How Disruptive Tech Is Rewiring Digital Marketing
By Drew Mabry, Founding Partner of Fast + Light
The last few years in marketing have felt like crossing a canyon on a rope bridge. On one side, the “pandemic era” forced us to get maniacally good at delivery—standing up SOPs, building teams, shipping work on time because demand and stimulus cash were colliding with a homebound world. On the other side lies what I think of as “AI culture,” which runs on innovation, velocity, and a fair bit of chaos. Somewhere in the middle of that swaying bridge is where most marketers live right now: excited, skeptical, and trying not to drop the bag while the planks wobble.
I’ve spent decades moving between hands-on multimedia, systems, CRM and BI, enterprise software, and consulting before co-founding an agency that works month-to-month, sits with clients twice a month, and talks about the hard stuff—revenue, KPIs, creative that actually converts, and the uncomfortable bits that get people fired faster than poor performance ever does. We call ourselves mechanics for your revenue for a reason. We don’t polish the chrome; we tune the engine.
AI is the latest engine part everyone wants to swap in. A lot of what’s in market today is really just prompting on top of LLMs, and plenty of it gets oversold. Still, the real shifts are undeniable: personalization at scale is finally moving from slideware to shop floor; precision targeting and real-time optimization are changing creative and spend decisions; analytics and attribution are evolving amid walled gardens and privacy postures. And, above all, the “human in the loop” has never mattered more, because AI is probabilistic, not deterministic. Ask it the same question five times and you’ll get five different answers. That’s power—and risk. The job now is using that power without outsourcing our judgment.
Below, I’m going to break down what this disruption looks like across three zones—advertising, SEO, and operations—drawing directly from what we discussed on the TechConnect Podcast. If you’re feeling the bridge sway under your feet, you’re not alone. But there’s a way across.
Advertising: From “More Variants” to “More Velocity With Judgment”
Marketing’s most visible surface area is still advertising, and AI is already changing the pace and the palette. For small and midsize brands—think two to a hundred million in annual revenue—the fantasy that AI will let you “replace the team” is just that. You still need the crew, the software, and the shared rhythm of feedback and approvals. What AI gives you is throughput. With a team of seven on a modest account, we can now create three to four times the creative volume we did before, craft sharper offers, explore more angles, and test harder and faster. You don’t throw out the playbook; you run more plays.
This is where “human in the loop” stops being a cliché and becomes a control system. AI can brainstorm copy variations and moodboards in minutes; it can also generate what I call AI slop—text with the telltale m-dash addiction and a tone that reads like it was milled in a word factory. You can feel when there was no human comb pass, no intent behind the syntax. The result is work that “looks” like marketing but carries none of the brand’s feeling. And in marketing, you get fired for feelings far more often than for numbers. The wrong shade of green, a tone-deaf noun, a word like “panties” where your audience hears something else—these become signals that you don’t “get” the brand, even if the KPI chart is smiling.
Influencers are another example of AI shifting the economics of effort. Anyone who’s actually run influencer programs knows the hidden cost: discovery, outreach, contracting, briefing, shipping product, creative back-and-forth, formatting, approvals, timing, compliance, and reporting. It’s a lot. AI-generated influencer personas let us pressure-test ten angles quickly, find the one narrative that actually resonates, and then, when it makes sense, bring in human creators with the right authenticity to scale that narrative. It’s not replacing people; it’s getting to the right people faster. And it can reduce brand-safety exposure because you’re exploring early with a synthetic front before you put your logo in someone else’s hands.
The platforms themselves are nudging us toward broader buys where their own systems optimize delivery. If Meta and Google want us to set wide targets and let their machines hunt, then the differentiator becomes creative craft and learning velocity. AI assists here, too. We can turn around more variants, shape headlines and hooks in context, and even pre-score likely winners based on past patterns. But none of that absolves the need for care. We enforce at least two sets of human eyes on anything that leaves our shop. We’re also piloting the boring, essential parts of quality: two separate AI pipelines tasked with checking each other’s math on analytics rollups so that a probabilistic error doesn’t slip through and corrupt the next month’s decisions.
If there’s a single principle to hold in advertising right now, it’s this: use AI to generate surface area—more bets, faster—without letting it set your standards. Let it propose. Let humans choose.
SEO: From Wide Nets to Narrow Moats
If advertising is the loud stage, SEO is the shifting ground underneath it. Organic traffic is down in many categories. People ask questions in chat interfaces and accept the answer without clicking. Reddit threads show up because people want human texture in responses. Meanwhile, a field with many names—GEO, GSO, AEO, pick your acronym—is forming around optimizing for generative engines. It’s early. It’s messy. And yes, it’s already influenceable.
The honest state of play is that what ranks in generative answers can still be nudged in ways that won’t last forever. If you publish a local “top podcasts” post and structure it correctly, you may find yourself cited alongside outlets that would have outgunned you a year ago. That’s a land-grab moment, not a long-term strategy. The rules will tighten. The question isn’t just “how do I get in the answer box today,” but “what kind of content earns durable inclusion when the box gets smarter?”
I’m convinced the posture shift is from wide to deep. The old reflex of “we can be everything to everyone” collapses under generative selection pressure. If AI’s job is to give the best single answer, then your job is to be the best single answer in a narrow, well-defined niche. That means deciding—really deciding—what you want to be known for, and accepting that focus is a moat. It also means rethinking structure. Schema, syntax, and clarity matter. We’ve all seen local-search basics finally sink in for small businesses—clean NAP consistency across properties, maintained Google Business and Yelp profiles, structured data that reduces ambiguity. Extend that thinking. Organize your content so an AI can extract a clean answer without inventing one. Use language that’s explicit about the question you’re addressing and the context you’re operating in.
There’s a debate about whether to write “for humans” or “for AIs.” My view is that you aim to serve both, with a tilt toward machines in the medium term. No-click experiences aren’t going away. On phones, in assistants, in SERP overlays, answers will be rendered and consumed without a visit. The paradox is that the same editorial values carry you whether a person is reading or a model is parsing: originality, specificity, and helpfulness. Bland listicles won’t win with readers or models. Real examples, clear definitions, and concrete outcomes stand out to both.
There’s also a cultural layer. We’re in that early, hackable era where experiments move the needle and the playbook isn’t settled. That’s fun, but it’s temporary. What endures is a posture of informed focus: choose the segment, speak to it like an expert, back it with structured clarity, and be ready to evolve when the rails move. If you want breadth, earn it by stacking depth.
Operations: Numbers Over Noise, Without Losing the Narrative
Most of the disruption that will actually change your P&L won’t be visible in your ads or your rankings; it will be in how your team works and how your decisions get made. This is where AI’s ability to move data around, find patterns, and eliminate drudgery is already saving real hours—and turning those hours into better outcomes.
We meet with clients twice a month. Historically, one team member would spend about forty hours compiling cross-channel analytics into a coherent story: Meta and Google ads data, Shopify sales, email cohorts, site behavior. That time wasn’t wasted, but it was work that begged to be automated. Our near-term goal is to let AI handle roughly thirty of those forty hours, and we’re close. The gains aren’t just in speed; they show up in formatting that makes the conversation easier, in surfacing anomalies we would have noticed later, in freeing humans to ask better questions.
Of course, the uncomfortable question follows: how do you know the AI is right? We’re addressing that by instrumenting redundancy. Two different systems, following the same logic, checking each other’s outputs each cycle. It’s an old engineering idea adapted to a probabilistic world. We’re also leaning on tools purpose-built for this era. Walled gardens have forced us to find cross-channel visibility elsewhere, and that’s where platforms like Triple Whale earn their keep with journey stitching that makes business sense to smaller brands. When we can see that a YouTube view assisted a newsletter sign-up that preceded a paid conversion, spend allocation stops being a turf war and becomes a weighting exercise. Their AI agent accelerates the read so we can reallocate earlier in the month rather than waiting for “statistical significance” in a world that changed last week.
All of this presses on a KPI shift. ROI remains a useful diagnostic, but it’s a brittle compass. Marketing Efficiency Ratio (MER)—what we spend against what we get, in the full context of costs and LTV—maps better to reality. Because we can pipe in COGS, shipping, financing constraints, and merchandising strategy, we aren’t arguing about a channel’s last-click glory; we’re evaluating whether the system is producing cash in a way that matches how the business actually operates. That’s how you answer the CFO’s question without playing dashboard roulette.
Adopting this way of working is as much a cultural gambit as a technical one. I shared on the show the story of a CEO who paused Mondays company-wide to go all-in on AI, invested a fifth of cash into the transition, and faced down massive internal resistance—especially from technical teams who couldn’t get on board. The cost was high; the payoff was real. Whether or not you’d make the same call, the lesson holds: the blocker won’t be model quality; it will be human appetite for change. Marketing teams, interestingly, often embrace these tools sooner because we’re mercenary about outcomes. When the answer says, “This ugly ad makes 70 percent of your money,” the ego response is to kill it; the operator response is to let it run and build a better one beside it. Numbers over ideas isn’t a rejection of creativity—it’s a demand that ideas earn their keep.
There are limits, and ignoring them gets you burned. If a brand insists on approving two hundred creative pieces monthly on a modest spend, AI can produce the assets, but approvals will choke the pipeline and nothing will ship. If leadership wants someone to “own AI,” they’ll miss the point that responsibility must be distributed: everyone who touches the output is on the hook for sanity-checking it. If you try to automate the full loop now, you’ll end up with pretty dashboards no one trusts. The right move today is to incur controlled technical debt in exchange for learning. Build the pipelines, accept some mess, and put humans at the gates where it matters.
Personalization at Scale, With People at the Center
The biggest promise finally coming good is personalization at scale—real personalization, not “Hi, {FirstName}.” The thought experiment I like is the “Jennifer Aniston sweater.” Years ago, we dreamed of seeing something on a show and buying it in a click. Today, we’re approaching something more uncanny: systems that can predict what a customer will want before they go looking and place it in their path at the right moment. That’s heady stuff.
It’s also the kind of power that can go sideways if you forget the human being receiving it. Predictive engines will line up offers, but a person’s sense of your brand is still formed by tone, imagery, and the feeling you create. That’s why the slop tells on itself. It’s why the difference between “underwear” and “panties” isn’t trivial in a particular context. It’s why micro-influencers with lived credibility still move customers in ways stock avatars cannot. Personalization doesn’t mean without people; it means for people. When we keep humans in the loop, we’re not slowing the machine; we’re giving it a conscience.
Strategy in an Innovation Culture
What’s changing now is not just a technology stack; it’s a company posture. In the pandemic, delivery culture won: process, predictability, hands on deck. In AI culture, innovation wins: experimentation, fast feedback, small smart bets that ladder up. You don’t boil the ocean to make a cup of tea. Start bottom-up where the ROI has been showing up first anyway—individuals with access to capable models. Replace the half-day report compilation with a pipeline. Swap the brainstorm that produced three headlines with one that produces thirty and the judgment to pick two. Build a moat around your knowledge and your first-party data, not around your org chart. That moat is portable across channels and algorithm changes because it’s based on what you actually know about customers and how you show up for them.
Most importantly, acknowledge that feelings are part of the work. People get fired over feelings in this business. Leaders win by making people feel excited about the future, not cornered by it. The way through is simple, if not easy: be positive, be specific, and be accountable. If the ad that makes the money is ugly, bless it and build a better one. If the model’s answer is off, catch it and fix the prompt or the data. If approvals are the bottleneck, reduce the surface area or change the approval policy. AI isn’t a wand; it’s a lever. It multiplies whatever process you already have—clarity or chaos.
Build the Moat, Cross the Bridge
We’re alive at a wild moment. In one lifetime we’ve gone from punch cards and beepers to mobile internet, gene sequencing, and now models that can draft, design, and decide with us. No one actually knows what happens next. That’s the fun of it.
Here’s what I do know, drawn straight from the work. In advertising, let AI expand your option set and accelerate your learning, while humans choose and curate. In SEO, shrink the aperture and become unambiguously the best answer in a niche, structuring content so machines can parse it and humans can feel it. In operations, move the numbers faster and make them truer, shifting from brittle ROI to business-real MER, instrumenting redundancy, and accepting some technical debt for the sake of learning. Keep humans in the loop not out of nostalgia, but because the loop is where brand, judgment, and accountability live.
Treat AI as the catalyst for a cultural shift from delivery to innovation. Start small. Ship often. Own your data. Focus on outcomes. And remember that personalization at scale isn’t a license to remove people; it’s an invitation to understand them better. If we build our moats around knowledge and care, the bridge stops wobbling. We don’t just make it across—we bring our customers with us.
About the SoCal Tech Forum
The SoCal Tech Forum: Building a Community of Innovators
The SoCal Tech Forum was founded with a vision to create a dynamic and inclusive space for technology enthusiasts, innovators, entrepreneurs, and students. Based in the Inland Empire, a region brimming with untapped potential, the forum serves as a centralized platform where individuals can come together, gain valuable experience presenting their ideas, and learn from seasoned speakers. Our mission is to amplify technology growth and foster entrepreneurship within the Inland Empire and beyond.
With over 400 members and growing, the SoCal Tech Forum is the premier hub for tech-minded individuals from across Southern California. Our community thrives on the hunger for knowledge, collaboration, and innovation. Each month, we host engaging meetups that feature diverse topics ranging from blockchain and artificial intelligence to business strategies for startups and beyond. By adapting our format regularly, we ensure our events remain fresh, relevant, and tailored to the interests of our members.
At the SoCal Tech Forum, we value feedback from our community to guide the topics we explore. Whether you're delving into highly technical subjects or engaging with non-technical content, our events promise something valuable for everyone. Together, we aim to elevate education, spark meaningful discussions, and tackle challenges that drive the tech ecosystem forward.
