forecast and gain an understanding of the future of social media. Understanding this trajectory presents great potential for planning the use and integrations of social media for positive business outcomes.
To conduct this inquiry, we use social media analytics from Infegy Atlas, along with numerous other archival sources. We organize our findings chronologically into four sections:
As computers were programmed to communicate with each other over networks, those conversations took place in text format. This was because they had neither the graphical processing power nor the storage to hold high-resolution images or videos. Text, on the other hand, is one of the lightest forms of data, making it a perfect medium for early computers.
However, as early as 1965, Gordon Moore, the co-founder of Intel predicted that the number of transistors that could fit on a chip would double every two years, essentially forecasting the rapid development of computing technology over the next 40 years (Kelleher, 2022). The growth of computing storage serves as a corollary to Moore’s Law, where hard drive storage has indeed grown exponentially over the last several decades (Hutchinson, 2011). Computing power circa 1997 was sufficient solely for the light data formatting of text, but rapidly developed to be able to support images.
On the network side, internet users used dial-up as the predominant network infrastructure. This interface could not handle the size requirements of more complex media like images. A 2007 Communication Workers of America Survey put the average US download speed at only 1.7 megabytes per second, fast enough for text-based websites, but not fast enough for heavier formats like photos or video (Cheng, 2007). During this era, American Online (AOL), a dial-up provider, had a virtual monopoly on the US internet market. AOL, however, did not anticipate the adoption of broadband as the primary vehicle for networking, and soon became obsolete. Infegy Atlas data depicts a decline in AOL’s influence between 2007-2008, measured by a reduction in AOL-related post volume. This post-volume continued to decline by 90% over the next 15 years (Figure 1, Infegy Atlas - AOL, 2022).
Figure 1: AOL Post Volume (2007-2022); Infegy Atlas data.
Top websites of the text-era reflected these network and computer hardware constraints. Professor Anthony Cocciolo of the Pratt Institute of Information Science conducted a survey using six hundred historical snapshots of the top 100 websites in the United States. He found that text-use on the internet peaked around 2005 through 2007, forming a median 31-33% of the overall page, only to drop drastically right after 2007 (Cocciolo, 2015). This becomes the threshold to image-centered social networking.
Early examples of social media websites reflect this text-first design. When Facebook launched, it allowed only one profile picture per account in 2004 (Shontell, 2014). This continued for the first few iterations of the site. However, as the site continued to develop, picture-based content began to replace text. The shift towards images began in 2005 as Facebook provided users with unlimited photo storage – it was the first social networking site to do so (Kanalley, 2011). Facebook was not only the first to integrate images into social networking, it did so well in advance of other SNSs. For example, Twitter only began supporting embedded pictures in 2010, five years after Facebook’s initial adoption (Barbato, 2014). Infegy Atlas shows a spike in post volume that corresponds to the expansion of features: we can see an estimated 1.7 million posts across all collected channels at its peak, with a 12% increase in Anticipation in the months leading up to the feature release (Figure 2, Infegy Atlas - Twitter Photo Anticipation, 2022).
Figure 2: Twitter Photos Anticipation (2007-2022); Infegy Atlas data.
The rise of the smartphone and the spread of fast, mobile networks sparked the explosion of the image-based internet that would dominate social media for the next eight years. These two factors worked together to make social media’s era of images. Smartphones gave millions of users a network-connected camera to produce and view images. Simultaneously, mobile networks and home broadband provided the data pipelines to share and download those images. These two factors combined, caused the exponential growth of images across social media.
In 2007, Apple released the iPhone, the world’s first modern smartphone. Its reception was lukewarm. Infegy Atlas captured the sentiment around this moment in history: both post-volume and sentiment were low and, as users complained about the iPhone’s exclusive functioning on AT&T networks, and its use of proprietary software (Figure 3 and 4, Infegy Atlas - iPhone Release Post Volume and iPhone Release Sentiment, 2022). Despite this, users admired its ability to connect to networks and browse the internet. The iPhone offered the first modern, touch-based, mobile-browsing experience. Android phones followed shortly after, the same year.
Figure 3: iPhone Post Volume (2007-2022); Infegy Atlas data.
Figure 4: iPhone Sentiment (2007-2022); Infegy Atlas data.
This explosion of smartphones meant an internet-connected camera in everyone’s pocket. Shipments of stand-alone digital camera devices fell from 120 million units at their peak in 2010 to 9 million units in 2020 (Guinness, 2021). While digital cameras dropped precipitously, Pew Research notes that American smartphone ownership grew from 35% of adults in 2011 to 70% of adults in 2015 (Pew Research Center, 2021). In the meantime, cameras in smartphones have become increasingly complex and powerful. The original iPhone launched with a 2 megapixel shooter that could only capture images, not video. By the end of 2015 – essentially the end of the social media photo era – Apple’s iPhone 6s featured a 12 megapixel camera that could shoot 4K video and was highly effective in lowlight conditions (Costello, 2019).
Small cameras embedded in modern mobile phones could not support the image-revolution on the internet without the concurrent expansion of fast broadband and the development of mobile internet. Moreover, this provision was made available to the common consumer.
Pew Research notes that January 2007 was the first time more than half of America had access to broadband internet via satellite or cable. That grew to approximately 70% by the end of 2015 (Pew Research Center, 2019). Mobile data technology and its access has grown even more impressively than broadband. In 2012, mobile internet traffic only represented 10.88% of global internet traffic (Desktop vs Mobile Market Share Worldwide, 2022). By 2015, that had risen to 37.14% and would rise even higher, to 54.09%, in the next two years. Mobile data speeds progressively increased during this time as well. Infegy Atlas, tracking mentions of different wireless technologies on social media, charts how the 2010s represent the move from 2G and 3G bands to 4G to finally 5G bands (Figure 5, Infegy Atlas - Wireless Post Volume, 2022). These massive data pipes allowed billions of photos to flow across networks from users’ new smartphones.
Figure 5: Wireless Internet Post Volume (2008-2022); Infegy Atlas data.
Encouraged by the desire of social media users to share images, and facilitated by the new phone and network hardware, a plethora of mobile apps designed to edit and share these photos began to emerge. The initial mobile apps specialized in photos. Instagram, founded in 2010 and acquired by Facebook in 2012, is the best known. Snapchat was founded a year later in 2011. It avoided an acquisition and, as of 2022, remains an independent company.
As predicted by Moore's Law, computer chips had developed in leaps and bounds, and image-focused apps took advantage of the massive advancements in computing power. Prior to mobile apps, users had to have a dedicated digital camera device on hand, take pictures, load them onto a computer, edit them, then post them. Smartphones and the cameras they powered streamlined this process to a few taps and a matter of moments. Users could now edit photos on the fly before posting them. 1.8 billion photos were uploaded daily to the internet in 2014 (Knibbs, 2014). A 2014 visual report from eMarketer, a marketing research firm, notes that 75% out of a 1.2 million post-sample involved photos, while 4% already incorporated videos (Maranga, 2014).
While users loved the transition from text to images, marketers and advertisers loved it even more. Visual content proved to garner more reactions. Moreover, it proved to be more engaging, and kept users on social media sites for longer. Writing for The Journal of Marketing Research, Professor Yiyi Li of the University of Delaware, contends that “the inclusion of an immediately viewable image in a tweet increases the number of retweets by 119.15% and the number of likes by 87.26% for tweets related to air traveling” (Li & Xie, 2019). Li goes on to explain that the engagement increases varied by content category, but were universally positive when images were included with tweets. Interestingly, this effect did not hold true for linked images, but only images directly embedded in the bottom of the post. The previously cited eMarketer report notes that sampled photos had an 87% interaction rate from fans (Maranga, 2014). Ultimately, the birth of social media marketing was one of the major outcomes of the transition from text to image on SNSs.
By 2015, phones and mobile network speeds were faster than ever. Despite the speed increases, Facebook and other social media platforms began noticing an engagement slowdown. Writing in The Information, reporter Amir Efrati writes that Facebook users slowed sharing original content by 5.5% from mid-2014 to mid-2015, continuing to a 15% year-over-year decline (Efrati, 2016). A late-2015 study by Locwise, a social media research firm, revealed that engagement on Instagram dropped by 15% across thousands of randomly sampled posts (King, 2015). In order to recapture user engagement (and advertising dollars), Mark Zuckerberg and other top social media executives announced a (now notorious) “pivot to video.”
Prior to Facebook’s pivot, several companies had tried and failed to sustain video-centric apps. A social media platform called Vine made the best-remembered attempt. Vine pioneered the short-format video where users would upload and watch 6-second videos in looped format. Vine launched in 2012, and was acquired by Twitter in October of the same year. Trend analysis on Infegy Atlas revealed that Vine’s popularity peaked in 2013-2014. In 2017, Twitter shut down Vine completely because marketing dollars migrated to other video-centric platforms.
Despite its short-lived existence online, Vine proved hugely influential to the modern, video-centered social media ecosystem, and actually launched the careers of many famous influencers like Jake and Logan Paul, David Dobrik, and Shawn Mendes. Infegy Atlas depicts that modern social media users are highly nostalgic about Vine, and depicts that it generated a higher average sentiment score after it was shut down, than it did while it was live (Figure 6, Infegy Atlas - Vine Sentiment, 2022).
Figure 6: Sentiment Around Vine (July 2012-July 2023); Infegy Atlas data.
Another short-lived video-platform was Beme, founded by Casey Neistat, the famous YouTube creator. Beme launched in 2015, and was designed to be an alternative to highly-edited social media content from Instagram or Vine. Beme users could only upload unedited videos to the platform to preserve an authentic feel to them. Like Vine, Beme failed to attract advertisers and fend off competing features like Instagram Stories (Glaze, 2021). Ultimately Beme was acquired by CNN in 2017 for $25 million.
In the midst of growing competition between video-centric social media platforms, Facebook CEO Mark Zuckerberg announced that his company would pivot to video, highlighting video content on news feeds over textual or photo-based posts. In 2015, Facebook designed several new features to highlight video on their platform like a “Suggested Video Interface,” news feeds that remain scrollable while playing videos, and a dedicated video feed (Constantine, 2015).
Facebook’s shift to video soon became very controversial. According to leaked Facebook internal engineering reports, Facebook data scientists estimated that the company oversold users’ video engagement by upwards of 80% (Kosoff, 2018). Acting on and publicizing these inaccurate metrics, Facebook continued pushing video-centric content, as well as the false pretense of video-engagement being higher than it was. This led to immense repercussions in digital media, especially journalism. The Atlantic Monthly calculated that at least 483 journalists lost jobs as newsrooms laid off traditional journalists to hire more video-editors and media producers (Meyer, 2018). Multiple publishers sued Facebook for publishing these misleading metrics (Oremus, 2018). Zuckerberg expressed “regret” for the way his platform handled the pivot during his 2018 US Congress testimony. Within digital journalism, the phrase “pivot-to-video” has become a derisive term for any poorly thought out strategic business pivot. Captured by Infegy Atlas, Facebook’s “pivot to video” has an overwhelming negative connotation on social media, with a 70-80% overall negativity rate (Figure 7, Infegy Atlas - Pivot to Video, 2022).
Figure 7: Sentiment Around Pivot to Video (July 2012-July 2022); Infegy Atlas data.
Facebook’s failure to effectively reorient the platform around video, created a vacuum – one that was quickly and efficiently filled by ByteDance, the Chinese technology company that founded TikTok in 2016. Since then, TikTok has dominated the mobile video space, acquiring an enormous amount of advertising dollars. Infegy Atlas data charts this meteoric rise, showing that TikTok post-volume has increased by 1244% over the last five years, while Facebook’s product offerings like Instagram and WhatsApp have declined by 9.5% during that same time frame (Figure 8, Infegy Atlas - The Rise of TikTok, 2022). British journalist Mark Sweney writes that TikTok’s advertising revenue in 2022 surpassed both that of Twitter and Snapchat combined (Sweney, 2022). Sweney explains that TikTok’s clever algorithm keeps the average TikTok user on the app for a mean of 19.6 hours per month, a four-fold increase from 2018 when they began recording data. This enormous growth is expected to continue over the next several years.
The idea for augmented and virtual reality (AR/VR) has been around longer than computers have existed. Ever since the development of photography in the 1800s, people have conceived of existing within a photographic world (History of Virtual Reality, 2016). However, AR/VR integration within social media is a relatively new concept.
Google launched Google Glass, a set of smart glasses, in 2014. This represented one of Silicon Valley’s first attempts at integrating an augmented world with the online social one. As part of the Glass OS, Google provided native access to Google+, Google’s now-defunct social network, and allowed a Twitter integration via a third-party app (Redhead, 2013). This app allowed users to read Tweets from their feed and dictate tweets to post while on the go.
Google Glass was ahead of its time, and did not achieve widespread adoption. First, it was too expensive, retailing initially at $1500. Second, non-users thought it was a violation of privacy due to the headset having an integrated camera and microphone, but no indicator to show when it was recording. This caused concerns, since people did not like being recorded without their knowledge or consent. Linguistic analysis on the conversation around Google Glass shows topics like “creepy,” “awkward,” and “privacy concerns” (Figure 9, Infegy Atlas - Google Glass, 2022). Third, people thought Google Glass was created for a niche, technology-focused audience. Even before its release, Wired Magazine categorized it with products like “Segways, pocket protectors, and Bluetooth headsets,” and labeled it as “too nerdy to live” (Wohlsen, 2013).
Figure 9: Google Glass negative topics (July 2012-July 2022); Infegy Atlas data.
Despite all these criticisms, Google Glass would influence other successful products like Snapchat’s Spectacles and Facebook/Meta’s RayBan Stories. Wary of the failure of Google Glass, but inspired by the potential of augmented and virtual realities, major social media and technology companies now approach AR and VR as the next frontier of technological innovation.
Facebook upended its entire structure and pivoted to building a virtual world, the Metaverse. They changed their entire brand to reflect this new foray into VR, rebranding as Meta in October 2021. This restructuring began back in 2014 with their acquisition of Oculus, a leading virtual reality company. Meta views these virtual worlds as the future of social networking. Meta’s Reality Labs, their main VR research arm, had a $10 billion annual budget in 2021 with plans to spend even more money in the future (Kastrenakes, 2021).
Despite this enormous investment, both the advertising and publishing industries remain skeptical. Bruised from Meta’s “pivot to video” debacle in 2018, The New Republic labeled the Metaverse launch as the “next pivot-to-video” (Shephard, 2022). Infegy Atlas data shows that conversation around the Metaverse began to peak close to its official announcement date but conversational volume dropped 50% in the early months of 2022 (Figure 10: Infegy Atlas - Metaverse, 2022). That being said, social listening also shows that the sentiment around the Metaverse announcement was 87% positive, suggesting that the larger internet community remains hopeful and anticipates the launch of Meta’s virtual world.
Figure 10: Metaverse Post Volume (August 2019-July 2022); Infegy Atlas data.
In contrast to Meta’s vision and investment around augmented reality, Apple has chosen to invest heavily in augmented reality, or providing an interactive virtual experience within the confines of the real world. Apple’s investment involves both hardware and software, with advanced hardware sensors allowing apps and features to take advantage of AR. Apple’s first foray was the Face ID feature which allowed users to unlock their phone by looking at invisible, thermal sensors. Apple used these same sensors to enable Animojis, or interactive, animated emojis that were mapped to the user’s facial expression. Additionally, Apple included LiDAR sensors (Light Detection and Ranging) on premium iPads and iPhones which enabled these devices to place virtual objects within physical spaces.
Apple operationalized many of those features with the upcoming Apple Vision Pro. The Apple Vision Pro is an upcoming standalone mixed-reality headset, set to launch in early 2024, integrating digital media with the real world through motion gestures, eye tracking, and voice input. Running on visionOS, derived from iOS, it features dual micro-OLED displays, Apple M2 and R1 CPUs, advanced tracking, and an outward-facing "EyeSight" display. Initial reception has been mixed due to its high price (They have priced it at $3,500, well above Google Glass), short battery life, and missing features, but it represents Apple's significant entry into immersive technology.
Figure 11: Apple Vision Pro release (August 2022 through August 2023); Infegy Atlas data.
Gone are the days of the text-heavy web that only had dial-up to back their transmission. The exponential growth of computing power and faster networking forged pathways to more complex formats on social media. This transition began with the growth of images on social media, and continued as platforms began to emphasize video over still images. Now, with the move towards augmented and virtual worlds, the future of social networks will be more immersive than ever. The next question is not whether social media will transition to the immersive worlds of virtual and augmented reality, but who will emerge as the top players in the space, and how the shift will influence business, marketing and planning.
Barbato, L. (2014, July 15). Happy Anniversary, Twitter! Here’s How Twitter’s Look Has Changed, From 2006 To Right Now — PHOTOS. Bustle. https://www.bustle.com/articles/31726-happy-anniversary-twitter-heres-how-twitters-look-has-changed-from-2006-to-right-now-photos
Cheng, J. (2007, May 29). CWA survey: average broadband speed in US is 1.9Mbps. Ars Technica. https://arstechnica.com/tech-policy/2007/05/survey-average-broadband-speed-in-us-is-1-9mbps/
Cocciolo, A. (2015, September 15). The rise and fall of text on the Web: a quantitative study of Web archives. Informationr.net. http://informationr.net/ir/20-3/paper682.html#author
Constantine, J. (2015, October 15). Facebook Tests Video Feed To Sidestep YouTube With Friendly Discovery. TechCrunch. https://techcrunch.com/2015/10/13/facebook-video-feed/?_ga=2.157740602.1222003649.1657569893-1374966353.1657569893
Desktop vs Mobile Market Share Worldwide. (2022, June). StatCounter Global Stats. https://gs.statcounter.com/platform-market-share/desktop-mobile/worldwide/#yearly-2011-2022
Efrati, A. (2016, April 7). Facebook Struggles to Stop Decline in “Original” Sharing. The Information. https://www.theinformation.com/articles/facebook-struggles-to-stop-decline-in-original-sharing?shared=eeb2ca
Glaze, V. (2021). Casey Neistat reveals what happened to his BEME app. Dexerto. https://www.dexerto.com/entertainment/casey-neistat-reveals-what-happened-to-his-beme-app-1621836/
Guinness, H. (2021, December 2). Camera sales are down, but the industry has survived similar slumps. Popular Photography. https://www.popphoto.com/news/the-camera-industry-is-changing/
Hutchinson, L. (2011, September 28). Information explosion: how rapidly expanding storage spurs innovation. Ars Technica. https://arstechnica.com/information-technology/2011/09/information-explosion-how-rapidly-expanding-storage-spurs-innovation/
Kanalley, C. (2011, August 2). INFOGRAPHIC: A History Of Facebook Photos. HuffPost. https://www.huffpost.com/entry/facebook-photos-infographic_n_916225
Kastrenakes, J. (2021, October 25). Facebook is spending at least $10 billion this year on its metaverse division. The Verge. https://www.theverge.com/2021/10/25/22745381/facebook-reality-labs-10-billion-metaverse
Kelleher, A. (2022, February 16). Moore’s Law – Now and in the Future. Intel. https://www.intel.com/content/www/us/en/newsroom/opinion/moore-law-now-and-in-the-future.html#gs.6fd3r1
King, C. (2015, August 24). Is Instagram’s Rapid Growth Slowing Down? Engagement And Follower Growth Down. Social Media Today. https://www.socialmediatoday.com/news/is-instagrams-rapid-growth-slowing-down-engagement-and-follower-growth-do/452861/
Knibbs, K. (2014, May 28). A whopping 1.8 billion images get shared and uploaded a day. The Daily Dot. https://www.dailydot.com/debug/mary-meeker-photo-report/
Kosoff, M. (2018, October 17). Was the Media’s Big “Pivot to Video” All Based on a Lie? Vanity Fair; Vanity Fair. https://www.vanityfair.com/news/2018/10/was-the-medias-big-pivot-to-video-all-based-on-a-lie
Li, Y., & Xie, Y. (2019). Is a Picture Worth a Thousand Words? An Empirical Study of Image Content and Social Media Engagement. Journal of Marketing Research, 57(1), 1–19. https://doi.org/10.1177/0022243719881113
Meyer, A. C. M., Robinson. (2018, October 18). How Facebook’s Chaotic Push Into Video Cost Hundreds of Journalists Their Jobs. The Atlantic. https://www.theatlantic.com/technology/archive/2018/10/facebook-driven-video-push-may-have-cost-483-journalists-their-jobs/573403/
Oremus, W. (2018, October 18). The Real Lie About Online Video Runs Deeper Than Facebook’s False Metrics. Slate Magazine; Slate. https://slate.com/technology/2018/10/facebook-online-video-pivot-metrics-false.html
Patricia Maranga. (2014, May 13). Social Photos Generate More Engagement: New Research. Social Media Marketing | Social Media Examiner. https://www.socialmediaexaminer.com/photos-generate-engagement-research/
Pew Research Center. (2019, June 12). Internet/Broadband Fact Sheet. Pew Research Center: Internet, Science & Tech. https://www.pewresearch.org/internet/fact-sheet/internet-broadband/
Pew Research Center. (2021, April 7). Mobile Fact Sheet. Pew Research Center: Internet, Science & Tech; Pew Research Center. https://www.pewresearch.org/internet/fact-sheet/mobile/
Redhead, C. (2013). Google Glass: How Could It Change Social Media and Recruitment? Undercover Recruiter. https://theundercoverrecruiter.com/google-glass-social-media-recruitment/
Sam Costello. (2019). How Every iPhone Ever Made Compares. Lifewire. https://www.lifewire.com/compare-iphone-models-1999430
Shephard, A. (2022, February 4). Facebook’s Sad, Hilarious Pivot (Again) to Video. The New Republic. https://newrepublic.com/article/165290/facebook-zuckerberg-metaverse-earnings-report
Shontell, A. (2014, August 25). The Only 8 Features Facebook Had When It Launched In 2004. Business Insider. https://www.businessinsider.com/facebooks-first-8-features-from-2004-2014-8
Sweney, M. (2022, April 9). The rise of TikTok: why Facebook is worried about the booming social app. The Guardian. https://www.theguardian.com/technology/2022/apr/09/rise-of-tiktok-why-facebook-is-worried-booming-social-app
Wohlsen, M. (2013, May 2). Guys Like This Could Kill Google Glass Before It Ever Gets Off the Ground. Wired. https://www.wired.com/2013/05/inherent-dorkiness-of-google-glass/