Adlanders Share the AI Hacks and Tips that Saved 2025

When it comes to AI in advertising, the heat and debate seems to be largely focused on the gen AI content/slop. But, quietly in the background, people have been finding smart ways to use AI-powered tools to make tedious tasks a little bit less annoying, to get to where they need to a little bit quicker. Not everyone has had the time or bandwidth to get tinkering with the platforms out there, so we figured it could be handy to pull together some tips, tricks and inspiration that you might want to steal to make 2026 that tiny bit less painful.

Connect the Dots with NotebookLM

Orli Millstein, director, content strategy, DAC
Tool: NotebookLM

My strongest AI unlock this year was building a living, briefing system with NotebookLLM. We're often juggling audience research, client decks, interview notes, and a pile of mismatched PDFs. None of it connects cleanly, and I’d lose valuable time trying to stitch it together before a pitch or presentation. NotebookLM changed that. I now feed every input into one workspace and have Notebook organize the chaos into a searchable, conversational library of client insight. I can ask it to surface patterns, strengthen a narrative direction, or translate audience tensions into clear language we’ve been able to use. It has quickly become both a pre-briefing tool and a reliable strategic partner whenever I need to shape a story with speed and clarity. As a bonus, I love being able to turn that same research into audio or video formats, which makes it easier to brief teams in different ways and explore ideas from new angles.

Hire AI Agents to Get Audience Data Quickly

Nina Matzat, general manager, Plan.Net Studios (part of Serviceplan Group)
Tool: AI Agents on Sōkosumi (sokosumi.com, via MCP integration); GWI; Claude

I needed consumer insights on AI adoption and target demographics for our marketing, but traditional research would have taken days and cost a ton in access rights and work hours.

The hack was using the Sōkosumi agent marketplace and the GWI Spark agent [an AI agent from audience research company GlobalWebIndex]. Additional superhero hack: using Sōkosumi’s MCP integration with Claude, I hired the GWI agent directly from my Claude chat – no platform switching, no workflow disruption. I simply asked for audience intelligence on AI tool usage across demographics, and the agent delivered the data from GWI's consumer database within minutes.

What would've been an expensive two-day research project collapsed into 30 minutes. I got psychographic profiles, adoption patterns, and behavioral insights that helped our marketing campaign.

Prep for Prospective Clients with Confidence

Mette Davies, new business director, true
Tool: Prospect Meeting Preparation GPT (built in ChatGPT)

As new business director at true, I spend a lot of time preparing for meetings with prospective clients. I built a simple GPT that pulls together everything I need before a first call. It helps me understand the client’s business, the people involved, the pressures they might be under, and where we could add value. It also drafts a clear agenda that I can refine and share ahead of the meeting.

The impact has been huge. What used to take a few hours now takes about 30 minutes, and I go into conversations feeling prepared, confident, and focused on the right things. It has removed the stress of starting from a blank page, making every discussion more productive. It’s not flashy, but it’s become one of the most useful tools in my week.

Use ChatGPT’s Generic ‘Insight’ to Your Advantage

Kev Chesters, CSO
Tool: ChatGPT

If you are ever briefed to write a positioning statement for a brand, then there is a very simple hack using the simplest AI tool of all: ChatGPT.

Let’s say the brief is to write an inspirational positioning statement for a brand in the jewellery sector. Simply pop onto the old ChatGPT and type, “write me an inspirational positioning statement for a jewellery brand".

Now this won’t tell you the answer for your client. It will tell you what NOT to tell them; because it’ll go out there, scrape the entire sector, and come back with the most obvious, anodyne, generic statement that it would be possible to create. It’ll be a summation of all the pap out there right now because the tragedy is that this statement will be precisely what 99% of the sector will be saying right now – including your client. It is a great way to see exactly why most positioning statements are a waste of time and money, and will show your client exactly where NOT to dig for treasure. Then you can start to get somewhere different and good.

AI Placeholder Voiceover to Stop Editors Waiting Around

Seba Morando, post-production director, DUDE Milan
Tool: ElevenLabs

In 2025, our post-production team stopped wasting time on VO scratches. Before ElevenLabs, even a simple temporary voiceover could slow down a half-day edit: someone had to record, send files, fix the timing, and re-record if a line changed. Out of 10 projects a month, this was a silent productivity killer.

With ElevenLabs, VO scratching has become instantaneous. Editors generate a clean, already synced narrative in minutes, try different tones and intensities, and present montages that ‘breathe’ already like finished pieces. This has released a huge amount of mental load: less waiting, less stress, more creativity. It has become one of the pillars of our workflow and a real improvement in the team's quality of life.

Generative Fill to Avoid Reshoots

Amy Eakin, editor, After Avenue Post
Tools: ElevenLabs, Adobe Podcast, Adobe Firefly/Photoshop Generative Fill

In post production, 2025 was the year AI stopped being a novelty and actually made our lives easier. ElevenLabs has been a game changer for us – we can create scratch VO that matches tone, pacing, and personality so clients immediately understand the feel of an edit and cut down on feedback.

Adobe Podcast has also saved us countless hours by cleaning production audio enough that early cuts feel polished and client-ready.

Visually, Adobe’s Generative Fill has helped us extend our studio cyc [cyclorama] and solve setup limitations without reshoots. And when we needed quick approval on a look, we used an AI image tool to generate a cleaned-up frame straight from a screenshot, good enough to get the green light before heading into colour. AI hasn’t replaced our craft; it’s just cleared the clutter so we can actually focus on the creative.

Fix Wrecked Audio, Kill Time-Sinks and Deliver Clean Work Under Pressure

Jay Nelson, editor, Cut+Run
Tool: DaVinci Resolve

The AI baked into DaVinci Resolve stopped being a gimmick long ago. Beat detection, dialogue silence removal, and audio clean-up routinely save me a day when editing a :30. AI transcription, edit-blending, and music splitting once required five separate applications; now that’s just another Tuesday morning. The moment DaVinci’s AI became essential was an interview with wrecked audio. I analyzed the disaster and regenerated a clean, exact performance that saved the project. Then I automated 16 channels of various sounds into a rough mix better than anything I had built manually, and we all made it home for dinner.

It’s not cheating. It’s evolution. And it’s fun. The industry is overcrowded with capable individuals, but the bar for expertise is higher now; efficiency and resourcefulness are integral to the craft. Those who embrace innovation will lead the way. How many of us used AI to hit 150 words? [Editor’s Note: you got me.]

Overcome the Brain Fog of the Blank Page

Valerio Rossi, innovation producer, UNIT9
Tool: a custom multi-agent workspace (Creator / Technician / Critic / Producer)

2025 was the year I stopped treating AI as a clever assistant and started using it as a thinking partner. I built a multi-agent workspace where four AI roles help me stress-test ideas fast: one pushes creativity, one checks technical feasibility, one challenges assumptions, and one drives execution.

The problem it solved? Early-stage fog. Projects used to start with vague briefs, exploratory meetings, and slow alignment. Now I arrive with options, risks mapped, and clearer direction. It hasn’t replaced collaboration – it’s made it sharper. Meetings are shorter, decisions are faster, and teams spend more time elevating ideas instead of defining them.

The concrete outcome: I delivered projects with more clarity, fewer false starts, and far less mental exhaustion. AI accelerated my process and freed more space for people to do their best work.

Test Your Prompt on a Bunch of Platforms at Once with Weavy

Akira Thompson, group creative director, Publicis Production
Tool: Weavy

I use Weavy AI as a central hub to test multiple image and video generators in one go. Instead of bouncing between different platforms, I build a quick node-based network where a single text prompt fans out to a dozen AI models at once: Runway, Veo, Firefly, etc. With one click, the same prompt is batch processed across every connected model, so I can compare outputs side by side and quickly choose the strongest result.

It removes the repetitive task of manually pasting the same prompt into separate tools, dealing with different interfaces, and logging in and out of multiple accounts. Weavy turns that whole multi-platform workflow into a single streamlined step, saving time and giving me a broader creative overview of what each model can do with the exact same brief.

Vibe Coding with Google Gemini to Create Handy Mini Apps

Tim Schatz, VP, data and innovation, H/L
Tool: Google Gemini

The Problem: We had a massive operational divide. Our analytics and engineering teams held the keys to automation while most other teams were stuck doing mundane data tasks by hand, or relying on Engineering. We wanted to build custom tools to fix this but simply didn't have the technical headcount to support every single micro-need across the agency.

The Hack: We turned our non-technical staff into "vibe coders" using Google Gemini. Since it sits securely inside our existing Google ecosystem, we skipped the IT red tape. Staff now use natural language prompts to generate working Python and App Scripts for Google Sheets without writing a single line of code themselves.

The Outcome: It has been a massive morale boost. Teams are building their own mini-apps to automate things like pacing checks, keyword research, and performance alerts. We essentially decentralized engineering and let the people closest to the work build the exact tools they need to survive the week.

Combine Chat GPT and Figma for Slick Interactive Prezzos in Minutes

Natalie Levy-Costa, UX lead, Left Field Labs
Tool: ChatGPT; Figma

I used ChatGPT to craft a prompt system that builds interactive presentations directly from my Figma working files. Instead of creating separate decks, I can now take designs straight from ideation into presentation, using a Mad Libs–style “fill in the blank” prompts to generate unique review experiences. No more juggling slide links or massive files, Figma Make lets me publish a clean, password-protected link to the final output.

Getting this workflow off the ground took some experimentation, but once I started thinking systemically, everything clicked. Now it’s a time-saver that frees designers from repetitive deck building and makes space for what actually matters: sharpening concepts and improving the work. Knowing Figma Make can assemble a presentation in minutes means I can stay focused, present more seamlessly and preserve a little sanity in the process.

Let AI Orchestrate So You Can Concentrate

Dan Huynh, associate director, product strategy, Kettle

AI Tool: Roo Code x VS Code

Product management work traditionally lives across Confluence, Jira, GitHub, and Slack. I was constantly pulling inputs from each system, translating stakeholder notes into PRDs [product requirement documents], rewriting specs into Jira stories, and wading through dense PR threads as a non-engineer. The context switching was relentless, and none of the work lived in one place I actually controlled.

Roo Code [an open-source coding agent] flipped that model. By plugging an MCP-powered AI agent into VS Code [Microsoft’s open-source development environment], my code editor became a true PM command centre. I keep all my documents in local Markdown files such as PRDs, meeting notes, and raw requirements. The agent reaches out to Confluence, Jira, and GitHub to fetch data, summarise conversations, publish updates, draft stories, and rewrite specs. Everything is unified, traceable, and instantly regenerable from my local workspace. The impact was significant: double-digit hours saved, much lower cognitive load, and sharper prioritisation. The AI handles the orchestration so I can focus on the product.

Vibe Coding Assistants to Piece Together the Info You Need for Clearer Starting Points

Daniel Bueckman, paid media & performance marketing lead, Intuit
Tool: AI-assisted coding tools

This year, we tapped into the benefits of AI-assisted coding tools, including vibe coding – no-code prototyping with plain language – by building custom AI assistants that could turn scattered knowledge from multiple decks, tools, and past analyses into something teams could act on quickly. It cut down the friction of piecing together performance context and gave us clearer starting points for decisions across media placements, audiences, and creative. Speed wasn't the only positive, either. With clearer, more confident starting points, our teams had more mental space to focus on the human side of the work: the strategy, ideas, and judgment calls that actually move a campaign forward.