your client brought ai to the chat.

now what?

your client brought ai to the chat.

now

what?

Pick any of the examples below for tips on how to handle the conversation.

  • Can you just build this?

    They sent you an AI mockup and assume the hard part is already done.

  • ChatGPT said this would take five minutes.

    They arrived with an estimate without understanding the real work required.

  • Here is the code AI gave me.

    It worked in their preview. Now they want you to put it on the site.

  • Here is some feedback.

    They ran your work through AI and the revision list just got a whole lot longer.

  • AI said Squarespace can do this.

    A chatbot already gave them a very confident answer without key details that make it difficult to create.

  • Shouldn’t this cost less now?

    AI made part of the work faster. They assume the whole job should cost less.

Start here:

1. Find what they actually want.
Pinpoint the layout, interaction, or idea they’re excited about.

2. Separate the picture from the build.
A mockup shows the idea. It doesn’t explain how it works.

3. Find the hidden work.
Check Squarespace features, custom code, integrations, mobile, and future editing.

4. Reset the expectation.
Focus on the best version to build—not an exact visual copy.

5. Give them a next step.
Assess the implementation, then come back with the approach you recommend.

Here’s what a response can look like:

Thanks for sending this over! I can see what you like about this direction, especially the way the layout and interaction are working together.

Before I promise an exact build, I want to break the mockup into the pieces it would actually require inside Squarespace. Some of it may be straightforward, while other parts could need custom code or another tool behind the scenes.

I’ll take a look at the best way to create the same experience on your actual site and come back with the approach I recommend, including anything that would affect the scope of your project.

Start here:

1. Don’t argue with the estimate.
AI may genuinely generate a first attempt in five minutes.

2. Separate generation from delivery.
Creating output quickly isn’t the same as completing the project.

3. Identify what it left out.
Consider integration, responsive behavior, testing, cleanup, and revisions.

4. Reset around scope, not minutes.
Price the work you’re responsible for delivering—not typing speed.

5. Give them a next step.
Review the actual setup, then provide your scope and price.

Here’s what your response can look like:

Thanks for sending this over! AI may be right that generating a first version is pretty quick.

Before I estimate it, I want to look at how the feature needs to work inside your existing Squarespace site. That can include responsive behavior, integration with the current design, testing, and any adjustments the generated solution needs.

I’ll review the actual setup and come back to you with the approach I recommend, along with the scope and price.

Start here:

1. Treat it like useful input.
Don’t immediately dismiss it. The code may be a useful starting point.

2. Separate working from appropriate.
Code can function correctly and still be wrong for this site. Find what part of this (if any) can adds value to the project.

3. Review what it actually does.
Check selectors, conflicts, mobile behavior, complexity, and future maintenance.

4. Reset “paste” to “evaluate.”
You might use it, simplify it, replace it, or skip it.

5. Give them a next step.
Test the code first, then recommend the best implementation.

Here’s what a response can look like:

Absolutely. Send it my way and I’ll take a look.

AI-generated code can be a useful starting point, but before I add it to the site I want to check what it’s actually changing, how it behaves on mobile, and whether it conflicts with anything Squarespace already handles or with code that’s already installed.

If it’s a good fit, I can use it. If there’s a cleaner or more reliable way to get the same result, I’ll recommend that instead.

Start here:

1. Don’t dismiss the feedback.
Your client was trying to evaluate the work and give you useful direction. Start there instead of making the conversation about whether they should have used AI.

2. Separate more feedback from better feedback.
Not every suggestion is likely to solve a real problem or supports the goals of the project, but some might so don’t dismiss them all at once.

3. Find the concern underneath the suggestions.
Look for the few things your client actually wants to change. Those concerns matter more than twenty AI-generated prescriptions for fixing them.

4. Reset what counts as a revision.
Some suggestions may introduce new layouts, features, content, or strategy that meaningfully change the scope. A longer revision list should never automatically become part of the original project.

5. Give them a clearer way forward.
Pull out the changes you recommend addressing, flag anything that would expand the scope, and confirm the direction before you start revising.

Here’s what that can look like:

I went through the feedback, and there are a few concerns in here that I’d like to address.

Before I implement all of the feedback list item by item, I want to separate the underlying changes you actually want from some of the additional suggestions AI generated around them. A few are straightforward revisions, while others would introduce a different layout or expand what we originally planned.

I’ll pull out the changes I recommend making, flag anything that would affect the scope, and send that back to you so we can confirm the direction before I start revising.

Start here:

1. Don’t turn it into a debate.
The AI may be wrong, right, or missing important context.

2. Separate possible from appropriate.
Squarespace doing something doesn’t mean that’s how you should build it.

3. Review the actual project.
Check the site, plan, integrations, mobile needs, and maintenance.

4. Ask the better question.
Move from “Can Squarespace do this?” to “How should this work?”

5. Give them a next step.
Recommend built-in features, custom code, or another service based on fit.

Here’s what your reply can look like:

It may be possible, and I’m happy to look at the option AI suggested.

The part I need to check is how that feature would actually work on your specific Squarespace site. Depending on the setup, the best solution might be something already built into Squarespace, a small amount of custom code, or a separate service that handles the job more reliably.

I’ll review the actual requirement and recommend the route that makes the most sense for your site setup.

Start here:

1. Acknowledge what changed.
Yes, AI may genuinely make part of your process faster.

2. Separate speed from value.
Your client is paying for judgment, implementation, testing, and responsibility.

3. Identify what actually disappeared.
Did AI remove meaningful work—or just speed up one step?

4. Reset around the deliverable.
Price the scope and responsibility, not minutes at the keyboard.

5. Give them a next step.
Clearly explain what’s included, what you’re delivering, and what it costs.

Here’s what your reply can look like:

AI definitely helps me move faster in parts of the process, and I use that when it makes sense.

My pricing is based on the work I’m responsible for delivering, which goes way beyond generating a first draft.

My skills are in evaluating the solution, adapting it to your site, testing it, handling revisions, and making sure the finished result works the way it should. For this project, the scope includes [deliverables], and the price for that work is [price].

*Submitting this form will not add you to my email list, but I will ask if you want to join. You can unsubscribe at anytime. No spam. No slop.

Have a wildcard situation not covered above?