If a tool can generate copy or images in seconds, it is reasonable to ask how that changes the price. Some work becomes faster. Other tasks remain necessary or become more important.
You should be able to see where efficiency has improved and which responsibilities remain. That gives you a fair basis for assessing the work and the terms of the collaboration.
A useful pricing conversation covers the process, agreed result and commercial model. “We use AI” says very little about what you receive or who checks the finished work.
What can be accelerated
A tool can help explore directions, draft variants, transcribe, organize information or create images. How useful it is depends on the task, the instructions and the criteria we use to evaluate the result.
Generation is only one part of a campaign. The team still needs to understand the audience, choose the argument, verify the offer, preserve the product’s identity and prepare the formats required.
Some projects can be completed very quickly. Others require further references, corrections and checks. Time should be measured or estimated honestly for the specific situation.
What happens after the first generation?
Imagine a fictional campaign for a bag. We need an image, message and adaptations for two environments. An attractive result needs checking before it becomes a campaign asset.
Generation is one step. Delivery requires more checks.
Illustrative example. The conditions are explained in the article.
Information
Product, audience, objective, conditions
Generation
Versions relevant to the requirement
Selection
Selection against the project criteria
Verification
Accuracy and consistency
Preparation
Formats, destinations and implementation
| Step | Verification question |
|---|---|
| Information | Do we know the product, the audience, the objective and the conditions of the offer? |
| Generation | Do we have variants relevant to the requirement? |
| Selection | Which variant best meets the project criteria? |
| Verification | Are the product, claims, copy and details accurate? |
| Preparation | Does the material work in the formats and destinations used? |
These steps do not require a fixed number of hours. They describe responsibilities. Automation can reduce effort at several points, and the project should benefit from that efficiency in the way agreed.
What may still need doing after a successful generation
For example, the tool can quickly produce the basic image. If the plan requires an integrated text composition and a set of separate components, the team has to prepare for both uses: crops, an image without text, matching copy and checks during implementation.
Some of this can be automated. The responsibility remains to deliver what the agreed campaign needs. Explaining those requirements before quoting helps show why generating an image and delivering a usable asset can take different amounts of time.
How efficiency should affect the price
With hourly billing, the discussion concerns actual billable time and the agreed rules. Claiming hours that were not worked is not acceptable. With a fixed-price project, the basis is the deliverable, scope, responsibility and terms agreed.
Efficiency can allow a lower price, more exploration, a shorter timeline or more thorough checks. These are options to negotiate and explain. Using AI does not, on its own, call for a universal discount, nor does it justify keeping a price unexplained.
The client needs to understand what they are buying. The agency needs to explain the value of the result and the responsibility it takes on.
What remains the team’s responsibility
A tool does not automatically know every condition of the business. It can produce an unverified statement, modify a product detail or misinterpret a requirement.
The delivery team sets the appropriate checks: information, composition, consistency, functionality and terms of use. Data and materials entered into the tools are managed according to the relevant project agreements and settings.
“That is what the AI generated” does not solve the client’s problem. Review and correct the deliverable before using it.
How we discuss this at NOT MORE.
We want the tools to help the project. We explain what we include, what can be accelerated and what checks remain necessary. We do not present the number of attempts or the complexity of the tool as a substitute for the value delivered.
If a task can be solved simply and efficiently, this is useful information for the proposal. If some requirements need additional work, we describe them specifically.
Three questions about an offer with AI
- What part of the process is accelerated and how is it reflected in the project?
- What deliverable and what checks are included?
- Who is responsible if corrections are needed?
What do you need from the AI-assisted project?
Start with the deliverable, its intended use and the materials you have. We can discuss the process required and how any efficiency should be reflected in the proposal.