Three Things In-House Teams Build Before a Gen AI Budget
Most in-house creative teams started with Generative AI by getting access to a platform, learning how it works, and then seeing the results. This is the usual approach, but it often means teams quickly hit a quality limit and can't move past it.
Teams that create photos a brand can actually use did some groundwork before signing up for a platform. Teams that only produce lots of generic photos skipped this step.
Brand Guidelines Don't Direct a Generative AI Session
Most teams miss one key step: creating a standard for generative direction and production workflows when working with Gen AI. This is a written description of what high-quality Generative Photography should look like for their brand, detailed enough to help write a good prompt.
This is not the same as the brand guidelines. Guidelines describe results, like 'warm, approachable, premium,' but they don't explain the lighting, the talent model's energy, or the composition that creates those results. If you tell an AI model like Nano Banana Pro to make something 'warm and approachable,' it will give you its own version of what this looks like, which could fit any brand, or fit none. We explained what this looks like in 'Why Your AI Product Photos Look Unrealistic (and How to Fix It).'
The generative direction standard turns the brand guidelines into the language you use to write prompts. It defines your brand's lighting style, the posture and expressions that match your image, and the flaws you will not accept.
Three Things to Build Before the Next Gen AI Budget
The first thing to add is a step between the marketing brief and the AI session. Usually, teams get a brief from marketing, but they skip writing down the Generative Direction that will guide the session before anyone starts using the platform. This step doesn't have to be long, but it should be written and shared with everyone involved.
The second thing to create is a written quality standard. This defines what a finished, publishable photo looks like for your brand and what doesn't make the cut. Without this, two team members could get the same brief and platform but produce work of different quality, and no one can say which is correct.
The third thing is Generative Curation. Many teams start by just picking the best photo from what the platform gives them. Generative Curation means checking every output against the written standard, only moving forward with the photos that meet the mark, and changing the prompt if none do. With this approach, a session can produce three to five times as many usable photos as a traditional shoot, so you don't have to settle for less.
The Standard Follows the Workflow, Not the Creative Professional
The main point is that this foundation takes the knowledge out of just one person's mind and shares it with the whole team. Without it, the quality of your brand's Gen AI work depends on who is using the platform that day. With it, any team member can work to the same standard. You can repeat it and improve it.
This approach also changes how you talk about budgets. A team that can show a written direction standard and a quality check is asking for a Gen AI budget to build a real capability. A team that just generates photos and hopes for the best is only asking for green-lighting a pilot to dabble with a platform subscription.
To conclude with a comparison, no brand would send a photographer to a shoot without a brief, a lighting plan, and the judgment to reject a bad shot. A Generative AI session is just like a photo shoot and needs those same three inputs. Today's successful teams are learning those skills to become Generative Photographers.
Download the free Generative Photography Cost Breakdown to see what the reallocation looks like across production categories: genphotography.com/asset/costbreakdown
