AI is changing how marketing gets done, not by replacing creative judgment, but by clearing space for it.
Across our network, we asked creative leaders how AI has reshaped their operations for this article. The pattern is clear. Drafts appear faster, research compresses, and production scales. The edge still sits with judgment, taste and timing. In short, AI does the doing. People do the deciding.
Modern marketing sits at the junction of brand, performance and operations. A brand creates long-term demand. Performance turns current demand into customers. Operations is the system that makes both work at speed and quality. AI has moved from a toy to a core infrastructure inside that system. It drafts, clusters, routes and watches. Automation then ships, schedules and reports.
The trick is balance. If you spend all your AI gains on more low-value output, you miss the point. Use the time to improve creative quality and brand consistency. Let the machines help you reach more people with work that is actually worth their attention.
Across teams, the same pattern shows up. AI speeds research, testing and basic production. That means faster cycles and more iterations with less waste. Microsoft’s Copilot research backs this up. Teams get to a first draft sooner, then spend more time refining and using judgment, not typing in silence (Microsoft WorkLab, 2024).
In practical terms, this looks like rapid copy variants, automated reporting, smart clustering of customer feedback, intent signals that trigger timely outreach, and one-click asset resizing. For content, it can transcribe workshops, summarise interviews, and serve up an outline that a strategist can hone. For paid media, it can suggest audiences, build drafts of creative, and push clean naming conventions into your ad accounts.
For brand teams, AI is now a decent creative thinking partner at the messy start. It can push language and tone, visual routes, and sample audience reactions so people can debate with something on the table.
When we are creating new brands, or repositioning existing ones, we use AI to create audience profiles to test strategic territories, to explore richer language and tone for rationales and positioning statements, and to demonstrate key visuals through image creation. That gives creatives and strategists more time to craft, interrogate and optimise their thinking.
It is not about doing everything with AI. It is about clearing the runway for judgment, taste and timing. That is where brands win.
AI is useful because it removes the grind and gives you time to think. Research runs faster. First drafts appear quickly. Formats and small variations come together with less effort. None of this is the creative edge. The edge is judgment, timing, and taste. That still sits with people.
Laura Weldon, Founder of Studio.LWD, has taken a similar line. Her team uses AI for the repetitive parts so they can concentrate on ideas and problem-solving.
By taking care of the repetitive stuff, such as admin, data analysis and research, AI gives our team the time and space to focus on the thinking, the problem solving and the ideas that really make a difference. It is another tool in the kit, not a replacement.
This is where creativity lifts. AI turns stills into motion so stories travel further. It tidies a messy transcript into a clean narrative. It drafts options so you can pick a direction, not stare at a blank page. The craft remains human. The throughput increases. The quality should too, if you use the time well.
Personalisation is moving from tokens to timing. The goal is to make contact when you have a clear reason. Say why this person, why now, and how you can help. That is how trust grows and how brands feel useful. McKinsey’s personalisation study found that 71 percent of consumers expect tailored experiences, and top performers drive materially higher revenue from getting it right (McKinsey, 2021).
AI helps here in two ways. First, it can scan signals to check fit before you reach out. Second, it can assemble messages grounded in real context rather than guesswork. Policy and platform changes also push the market towards quality. High-volume spray and pray will keep getting filtered or flagged. Useful, authenticated messages will land.
I want messages that are specific to a person or an account, based on signals that show what they are trying to do right now. Before: 1,000 cold emails with a 0.8 percent reply. After: 120 targeted notes tied to recent behaviour, 7.5 percent reply, 3 qualified calls. Over time that discipline lowers volume and raises quality.
Relevance comes first. Then you can widen horizons without being creepy. Show something new if it helps. Keep the frequency sensible. Ask yourself the simple test Anshul uses. If you received this message, would you find it useful? If not, fix it or do not send it.
The old split still guides many teams. Roughly 60% of investment goes into long-term brand building and 40% into activation. AI does not scrap that. It changes the economics. You can now produce and test more for the same budget. You can also waste more if you let volume trump thinking.
Use AI to enhance both sides. For brand, generate and evaluate routes faster, then put human craft into distinctive assets, codes and story. For performance, use signals to trigger timely journeys, keep copy fresh, and maintain channel hygiene without heroics. Creative quality remains the biggest driver of returns. Let the machines boost your hit rate by speeding iteration, but keep humans on the idea.
Conversion rate optimisation sits at the centre as a multiplier. Cleaner journeys and better feedback loops help everything work harder. AI can spot friction, draft microcopy, and suggest next steps. People make the call on trade-offs and tone so the experience stays on brand.
AI adoption often stalls because teams do not know where to begin. Tool overload is real. The fix is to start small and real. Pick one journey or task that matters. Build a thin version with clear guard rails. Measure it. Improve it. Then add the next piece.
Teams get stuck for normal reasons. Fear of change. Tool overload. Not knowing where to start. I prefer small, real exercises. Pick one journey that matters, build a thin version that runs with real product and price, and measure it. When people see their own work move faster and perform better, the anxiety drops and the quality goes up.
James applies the same pattern in brand creation. Quick audience profiles to test strategic territories. Draft language to sharpen a position. Sample visuals to provoke debate. These prototypes turn meetings into decisions.
Laura’s team keeps a rolling test and learn rhythm. Some tools stick. Others do not. The point is not to chase every new thing. It is to find what genuinely helps your workflow. Microsoft’s field studies suggest the biggest gains land when teams embed AI into daily habits rather than run it as a side experiment (Microsoft WorkLab, 2024).
Guard rails matter too. Define brand tone, claims you can make, distinctive assets, and lines you will not cross. Keep a small library of prompts that reflect those rules. Keep human checks on headlines, key visuals and claims. Automate the rest.
I write down what good looks like for the brand. Tone, claims we can make, how we use distinctive assets, and what we do not touch. I keep human checks on headlines, key visuals, and claims. I am happy to automate the rest as long as these touch points are reviewed properly.
Three shifts to watch. First, one-to-one messages will feel normal across more journeys. The tools will feel more conversational for creative direction. Production will keep speeding up. That creates space for concept and craft.
Second, discovery is moving from search pages to answer engines. Generative Engine Optimisation will sit alongside SEO. Your brand needs clean data, clear language, and authoritative assets that AI systems can cite with confidence. The same applies to retail and marketplace surfacing.
Third, embedded partnerships will matter more than ever. Agencies and clients will co own prompts, workflows and quality controls. Live data from the client and standards from the agency. Quarterly reviews that tune the system based on real output, not deck theatre.
Partnerships work best when the client and the agency co own the system. We maintain prompts, workflows, and quality control. They bring live data and feedback. Every quarter we look at output against brand standards and commercial results, and we tune the system.
Plenty of large brands are already moving. Klarna’s AI assistant has shifted service and commerce behaviours. Mars has trialled AI to pre-test creative and speed up content for distinct brands. JPMorgan has used AI to spot signals in markets and inform how and when it communicates. The detail varies by sector, yet the pattern is the same. Use AI to compress the distance from idea to execution. Keep humans on meaning and trust.
What excites me most is where AI is heading next. We’re starting to see tools that can help predict how people might feel about a campaign before it launches. That emotional intelligence is where it gets interesting. The goal isn’t to make creativity robotic, but to make it more informed, more human, and more impactful.
It is not abstract. Anshul gave us a live example. A new inquiry came in at 7am. A call followed half an hour later. The client needed imagery back the same day for a committee. A traditional CGI route would have taken one to two weeks. An AI-led approach made the turnaround possible. That is opportunity unlocked by speed.
On strategy, he often speaks ideas into a recorder, gets a transcript, and asks AI to group themes and remove repeats. First draft time for strategy notes falls from two hours to about 20 minutes. The thinking stays his. The typing does not.
AI is an enabler. It removes busywork so marketers can lift their heads and make better choices. The wins show up in two places. Faster operations that let you test and learn. Stronger brand work because you invest your time where humans add the most value. That is judgment, taste and story.
Our view at Door4 is simple. Treat AI like a fast junior with perfect recall. Set standards. Build small, then scale. Pair the brain that drafts, clusters and chooses with the engine that moves files and ships. Keep people in control of claims, concepts and craft. Do that, and AI will not squeeze creativity. It will give it room to breathe.
Our latest articles, features and ramblings.
We explore performance marketing, AI, communications and optimisation.
