A client can now open a design agent, type a sentence, and get a reasonable-looking first draft of their app in minutes. So the question is coming up in proposal meetings across the region: "If AI can do the screens, why does the quote still look like this?"
It's a fair question. Most pricing advice written about it targets US and European agencies billing by the hour. In MENA, and especially when selling to the Gulf, most clients expect fixed-scope quotes, procurement wants line items, and "value-based pricing" can sound like a vague excuse for a high number. Here's how I think about pricing design work when the first draft costs almost nothing.
Key takeaways
- Screens are cheap now. Decisions, research, systems and accountability are not, so price those.
- Structure quotes around decisions and outcomes, not screen counts.
- Make bilingual and RTL work a visible line item instead of hiding it in the total.
- Be open about how you use AI. Clients trust a clear process more than a quote that hides it.
The client's new question: "Can't AI do this?"
Partly, yes. AI tools can produce layouts, variations and placeholder flows faster than any designer. If a quote is built on "number of screens × days per screen", the client is right to push back, because the work being counted is the work that got cheap.
The answer isn't to defend the old math. It's to change what the quote counts.
What's cheap now, and what still takes time
Cheap: first-draft layouts, visual variations, standard flows like sign-up or settings, placeholder copy, basic prototypes.
Still expensive:
- Deciding what to build. Which features go in the MVP, which flows matter, what to cut.
- Research. Talking to real users and turning what they say into design decisions.
- Systems. A design system that developers and agents can build from without drifting.
- Edge cases and states. Errors, empty states, loading, permissions: the parts AI drafts usually skip.
- Bilingual QA. Checking that the Arabic version reads and works like a native product.
- Accountability. Someone who answers for the result when it ships.
Quote decisions and outcomes, not screens
Instead of a screen count, I'd structure a fixed-scope quote around phases that each end in a decision or deliverable the client can see:
- Discovery: user interviews, competitor review, and a prioritized feature list. Ends with: what we're building and why.
- Core flows: the three to five flows that decide whether the product works. Ends with: a tested prototype.
- Design system: tokens, components and documentation, in both languages. Ends with: a system the dev team can build from.
- Full UI and states: every screen, every state. Ends with: dev-ready files.
- Build support: design QA during development. Ends with: a shipped product that matches the design.
This still works for procurement, because each phase is a fixed line item with a clear deliverable. It just counts the right things.
Bilingual and RTL work is a line item, not a hidden cost
Many agencies treat Arabic as "the same screens, flipped" and absorb the cost. It isn't the same work. Mirroring rules, Arabic typography, mixed-direction text, dialect review and RTL QA are real effort, and they're what separates a native-feeling product from a translated one.
Put it in the quote as its own line. Clients who care about Arabic users will see why it's there. Clients who don't should know what they're giving up.
Talk about AI openly
Don't hide that you use AI tools, and don't treat it as a selling point either. Explain where it helps and where it doesn't:
We use AI to explore options and draft faster, which keeps your budget on research, decisions and quality. Every screen that ships is reviewed and owned by a designer.
This also answers the "Can't AI do this?" question before the client asks it.
A sample proposal structure
- The problem, in the client's own words
- What success looks like (two or three measurable goals)
- Phases, each with a deliverable and a fixed price
- Bilingual / RTL work as a separate line
- How we use AI in the process
- What's not included
- Timeline and review points
[Waleed: add how Breem's own proposals changed this year, or a short anonymized example.]
See it in real projects
The projects in my portfolio were scoped this way: around decisions, systems and bilingual quality rather than screen counts. For how I approach early-stage products, read how to design an MVP.
Sources: The Billable Hour's Reckoning (Metapress), AI-Era Agency Pricing Models: A 2026 Decision Guide (Digital Applied), Is Your Organization Messing Up AI Adoption? (Smashing Magazine).
FAQs
Should design agencies lower prices because of AI?
Not across the board. Prices for work that AI made cheap, like first-draft layouts, should come down. Research, decisions, design systems, bilingual quality and accountability still take the same expertise and should be priced for it.
How should I price Arabic and RTL design work?
As its own line item. RTL layout, Arabic typography, mixed-direction text and native-speaker review are separate work from the English design, and listing them shows the client what they're paying for.
Do you work on fixed-scope projects?
Yes. I scope projects in phases, each with a clear deliverable and a fixed price, which works well for clients across MENA and the Gulf.