Digital Marketing Inquiry from the Fall 2025: What the Latest Research Tells Us

The final quarter of 2025 has brought a watershed moment in digital marketing scholarship. As we sift through the September and October journal publications, a clear picture emerges: we’re witnessing the convergence of artificial intelligence, consumer psychology, and commerce in ways that are fundamentally reshaping how brands connect with customers.

The Journal of Marketing Analytics devoted its entire September 2025 special issue to digital marketing and e-commerce analytics—a signal that these topics have moved from emerging trends to core strategic imperatives. Combined with breakthrough work published in the International Journal of Advertising, Journal of Global Marketing, and other peer-reviewed venues, we now have a robust evidence base for understanding where digital marketing is heading.

Here are the major themes and critical learnings that every marketer should understand.

Theme 1: The Analytics Arms Race Is Getting Serious (And More Accessible)

Perhaps the most cited paper from this cycle—Hair, Babin, Ringle, and colleagues’ tutorial on Covariance-Based Structural Equation Modeling (CB-SEM) using SmartPLS 4—has already garnered 7 citations and over 15,000 accesses in just weeks. Why does a methodological paper matter to practitioners?

The takeaway: Sophisticated statistical modeling is no longer the exclusive domain of data scientists. Modern tools are democratizing advanced analytics, allowing marketing teams to:

• Map complex customer journey paths with precision
• Test how multiple touchpoints interact to drive conversions
• Model the cascading effects of personalization on loyalty
• Validate which marketing mix elements truly move the needle

This matters because attribution has never been harder. With customers bouncing across devices, channels, and contexts, simple last-click models are hopelessly inadequate. CB-SEM provides a framework for understanding the structural relationships between marketing actions and business outcomes—essentially, it helps you build and test theories about what actually works.

Practical application: If your marketing analytics stack still relies primarily on descriptive dashboards and basic regression, you’re leaving strategic insights on the table. The barrier to entry for advanced modeling has dropped dramatically. Teams that invest in upskilling (or partnering with analytics-savvy vendors) will gain competitive advantage in optimizing increasingly complex digital ecosystems.

Theme 2: Conversational Commerce Is Crossing the Chasm

Schultz and Kaiser’s research on consumer value dimensions in conversational and mobile commerce (5 citations, 2,500+ accesses) provides empirical validation for what many of us have suspected: AI-driven shopping assistants are moving from novelty to necessity.

Their findings illuminate how consumers evaluate chatbots and voice assistants across multiple value dimensions:

• Functional value: Does it actually help me find what I need faster?
• Trust: Will this thing recommend products in my interest or the brand’s?
• Risk perception: Am I comfortable sharing data with an AI agent?
• Hedonic value: Is this experience enjoyable, or does it feel robotic and cold?

The critical insight: Consumers don’t evaluate conversational commerce on a single axis. A chatbot that’s highly functional but low on trust will fail. One that’s enjoyable but ineffective won’t convert. Brands need to design for multi-dimensional value.

What this means for your strategy:

If you’re implementing AI shopping assistants or conversational interfaces, you can’t just optimize for task completion rates. You need to instrument measurement across all value dimensions. Track not just “Did the customer complete checkout?” but “Did they trust the recommendations?” and “Was the interaction pleasant?”

The brands winning in conversational commerce are those that nail the balance—creating assistants that feel helpful without feeling invasive, efficient without feeling soulless.

Theme 3: Livestream Commerce Gets Academic Validation (Finally)

Three separate papers in this cycle examine livestream shopping from different angles, signaling that this format—already massive in Asia—is being taken seriously by Western marketers and researchers.

The credibility equation: Dao, Bui, Hoang, and Martinez studied cosmetic surgery livestreams (yes, really) and found that success requires a dual-credibility approach:

• Doctor credibility drives functional trust (“This person knows what they’re talking about”)
• Celebrity credibility drives emotional engagement (“This person is aspirational and entertaining”)

The most effective livestreams leverage both. A plastic surgeon alone might be credible but boring. An influencer alone might be engaging but lack authority. Together, they create a conversion engine.

The operational playbook: Zhuang and Xu’s research reveals which streamer tactics actually drive purchase value:

• Watch time is predictive (but not sufficient on its own)
• Product assortment strategy matters enormously
• Multi-channel presence amplifies impact
• Broadcast frequency affects purchase behavior non-linearly (more isn’t always better)

Strategic implications:

Livestream commerce isn’t just “video selling.” It’s a distinct channel with its own dynamics. Success requires:

1. Casting strategy: Match host credibility to product category (technical products need expertise; lifestyle products need aspiration)
2. Content strategy: Optimize show length, product mix, and pacing based on category and audience
3. Omnichannel integration: Viewers who see you on multiple platforms convert at higher rates
4. Measurement rigor: Track beyond just sales—monitor engagement patterns, repeat viewership, and audience composition

If you’re in retail, consumer goods, or any visual/demonstrable product category and you’re not experimenting with livestream formats, you’re ceding ground to competitors who are.

Theme 4: The Personalization Paradox Intensifies

Perhaps no topic is more fraught than hyper-targeted advertising. Research published in the International Journal of Advertising in September directly tackles the “creepy versus helpful” line in algorithmic personalization.

The fundamental tension: consumers want relevant ads but feel surveilled when targeting is too precise. Every marketer knows this intellectually, but now we have empirical frameworks for understanding where the line is.

Key findings:

• Receptivity to personalization varies by context (people accept health-related targeting differently than entertainment-related targeting)
• Transparency about data use shifts perception dramatically
• Prior negative experiences with personalization create lasting resistance
• The “value exchange” must be explicit and fair

The October follow-up: Research on anthropomorphic AI agents in advertising (published October 15) adds another wrinkle. When AI agents look or sound human, consumer expectations—and reactions to targeting failures—intensify. A badly targeted ad from a humanoid AI agent feels more intrusive than the same ad from an obviously algorithmic system.

What marketers must do:

1. Implement progressive disclosure: Don’t hit users with hyper-personalized experiences immediately. Build trust first.
2. Make the value exchange explicit: If you’re using location data to personalize, show the benefit clearly (“We’re showing you stores within 10 minutes of you”)
3. Provide meaningful control: Opt-out mechanisms must be genuine, not dark patterns
4. Context-match your personalization intensity: High-stakes categories (health, finance, children) require extra care; low-stakes categories (entertainment, general retail) have more latitude
5. Test your AI agent design carefully: If you’re using anthropomorphic agents, monitor for the “uncanny valley” effect where human-likeness backfires

The regulatory environment is tightening (GDPR, CCPA, AI Act in the EU). Brands that get ahead of this by earning permission for personalization rather than exploiting technical capabilities will build sustainable competitive advantage.

Theme 5: Digital Marketing Capability Is Now a C-Suite Conversation

Research published in the Journal of Theoretical and Applied Electronic Commerce Research examines the link between “digital marketing capability” (DMC) and actual firm financial performance using data from Chinese manufacturing firms.

Why this matters: CMOs have long struggled to demonstrate ROI on marketing technology investments. This research provides empirical ammunition: firms with stronger digital marketing capabilities—measured across technology infrastructure, data analytics competency, and organizational alignment—show measurably better financial performance.

The components of DMC that drive results:

• Technology stack sophistication: Not just having tools, but having integrated, modern tools
• Data literacy across the organization: Marketing can’t be the only function that understands digital data
• Agility in campaign execution: Speed of testing and iteration matters
• Customer data platform maturity: Unified customer views drive cross-channel effectiveness

Practical implication: When building your 2026 marketing budget, frame technology and capability investments in DMC language. Don’t ask for “a new CDP” or “marketing automation upgrade.” Ask for “investment in digital marketing capability that will improve customer acquisition efficiency, increase customer lifetime value, and drive measurable revenue growth.”

Back your request with benchmarking data: How does your DMC compare to competitors? Where are the gaps? What’s the expected ROI based on closing those gaps?

Theme 6: Deep Learning Transforms E-Commerce Forecasting

Ramos, Martinez, and colleagues applied deep learning models to U.S. e-commerce sales data, creating forecasting systems that predict sales, labor hours, and cost structures with remarkable accuracy.

Why this matters for practitioners:

Traditional forecasting models struggle with e-commerce because of:

• Seasonality that doesn’t follow retail calendar norms
• External shocks (social media virality, influencer effects, algorithm changes)
• Cross-channel substitution effects
• Rapid trend cycles

Deep learning models can capture these complex, non-linear patterns that traditional time series methods miss.

The strategic opportunity:

Better forecasting doesn’t just help supply chain—it transforms marketing strategy:

• Dynamic budget allocation: Shift spend in real-time based on predicted demand surges
• Promotional timing: Launch campaigns when models predict maximum receptivity
• Inventory-aware marketing: Avoid promoting products that will be out of stock; push products where inventory is building
• Staffing optimization: Align customer service capacity with predicted inquiry volume

Getting started: You don’t need a Ph.D. in data science. Cloud platforms (AWS, Google Cloud, Azure) offer forecasting services with deep learning models built in. Start with a pilot on a single product category or channel. Validate accuracy against your existing forecasting methods. Scale what works.

Theme 7: Sustainability and ESG Enter Digital Marketing (For Real This Time)

The editorial launching the September 2025 special issue in Journal of Marketing Analytics explicitly frames sustainability and ESG as core concerns for digital marketing’s future—not just peripheral “nice-to-haves.”

The shift: Previous research on sustainability in marketing focused on product claims, packaging, and supply chain messaging. Now, the focus is on the digital channel itself:

• Energy footprint of digital advertising: Programmatic auctions, video streaming, and AI models consume significant energy
• Data transparency as ESG issue: How companies collect, use, and protect customer data is increasingly viewed through an ESG lens
• Algorithmic fairness: Biased targeting and discriminatory outcomes are governance failures
• Digital waste: Email campaigns, abandoned content, legacy marketing tech stack bloat

What forward-thinking marketers are doing:

1. Measuring carbon footprint of campaigns: Tools now exist to estimate energy consumption of various media types and tactics
2. Prioritizing efficient formats: Static ads over video when appropriate; optimized code for websites and landing pages
3. Data minimization: Collecting only necessary data, deleting old data, providing genuine control to users
4. Algorithmic audits: Testing for bias in targeting, messaging, and measurement systems
5. Sustainability reporting in marketing metrics: Including ESG KPIs alongside traditional performance metrics

This isn’t virtue signaling—it’s risk management. Regulatory scrutiny is increasing, and consumer expectations are shifting, particularly among younger demographics.

Theme 8: Co-Creation and Storytelling Redux

October research on co-created advertising and narrative transportation theory (published October 30 in International Journal of Advertising) revisits a classic marketing topic with fresh digital-era insights.

The core finding: When consumers participate in creating brand narratives—through UGC campaigns, creator partnerships, or participatory content—narrative transportation (the feeling of being “swept up” in a story) is significantly stronger than with traditional advertising.

Why it works:

• Ownership: People who co-create content feel invested in its success
• Authenticity: User-generated stories feel more genuine than brand-produced ones
• Diverse perspectives: Co-creation surfaces narratives the brand might never have conceived
• Community building: Participatory campaigns create bonds between participants, not just with the brand

Examples of sophisticated co-creation:

• Lego Ideas: Customer-designed sets that become real products
• Starbucks “White Cup Contest”: Customers decorate cups, winning designs become limited editions
• GoPro user-generated hero videos: The brand’s entire content strategy is co-created
• Glossier’s product development process: Community input directly shapes product roadmap

Implementation framework:

1. Choose authentic invitation: Don’t fake co-creation; give real creative control
2. Provide scaffolding: Structure and constraints help, total freedom paralyzes
3. Celebrate all contributions: Not just winners; showcase the community
4. Close the loop: Show how input shaped outcomes
5. Compensate fairly: If you’re using UGC commercially, make the value exchange clear

Looking Forward: What This All Means for 2026 Planning

As you build your 2026 digital marketing strategy, these research insights suggest several imperatives:

1. Invest in analytical capability, not just analytics tools. The sophistication of available methods is increasing; the constraint is organizational capability to use them.

2. Treat conversational interfaces as a distinct strategic channel, not just a feature. They require dedicated design, measurement, and optimization approaches.

3. Experiment with livestream formats if you haven’t already. The evidence base is now strong enough to justify serious investment, particularly for visual/demonstrable products.

4. Rebuild trust around personalization by being transparent about data use, providing real control, and respecting context. The regulatory and consumer sentiment environment is shifting against surveillance-based marketing.

5. Frame technology investments as capability-building that drives measurable business outcomes. Connect marketing tech to revenue, not just efficiency.

6. Incorporate sustainability and ESG considerations into your digital marketing operations. This will move from optional to mandatory over the next few years.

7. Create real opportunities for customer co-creation in content and product development. The narrative power of participatory marketing is too strong to ignore.

---

The Bottom Line

The September-October 2025 research cycle marks a maturation point for digital marketing as an academic discipline. We’re moving beyond “does digital marketing work?” (obviously yes) to far more nuanced questions about how it works, when different tactics are most effective, and why consumers respond the way they do.

For practitioners, this is good news. We now have evidence-based frameworks for making better decisions across the full spectrum of digital marketing—from the technical (SEM modeling, deep learning forecasting) to the psychological (personalization receptivity, narrative transportation) to the operational (livestream strategies, capability building).

The brands that will win in 2026 and beyond are those that treat these academic insights not as ivory tower abstractions but as actionable intelligence. Read the research. Test the frameworks. Measure rigorously. Iterate constantly.

The digital marketing revolution isn’t coming—it’s already here. The question is whether you’re keeping pace.

---

What themes from this research are you most excited (or concerned) about? How is your organization adapting to these shifts? Let’s continue the conversation in the comments.