⚡ Quick Summary
- User engagement on Instagram, Threads, and LinkedIn fell measurably in 2024, with session lengths declining 3–5% among key demographics as ad load and recommended content crowded out followed accounts.
- The root cause traces back to Apple's 2021 App Tracking Transparency framework, which cost Meta an estimated $10 billion in annual ad revenue and triggered aggressive in-platform monetisation strategies.
- LinkedIn's declining engagement directly threatens the quality of Microsoft 365 integrations — including Dynamics 365 CRM, Outlook contact cards, and Teams — that rely on active LinkedIn data signals.
- Competing platforms including YouTube, Substack, and Bluesky are capturing users and creators migrating away from algorithmically saturated feeds, with YouTube ad revenue growing 15.5% YoY in Q3 2024.
- Security teams should treat increased algorithmically recommended content on professional platforms as an expanded social engineering attack surface requiring updated threat models and employee training.
What Happened
New data tracking user behaviour across major social platforms has confirmed what many users have long suspected: time spent on Instagram, Threads, and LinkedIn all declined measurably over the past year. The culprit, according to analysts and user sentiment data alike, is not a lack of interest in social networking itself — it is the aggressive algorithmic prioritisation of advertisements, sponsored content, and algorithmically recommended posts over content from accounts users have deliberately chosen to follow.
On Instagram, Meta's flagship photo and video platform with over two billion monthly active users, the ratio of recommended-to-followed content in a typical feed session has reportedly shifted dramatically. Internal testing data and third-party analytics firms including Sensor Tower and data.ai both noted declining session lengths in 2024, with average daily time-on-app falling by an estimated 3–5% year-over-year among users aged 25–44 — historically the platform's most commercially valuable demographic.
Threads, Meta's text-based platform launched in July 2023 as a direct rival to X (formerly Twitter), saw its initial explosive growth — it reached 100 million sign-ups in under five days — give way to a pattern of declining return visits. The platform's aggressive injection of recommended posts from accounts users never followed has been widely cited in user feedback and app store reviews as a primary frustration.
LinkedIn, Microsoft's professional networking platform with over one billion members globally, has faced similar criticism. The platform's feed has become increasingly saturated with sponsored posts, algorithmically surfaced "suggested" content, and what users have termed "engagement bait" — motivational posts and viral professional anecdotes that crowd out genuine industry updates and connections' professional activity. Engagement metrics tracked by third-party tools such as Shield Analytics showed measurable declines in organic post reach throughout 2024.
The pattern is consistent and telling: platforms optimising aggressively for advertising revenue in the short term are measurably eroding the user experience that made them valuable in the first place.
Background and Context
This decline did not emerge in a vacuum. It is the culmination of a strategic pivot that began in earnest across major social platforms around 2020–2022, accelerated by the twin pressures of post-pandemic advertising market contraction and the seismic disruption caused by Apple's App Tracking Transparency (ATT) framework, introduced in iOS 14.5 in April 2021.
ATT effectively required apps to obtain explicit user permission before tracking activity across third-party apps and websites — a change that Meta alone estimated cost it approximately $10 billion in lost advertising revenue in 2022. To compensate, Meta and its peers began leaning more heavily on on-platform behavioural signals and expanding the volume of ads served per session. Instagram, which had maintained a relatively restrained ad load in its early years post-acquisition (Facebook acquired Instagram for $1 billion in 2012), began increasing ad frequency significantly from 2022 onwards.
Simultaneously, TikTok's meteoric rise — it surpassed one billion monthly active users in September 2021 — demonstrated the commercial power of the interest-graph model: showing users content based on inferred preferences rather than explicit social connections. Meta responded by fundamentally reorienting Instagram's algorithm toward a similar discovery-first model. In 2022, CEO Adam Mosseri publicly acknowledged the shift, stating that Instagram was "no longer a photo-sharing app" but a broader entertainment platform. The backlash from creators and users was immediate and vocal, leading to a partial rollback — but the underlying algorithmic direction did not reverse.
LinkedIn's trajectory mirrors this pattern. Microsoft acquired LinkedIn for $26.2 billion in December 2016, and for several years maintained a relatively clean professional feed. But as LinkedIn's revenue — which crossed $15 billion annually by fiscal year 2023 — became an increasingly important contributor to Microsoft's commercial cloud segment, pressure to monetise the feed more aggressively intensified. The introduction of LinkedIn's own short-video feed in 2023, directly mimicking TikTok's format, further diluted the professional utility that distinguished the platform.
The broader context is a social media industry caught between two incompatible goals: maximising advertising yield per session and maintaining the authentic social utility that justifies user attention in the first place.
Why This Matters
For businesses and IT professionals, the declining engagement on these platforms carries implications that extend well beyond marketing department concerns. LinkedIn, in particular, occupies a unique position in the enterprise technology ecosystem — it is simultaneously a recruitment platform, a B2B marketing channel, a professional learning resource (via LinkedIn Learning), and increasingly, a Microsoft 365 integration point.
Microsoft has invested heavily in weaving LinkedIn data into its broader productivity and CRM stack. LinkedIn Sales Navigator integrates directly with Microsoft Dynamics 365 and, via connectors, with Salesforce. LinkedIn's profile data surfaces within Microsoft 365 contact cards, Teams conversations, and Outlook. If LinkedIn's user engagement continues to erode — if professionals visit less frequently and interact less authentically — the signal quality of that data degrades, and with it the value of these integrations for enterprise sales and HR teams.
For IT departments managing Microsoft 365 deployments, this is worth monitoring. Organisations that have built recruitment workflows, sales intelligence pipelines, or learning and development programmes around LinkedIn's platform are exposed to platform risk if engagement decline accelerates. Diversifying talent acquisition channels and not over-indexing on LinkedIn for employee development — supplementing it with internal learning platforms or Microsoft Viva's learning modules — is a prudent response.
From a security and data governance perspective, the algorithmic shift also raises concerns. As platforms push more recommended content, users encounter more content from unknown sources, increasing exposure to misinformation, phishing lures disguised as professional content, and social engineering attempts. Security awareness training programmes should increasingly address the risks of engaging with algorithmically surfaced professional content on LinkedIn — a vector that corporate security policies have historically underweighted compared to email phishing.
For businesses investing in enterprise productivity software stacks, the lesson is also a broader one about platform dependency: building critical workflows on third-party platforms whose incentive structures are misaligned with user value creates long-term strategic risk.
Industry Impact and Competitive Landscape
The beneficiaries of this algorithmic fatigue are varied and instructive. YouTube, which has maintained a more balanced approach to recommended versus subscribed content — and whose subscription-based YouTube Premium tier creates a direct financial incentive to keep users satisfied — has seen sustained engagement growth. According to Google's earnings disclosures, YouTube's advertising revenue grew 15.5% year-over-year in Q3 2024, suggesting the platform is capturing attention migrating away from Meta's properties.
Substack and newsletter platforms have experienced a renaissance, as creators and professionals frustrated with algorithmic unpredictability seek direct audience relationships via email. Substack reported crossing 35 million active subscriptions in 2024. This trend directly threatens LinkedIn's creator ecosystem and its ambition to become the default platform for professional thought leadership.
X (formerly Twitter), despite its turbulent post-Elon Musk acquisition period, has paradoxically benefited among certain professional and tech communities who value its chronological and interest-based feed options — even as its overall user base and advertiser relationships remain volatile. Bluesky, built on the open AT Protocol and offering users genuine algorithmic choice, grew to over 20 million users by late 2024, appealing specifically to users who want transparency and control over their content feeds.
For Meta specifically, the competitive threat is existential at the margins. While Instagram and Facebook's combined scale — over three billion daily active users across Meta's family of apps — provides enormous inertia, the 18–24 demographic that platforms compete most fiercely for is showing the clearest signs of migration toward TikTok, YouTube Shorts, and emerging platforms. Meta's own internal research, portions of which became public via the 2021 Frances Haugen disclosures, acknowledged that younger users found Instagram less compelling than older cohorts.
Microsoft's response to LinkedIn's engagement challenges will be worth watching. The company has the unique advantage of being able to embed LinkedIn utility directly into tools professionals already use daily — Teams, Outlook, and Microsoft 365 Copilot. Deepening these integrations may prove more valuable than competing on feed quality alone. Organisations already running affordable Microsoft Office licences are well-positioned to leverage these native integrations without additional platform costs.
Expert Perspective
From a strategic analysis standpoint, what we are witnessing is the maturation crisis of the attention economy's dominant model. The ad-supported social platform, as currently architected, contains a structural contradiction: its revenue mechanism (advertising) is optimised by maximising impressions and time-on-platform, but its value proposition (social connection and relevant information) is degraded by the same maximisation behaviour.
Industry analysts at firms like Forrester and Gartner have been flagging "platform fatigue" as a meaningful enterprise risk category since 2022. What the 2024 engagement data confirms is that this fatigue is not merely anecdotal — it is measurable and accelerating among precisely the high-value professional demographics that platforms charge premium CPMs to reach.
The technical architecture of modern recommendation systems — transformer-based models trained on engagement signals, increasingly augmented by large language model capabilities — is extraordinarily good at capturing short-term attention but demonstrably poor at preserving long-term user satisfaction. When every piece of content is optimised for the click, the scroll, or the reaction, the aggregate feed becomes a relentless stimulus environment that users eventually find exhausting rather than rewarding.
The platforms that will win the next decade are those that find a sustainable equilibrium — likely through diversified revenue models that reduce dependence on pure ad volume. Apple's services model, which bundles content and platform access through subscription, offers one template. The question is whether Meta, LinkedIn, and their peers have the structural flexibility and shareholder tolerance to pursue that transition before engagement erosion becomes a genuine revenue crisis.
What This Means for Businesses
For business leaders and marketing teams, the 2024 engagement data should prompt a serious audit of social media investment strategy. Platforms with declining organic reach and rising ad costs represent deteriorating ROI — and the trend lines suggest this will worsen before it improves.
Practically, businesses should consider three adjustments. First, diversify content distribution away from algorithm-dependent platforms toward owned channels: email newsletters, company blogs, and direct community platforms like Slack Connect or Microsoft Teams external communities offer more predictable reach. Second, for LinkedIn specifically, focus investment on LinkedIn's more structured products — job postings, LinkedIn Learning licences, and Sales Navigator — rather than organic content marketing, which is increasingly subject to the same algorithmic unpredictability affecting other platforms.
Third, IT departments should reassess the security posture around employee use of professional social platforms. As recommended content volumes increase, so does the attack surface for social engineering. Updated acceptable use policies and security awareness training that specifically addresses LinkedIn-based phishing — a growing vector for credential harvesting targeting enterprise employees — should be prioritised.
For organisations managing Microsoft 365 environments, ensuring staff are fully utilising native productivity and communication tools reduces dependency on external platforms. Teams with a genuine Windows 11 key and a properly licensed Microsoft 365 stack have access to Viva Engage, Viva Learning, and Teams — tools that can replicate many of LinkedIn's internal knowledge-sharing functions within a governed, secure enterprise environment.
Key Takeaways
- Time spent on Instagram, Threads, and LinkedIn declined measurably in 2024, driven primarily by algorithmic prioritisation of ads and recommended content over followed accounts.
- The shift is a direct consequence of platforms compensating for post-ATT advertising revenue losses by increasing ad load and adopting TikTok-style interest-graph recommendation models.
- LinkedIn's declining engagement has specific enterprise implications, degrading the signal quality of Microsoft 365 integrations that rely on LinkedIn activity data for sales intelligence, recruitment, and professional networking.
- Competing platforms — YouTube, Substack, Bluesky, and newsletter ecosystems — are capturing attention and creator investment migrating away from algorithmically aggressive platforms.
- Security teams should update threat models to account for increased social engineering risk via algorithmically surfaced professional content on LinkedIn and similar platforms.
- Businesses should diversify content distribution toward owned channels and evaluate whether LinkedIn's structured enterprise products (Sales Navigator, LinkedIn Learning) deliver better ROI than organic social investment.
- The structural contradiction between advertising maximisation and user value creation represents a long-term existential challenge for the ad-supported social platform model.
Looking Ahead
Several developments in 2025 will determine whether this engagement decline is a correctable course adjustment or a structural inflection point. Meta's Q1 and Q2 2025 earnings calls will be closely watched for any acknowledgement of engagement metric softness — the company has historically been guarded about releasing granular time-on-platform data.
Regulatory pressure may also accelerate change. The EU's Digital Services Act (DSA), now in full enforcement, requires very large online platforms to offer users algorithm-free, chronological feed options. Compliance with DSA provisions — and potential extension of similar requirements in UK and US regulatory frameworks — could structurally alter how platforms balance recommendation versus followed content.
Microsoft's Build 2025 conference and LinkedIn's anticipated product announcements around deeper Microsoft 365 Copilot integration will indicate whether Microsoft is doubling down on embedding LinkedIn utility into the productivity stack as a hedge against platform engagement risk.
Finally, the emergence of genuinely open social protocols — AT Protocol (Bluesky), ActivityPub (Mastodon, Pixelfed) — as viable alternatives will continue to pressure incumbent platforms to either improve user experience or risk accelerating creator and audience migration to decentralised alternatives that structurally cannot replicate the same algorithmic overload model.
Frequently Asked Questions
Why are users spending less time on Instagram and LinkedIn?
The primary driver is the aggressive algorithmic shift toward showing users ads, sponsored posts, and recommended content from accounts they never chose to follow — at the expense of posts from their actual connections and followed accounts. This shift intensified from 2022 onwards as platforms sought to compensate for advertising revenue losses caused by Apple's App Tracking Transparency framework, which disrupted cross-app tracking and cost Meta alone an estimated $10 billion annually. Users, particularly in the 25–44 age bracket, are responding by shortening sessions or visiting less frequently.
How does LinkedIn's declining engagement affect Microsoft enterprise customers?
LinkedIn is deeply integrated into Microsoft's enterprise stack. Its data surfaces within Microsoft 365 contact cards, Outlook profiles, Teams conversations, and Dynamics 365 CRM via Sales Navigator. If LinkedIn engagement declines — meaning professionals post less, interact less authentically, and visit less frequently — the quality and currency of that data degrades. This reduces the practical value of LinkedIn-powered features in Microsoft 365 for sales teams, HR professionals, and anyone relying on LinkedIn signal data within their productivity workflows.
What should IT departments do in response to this trend?
IT departments should take three actions. First, reassess security awareness training to include LinkedIn-specific social engineering threats — as more algorithmically surfaced content appears in professional feeds, the risk of phishing lures and credential harvesting attempts disguised as professional content increases. Second, evaluate whether critical talent acquisition, sales intelligence, or learning workflows are over-dependent on LinkedIn's platform, and identify Microsoft-native alternatives such as Viva Learning and Viva Engage. Third, review the organisation's social media acceptable use policy to reflect the changed risk environment of recommendation-heavy professional platforms.
Are there alternative platforms benefiting from this engagement decline?
Yes. YouTube has seen sustained advertising revenue growth, rising 15.5% year-over-year in Q3 2024, as its more balanced recommendation model retains users. Substack crossed 35 million active subscriptions in 2024 as creators seek direct, algorithm-independent audience relationships via email newsletters. Bluesky, built on the open AT Protocol with user-controlled algorithmic choice, grew to over 20 million users by late 2024. These platforms share a common characteristic: they either offer users genuine control over their content experience or provide a direct financial incentive (subscriptions) to prioritise user satisfaction over raw ad impression volume.