Social Media Marketing Is Changing Faster Than You Think

The End of the Social Media Playbook

A few years ago, if you wanted to sell something on social media, you had a simple formula: post often, post pretty pictures, and pay for ads. That formula is now a liability.

In 2020, a team of marketing researchers led by Yogesh K. Dwivedi at Swansea University gathered insights from dozens of experts across academia and industry. Their paper, published in the International Journal of Information Management, is not a typical academic study. It is a kind of warning. The authors argue that the entire foundation of social media marketing is shifting beneath our feet, and most companies have not noticed.

The paper has been cited over 2,400 times. That is not because it is popular. It is because it correctly predicted the chaos to come.

The Old Rules Were Built on a Lie

For a long time, social media marketing operated on a simple bargain. Brands would create content, users would see it, and some would buy. The platform controlled the feed. The brand controlled the message. The user was a passive consumer.

Dwivedi and his coauthors (Elvira Ismagilova, David L. Hughes, Jamie Carlson, and others) argue that this model is broken. The reason is not complicated: users got tired of being sold to. They developed what the authors call "ad avoidance" behaviors. They scroll past sponsored posts. They install ad blockers. They trust strangers on Reddit more than they trust a brand's official Instagram account.

The paper identifies a core tension. Social media platforms need to make money from advertising. Brands need to reach customers. But users increasingly resent both. The old playbook assumed that more content meant more sales. The authors found that the opposite is often true. Intrusive and irritating online brand presence actually drives customers away.

The Algorithm Does Not Care About Your Content Calendar

Here is the finding that should terrify every marketing director: the platforms have changed what they reward.

Dwivedi et al. (2020) document a shift in how social media algorithms work. In the early days, platforms rewarded frequency. Post five times a day, and the algorithm would show your content to more people. That era is over.

Modern algorithms prioritize engagement signals that are harder to fake. They look for comments, shares, saves, and time spent. A single post that sparks a conversation is worth more than fifty posts that get ignored. The authors call this the move from "broadcast" to "conversation" marketing.

The paper draws on expert interviews and a systematic review of existing research. The methodology is not a single experiment. It is a synthesis of what leading thinkers across multiple disciplines agree is happening. The authors asked experts to identify the biggest challenges and opportunities. The consensus was clear: the old metrics (likes, impressions, reach) are dying. The new metrics (conversation quality, trust, community sentiment) are harder to measure but matter more.

Artificial Intelligence Is Not a Tool. It Is a New Player.

Most companies treat AI as a way to write captions faster or schedule posts cheaper. Dwivedi and his team saw something else.

The paper devotes significant attention to artificial intelligence in marketing. The authors argue that AI is not just another automation tool. It fundamentally changes the relationship between brand and customer. AI can now generate personalized content at scale. It can predict what a user will buy before the user knows. It can even simulate conversations.

But here is the catch the authors emphasize: AI also creates new risks. When a brand uses AI to generate content, it loses control over tone and accuracy. When an AI chatbot interacts with a customer, it can say something offensive or wrong. The authors found that negative electronic word of mouth spreads faster and lasts longer than positive word of mouth. One AI mistake can undo months of careful branding.

The paper does not offer easy answers. It poses research propositions, not solutions. One of the most provocative: "Will consumers trust AI generated content as much as human created content?" The authors do not know. Neither does anyone else.

Augmented Reality Is Not a Gimmick

Most people think of augmented reality (AR) as a Snapchat filter or a Pokemon Go relic. Dwivedi et al. (2020) argue that AR marketing is one of the most underrated shifts in the industry.

The paper cites research showing that AR allows customers to try products before buying them. IKEA lets you see a sofa in your living room. Sephora lets you try lipstick on your face. These are not novelties. They solve a fundamental problem of online shopping: uncertainty.

When a customer can visualize a product in their own space, they are more likely to buy and less likely to return. The authors found that AR reduces cognitive load. It makes decisions easier. And it creates a sense of ownership before the purchase happens.

The implication is uncomfortable for traditional social media marketing. A static photo of a product on a beautiful model does not work as well as an AR experience that puts the product in the customer's hands. The authors predict that AR will become a standard feature of social commerce within five years. That prediction was made in 2020. It is happening now.

Mobile Marketing Is Not a Channel. It Is the Only Channel.

The paper includes a section on mobile marketing that reads like a eulogy for desktop advertising. Dwivedi and his team document that mobile devices have fundamentally changed when and where people consume content.

People check their phones while watching TV. They check them in line at the grocery store. They check them in bed. The authors found that mobile marketing is not just about smaller screens. It is about shorter attention spans and fragmented contexts. A user might see an ad while walking down the street, then not buy until three days later on a laptop.

This creates a measurement problem. The old attribution models assumed a linear path from ad to purchase. The authors found that the path is almost never linear. It is messy. It involves multiple devices, multiple platforms, and multiple distractions.

The paper argues that marketers need to stop thinking about "mobile marketing" as a separate category. Mobile is the default. Everything else is a variation.

B2B Marketing Is Finally Admitting It Is Social

For years, B2B companies told themselves they were different. They did not need social media. Their customers were professionals who made rational decisions based on spreadsheets and RFPs.

Dwivedi et al. (2020) dismantle this assumption. The paper includes a section on B2B marketing that makes a simple point: business buyers are still humans. They still scroll LinkedIn. They still read reviews. They still trust recommendations from peers more than they trust sales calls.

The authors found that B2B social media marketing faces a different set of challenges than B2C. The content needs to be more technical. The sales cycle is longer. The decision makers are multiple. But the underlying principle is the same: social media is where trust is built or destroyed.

The paper cites research showing that B2B buyers who engage with a brand on social media are more likely to make a purchase. The mechanism is not advertising. It is relationship building. The authors argue that B2B companies need to stop using social media as a broadcast channel and start using it as a listening channel.

What the Research Does Not Prove

The Dwivedi paper is a review and a research agenda, not a controlled experiment. It does not prove that any specific strategy works. It does not give you a formula for viral content. It does not tell you how many times to post per week.

What it does is identify the questions that matter. The authors are honest about what they do not know. They call for more research on the ethical implications of AI in marketing. They ask whether personalization can go too far. They wonder if social media platforms will eventually destroy the trust they depend on.

These are not weaknesses. They are the mark of good science. The paper is a map of the unknown, not a manual for the known.

What This Actually Means

• ▸Stop optimizing for volume. Posting more content does not help. Posting content that starts conversations does. Measure comments and shares, not impressions and likes.

• ▸Treat AI as a partner, not a replacement. AI can generate drafts and analyze data. It cannot build trust. If your AI makes a mistake, the human cost is high. Have a human in the loop.

• ▸Invest in AR and interactive experiences. Static images are dying. Let customers try your product in their own environment. The data shows it reduces returns and increases purchase confidence.

• ▸Accept that mobile is the only screen that matters. Design for the phone first. Assume the user is distracted. Make the path to purchase as short as possible.

• ▸B2B is not exempt. Your business customers are on social media. They are making decisions based on what they see there. If you are not showing up authentically, your competitors are.

• ▸Trust is the only sustainable advantage. Algorithms change. Platforms die. But a customer who trusts your brand will follow you to the next platform. That trust is built one conversation at a time, not one ad impression at a time.

The old playbook is not just outdated. It is actively harmful. The brands that survive the next five years will be the ones that understood this before everyone else.