Seven before/after AI use cases in an agency environment
1. Consultation and insights compilation
After a discovery session or client consultation, we used to spend hours manually transcribing discussions (in whole or part), pulling out themes, and building a presentation to share back what we heard.
It was thorough. It was also very slow.
Now AI handles the transcription automatically, and goes a step further by surfacing suggested insights from the conversation. Our strategists review and refine those suggestions (this part still absolutely requires a human brain), but we get to spend our time on interpretation rather than logistics.
The result is a sharper, faster insights process that lets us move from conversation to strategic direction without losing momentum.
2. Audience research analysis
Audience research has always been time-intensive: compiling survey data, running crosstab analysis, cross-referencing multiple sources to build a fuller picture of who we're talking to and what drives their decisions.
AI has meaningfully changed the volume and speed of what's possible here. Automated crosstab analysis and demographic pattern recognition mean we can identify trends in data that would have taken much longer to surface manually. And because AI can consider multiple sources simultaneously, we're drawing from a wider pool of inputs without proportionally increasing the time it takes to do it.
Our planners still interpret the findings and make the strategic calls. Now they're working with more complete information, faster.
3. Performance reporting
Manual reporting used to eat a significant chunk of time that was better spent on strategy.
Building campaigns from scratch each time, researching audiences using platform suggestions and Google data alone: these were necessary processes. But they were a slog.
AI-assisted reporting has given our digital team time back. Standardized campaign conventions mean we're not reinventing the wheel with every new brief. And better audience intelligence up front means our targeting decisions are informed by more than what the platforms themselves want us to do.
More time thinking about strategy means better strategy. That’s just math.
4. Concept exploration and demonstration
Brainstorming and early concept development is where creative momentum can stall — or take off. When your visual references are limited to stock image libraries and you're manually building composites to communicate a rough idea, it slows down the creative process at exactly the moment it should be clipping along.
AI has changed the front end of our creative development in real ways. Unique image generation means we're no longer constrained by what stock libraries have on hand. AI-assisted brainstorming helps us pressure-test directions faster and get to a point of view more efficiently.
There's still significant refinement time involved once a direction is chosen — good creative doesn't skip that step, and doesn’t leave it to a machine. But we get there with more options explored and an increase in our ability to convey the exact direction we've landed on.
5. Animatics and storyboards development
This one surprises people. Developing storyboard imagery (which is a specific-enough ‘for instance’ within #6 on our list, below, that it deserves some solo attention) used to be genuinely time-intensive.
Building out visuals to communicate a concept before production even starts is a necessary step so clients can envision and approve the intended outcome before you start the (expensive) step of producing material. It also allows production partners to see our vision better and provide more accurate cost and time estimating for production — which often means a faster, smoother turnaround.
Still, creating a representative product that isn’t ‘the product’ represents a disproportionate investment of our client’s money and our team’s time. And revisions are often limited, because every change means more time/money.
AI-generated imagery has changed the economics of this stage significantly. We can produce storyboard visuals faster, iterate on them more freely, and — critically — test more creative approaches within the same budget. Clients get to see more options. We get to explore more directions. Everyone makes better decisions before the most expensive part of production begins.
6. Creative execution
Once a concept is approved and we're building final assets, AI comes back into play. Sourcing imagery used to mean hours in stock libraries, often not finding exactly what was needed and making composites, or compromises. Alternatively, capturing original photography and video adds cost and time.
AI generation and editing tools have opened up new possibilities at this stage: generating unique images, customizing stock faster, upscaling and refining imagery that might otherwise have required a reshoot.
That doesn't mean we've eliminated photography and video from our process — authentic, original creative still has real value. But it does mean we have more tools to solve production problems quickly and cost-effectively.
7. Longer-form content generation
Writing at scale is hard. It requires subject matter expertise (SME), a strong editorial process, and enough capacity to keep up with publishing cadences that would frankly overwhelm a small team without some kind of support.
AI has changed our content process by accelerating the drafting stage and making the overall workflow more manageable. Less intensive SME input upfront, faster turnaround on first drafts, and a timelier review process means content moves through the pipeline without the bottlenecks that used to push deadlines.
The writing gets reviewed, refined, and approved by humans — always — but the team is spending their time on the parts of the pipeline that require their expertise.