Originally published November 2022. Updated June 2026 to grade our predictions against what actually happened — and to update the picture for where digital marketing stands heading into 2027.
Four years ago, we published six predictions about where digital marketing was heading. The article went live on 15 November 2022. ChatGPT launched on 30 November 2022. It is safe to say the subsequent four years exceeded our model.
This is the honest update — a proper scorecard, not a polished retrospective that quietly edits out the misses. We will grade each prediction, explain what actually happened, and close with what we think the next four years hold.
The 2022 predictions — graded
01. Artificial intelligence and machine learning — ✅ Correct (massively underestimated)
We predicted AI and ML would become "increasingly prevalent" in digital marketing. That is technically true in the same way that saying the Atlantic Ocean is "fairly large" is technically true.
The enterprise use cases we cited — personalised email, ad spend optimisation, customer behaviour analysis — did expand. But the scale and nature of the shift exceeded everything we outlined. The arrival of generative AI rewired the entire practice, not just the edges of it. Content production, ad creative, SEO strategy, customer support, social media management — every function now has AI operating inside it or replacing parts of it entirely.
We were right about the trend and wrong about the pace by at least a factor of five. The businesses that read this in 2022 and said "AI is coming, we'll start preparing" were already behind by mid-2023.
02. Personalization — ✅ Correct
Personalization has become the floor, not the differentiator. In 2022, we cited email open rates and click-through uplift as evidence it worked. In 2026, the question is not whether to personalise — it is how granularly, across how many channels, powered by what data.
AI-native personalization (real-time, per-user, spanning email, web, push notifications, and paid) is now accessible to mid-market brands, not just enterprise. We got this right. The ceiling on personalization has moved further than we described.
03. Video marketing — ✅ Correct
Understated if anything. Short-form video did not just grow — it restructured the attention economy. TikTok became the dominant discovery platform for under-35 demographics globally. YouTube Shorts and Instagram Reels followed at scale. The Wyzowl figure we cited (81% positive business impact) would look conservative against 2026 benchmarks, where video-led brands see engagement differentials against text-first brands that simply were not achievable in 2022.
The one thing we missed within this prediction: AI-generated video production. The cost of producing short-form content has collapsed. A two-person team with current AI tooling can produce what required a five-person team and a studio budget in 2022. The edge is no longer production quality — it is positioning depth and audience specificity.
04. Social media marketing — ✅ Correct, but the landscape shifted
Social media is still essential. What changed: the platform map looks completely different from the one we had in mind. X (Twitter) lost advertiser confidence after the 2022 ownership change and never fully recovered as a marketing channel. TikTok became a serious commerce platform. LinkedIn's organic reach — consistently underrated — became exceptional for B2B as AI-generated noise flooded other channels with undifferentiated content.
The prediction held. The execution playbook needed rewriting every eighteen months.
05. Voice search optimisation — ⚠️ Half right, wrong delivery mechanism
We predicted Alexa and Google Home would drive voice search dominance, requiring businesses to optimise for conversational, natural-language queries. Smart speaker adoption plateaued. Voice search as a standalone marketing strategy quietly faded from most agencies' conversations by 2024.
But the underlying prediction was correct. The insight — that people would increasingly ask natural-language questions and expect direct answers rather than a list of links — was real. It just arrived differently than we expected.
The "answer engine" of 2026 is not Alexa. It is Google's AI Overviews, Perplexity, ChatGPT's search mode, and Claude. The optimisation strategy we described — clear, direct, conversational content that answers questions completely — is now called Generative Engine Optimisation (GEO). The techniques are largely the same. The platforms are unrecognisable from what we had in mind when we wrote this.
06. The metaverse — ❌ Missed
We will be direct about this one. The metaverse prediction did not land.
Meta spent more than $40 billion building Horizon Worlds. It attracted modest adoption and extensive public ridicule. Apple launched Vision Pro in February 2024 to strong critical response and weak consumer sales — priced at $3,499 during a cost-of-living crisis. VR and AR have not become the primary commerce environment that Meta's marketing (and our 2022 article) suggested.
What we got right within the miss: the underlying user need we were pointing at — more immersive, more contextual, more personalised commerce — is real and is being met. It just arrived through AI, not VR. Virtual try-on tools, AI-generated product visualisation, and conversational shopping experiences are the actual version of what we were describing. The technology changed. The need was real.
What we missed entirely
Four correct calls, one half-right, one miss — and we still did not predict the most significant structural shift: the breaking of the traffic model.
Digital marketing from 2005 to 2022 ran on one loop: create content → rank on Google → attract clicks → convert visitors. The entire industry — SEO, content marketing, landing page optimisation, analytics infrastructure — was built around that loop.
That loop is fracturing. Google's AI Overviews answer queries directly on the results page. Perplexity, ChatGPT, and Claude give comprehensive answers without routing users anywhere. Studies from early 2026 show over 60% of Google searches now end without a click. The number is rising.
The correct response is not panic. The goal of marketing was always to reach the right person at the right moment with the right message. Previously that meant ranking for keywords and earning clicks. Now it also means being the source that AI tools cite — being mentioned and attributed in AI-generated answers even when no one clicks through to your site.
That requires different content (authoritative, specific, original), different technical infrastructure (structured data, entity clarity, machine-readable formats like llms.txt), and a different success metric (brand mentions, citation frequency, branded search volume — not just organic click count).
Nobody in 2022 was writing about this. We were not either. We should have been.
The 2026 real picture
Generative Engine Optimisation (GEO)
The practice of optimising for AI citation engines alongside traditional search. The brands appearing in ChatGPT's answers, Perplexity's cited sources, and Google's AI Overviews are winning brand awareness at zero cost-per-click. GEO requires structured content, clear entity definition in Schema.org markup, and authoritative backlinks — not tricks, just harder and more durable fundamentals than most agencies want to commit to.
Agentic marketing operations
The next layer beyond AI tools is AI agents — systems that do not assist marketers but run marketing functions autonomously. Lead qualification agents that work around the clock. Content engines that produce, schedule, and distribute without human intervention. Reporting systems that surface performance anomalies before any human notices. The businesses building these now are operating at a leverage ratio their competitors cannot match manually.
We have deployed these for clients across real estate, retail, and industrial sectors. The operational difference between a team with well-built agents and a team without is not marginal. It is categorical.
First-party data as the primary moat
Third-party cookies are gone. Platform algorithm reach is unreliable and shrinking. The brands with rich first-party data — direct email lists, SMS subscribers, logged-in customer bases — have a structural advantage that no advertising platform can replicate or remove. Building owned audiences is not a nice-to-have in 2026. It is the only durable channel.
Creator-brand convergence
The creator economy did not replace brand marketing — it merged with it. The most effective brand content in 2026 is indistinguishable from creator content: opinionated, specific, filmed like authentic UGC, and distributed through individual voices rather than brand accounts. Brands that have understood this consistently outperform brands running polished "brand" content on engagement, reach, and conversion.
Predictions for 2026–2030
New predictions, made with full awareness that some will be wrong. We will grade these too in four years.
Agentic operations become the baseline, not the edge. By 2028, having AI agents handling lead intake, customer follow-up, reporting, and content distribution will be a hygiene factor, not a competitive advantage. Companies that do not have them will be structurally slower and more expensive to operate than those that do.
AI-mediated shopping restructures discovery. A meaningful share of consumer purchases will be initiated through AI agents rather than direct search within three years. A user asks their personal AI to find the best product for a specific need and budget. The AI queries product data, compares options across merchants, and presents a shortlist. The brands with structured product data, real-time pricing, and machine-readable catalogues will be surfaced. The ones optimising for human Instagram feeds may be invisible to the systems routing purchasing decisions.
Brand becomes the only durable SEO moat. As AI commoditises content production and zero-click continues to rise, businesses with genuinely recognisable brands will hold traffic. Branded search — people searching your name directly — will become the most valuable signal in analytics. Everything that builds real brand: consistent story, product quality, community, earned credibility — will have higher marketing ROI than at any point since the early internet.
The metaverse arrives — through AI, not VR. The immersive commerce vision from 2022 will be delivered through AI-generated environments, spatial web experiences, and conversational product interaction rather than headsets and virtual malls. Slower arrival. Different shape. The underlying need — more contextual, more personal shopping — is real, and AI is the actual delivery mechanism, not Meta's hardware.
The honest close
Four of six predictions correct. One half-right with the wrong mechanism. One significant miss on the metaverse.
The scorecard we are proudest of is not the predictions themselves — it is that we made them publicly, with our name on them, and are grading them honestly now. Most agency content does not do this. Most agency content is written to sound authoritative, not to be held accountable.
The businesses that will look back at 2026 the way we are looking back at 2022 — recognising they made the right moves early — are the ones investing in these foundations now, before they feel obvious. The 2022 fundamentals are still the floor: good content, genuine social presence, video that earns attention, personalised customer experience. The ceiling in 2026 is substantially higher.
Fourlines Agency has been building brands and the digital infrastructure that runs them since 2017. If you want to understand how these shifts apply to your specific business — not a generic deck, a real conversation — start here.