Sharing information about how you can use AI to help formulate and workshop ideas for your designs
Please check out our article "Does BE Ultimate use AI in the design process?" before reading this.
When chatting designs with your team, you might be tempted to use AI tools to assist you. Here's some advice for how you can make the most out of that and avoid common mistakes that we've seen.
What you can use AI for:
- Early visualization / proof of concept: If you have a rough idea, or you’re communicating to the design team what you’re thinking, use AI as a quick way to generate visual drafts or mood-boards. For example: “Here’s the copy, here’s the concept, show me three visual directions with different moods.” This helps give our designers something concrete to react to rather, than starting from a blank page.
- Exploration of styles: Use AI to surface alternative aesthetics, colour treatments, layout variants, or unexpected approaches you might not have considered. This is especially valuable if you’re sharing the results with people outside the process and want to show, rather than just tell, what you are considering
- Elevating shared assets: If you have images/visuals that need refinement (for example: background removal, upsizing, cleaning, variant generation), AI can help improve the asset you’re sending to design. That means the design team receives higher-quality inputs, and we all save time.
- Faster iteration: Because AI generates many variants quickly, you can use it to compare “what if” scenarios: different layouts, different fonts styles, colour palettes, visual moods, especially when you’re still exploring and not sure what final you want.
What AI cannot (yet) do, and how you should approach that:
- Production-ready design: AI-generated visuals are rarely ready to print! They often require adjustment: correct resolution, consistent typography, brand alignment, accessibility, final layout, file structuring for production, etc. We should treat AI outputs as a starting point, not the final version.
- Human judgement and brand fit: AI doesn’t have the lived experience, brand context, strategic thinking, or nuanced design taste that your team might have. So while it might generate something visually interesting, it may be misaligned with your team.
- Creative ownership and rights clarity: When you use AI-generated visuals, we must be mindful of licensing, attribution, model training data, and ensure the output meets our standards (both legally, and ethically!).
- End-to-end automation: Don’t assume that because you used AI, the project is fully done. There still is a design handoff, review, revision, and sign-off process. AI accelerates the early stages, but the design team still plays a major role in completing and refining the final deliverable.
Best-practice workflow suggestion
If you are going to lean on AI during the design process, here is a sample of how that might look:
- You have an idea for a new design, and want to get some thoughts from the rest of the leadership team. You use an AI tool to generate 2-3 visual directions (with prompts like: “modern, bold, high-energy with abstract shapes and blue/teal palette”).
- You folks select your preferred direction and send both the prompt + generated visuals to your sales rep here at BE Ultimate, adding any extra context that we might need to know
- BE Ultimate using your inputs to put together an idea, fully fleshed out and rendered on our templates.
- You take this idea back to the leadership team and they provide you with a bunch of feedback: for some feedback (eg. "how would this look like in black, instead of teal?") you can use simple prompts to create quick visuals, helping decide if ideas are good and worth sharing back to us, or not worth exploring further.
Some tips we've found to help make this process as smooth as possible:
- Be specific in your prompt: style references, mood words, colour palette, output type (banner, social tile, illustration) help guide better AI results
- Use AI generated visuals as a communication tool: they help you clarify what you mean rather than leaving things vague.
- Always pair the AI output with a design brief and context: audience, application, brand rules, where it will appear.
- Don’t skip the hand-off to design: even if the AI result looks great, our design team still needs to validate, refine and make it production ready.
- Time-box the AI exploration: it’s easy to stay in “generate variants” mode far longer than necessary; treat the AI part as exploration, not finalization.
Why this approach matters for us
BE Ultimate values both innovation and brand excellence. Using AI in the workflow lets us move faster, explore more options, and elevate quality earlier, all of which help us deliver stronger outcomes. However, our commitment to brand integrity, curated design, consistent execution, and thoughtful human oversight means the design team remains central. The goal is human + AI collaborating, not AI substituting human creativity.