Pricing AI-augmented work
AI made the render ten times faster. The trap is letting that ten-times collapse straight into your fee. Speed is yours to keep - value is what you sell.
She delivered six concept options in a day. Then she billed for one - and wondered why she was poorer than ever.
A Bengaluru interior designer adopted AI mood boards and started turning around concept directions in an afternoon that used to take a week. Her clients were delighted. Her bank balance wasn't. Because she still priced on _hours spent_, and AI had crushed the hours, she had quietly cut her own fee by eighty percent for delivering more value than before. The client got six options instead of one and paid less for them. That's not efficiency. That's giving your margin away with a smile. The fix isn't to slow down. It's to stop selling time.
Stop billing for the hour. Start billing for the outcome.
If you bill time and AI cuts the time, you've handed the client your productivity gain
Here is the pricing trap in one line: AI reduces your effort, but it does not reduce the client's value. The client who needed six convincing concept directions to choose between still needs them, still benefits from them, and is arguably better served when you deliver more options, faster.
If your fee is `hours x rate`, every hour AI saves is a rupee you remove from your own invoice while delivering the same or better result. You've turned a productivity revolution into a price cut for the client and a pay cut for yourself. Faster delivery is a gift to your business - more projects per month, better margins, a happier client - but only if your pricing model can hold onto the gain instead of leaking it.
The answer is not to fake slowness or pad timesheets. It's to price what you're actually selling: the outcome, the judgement, the options, the certainty - none of which got cheaper just because the render got faster.
The client never bought your hours. They bought a building they trust and options they can choose between. AI didn't make those worth less.
Sell outcomes and packages, not a clock
Value-based pricing ties your fee to the worth of the result, not the time it took. In Indian practice, architects already do a version of this with percentage-of-construction-cost fees and interior designers with fixed per-room or per-square-foot packages - both decoupled from your hours. AI makes that decoupling more important, not less: the faster you work, the more the time-based model punishes you, and the more a value or fixed-fee model rewards you.
The sharpest move AI unlocks is packaging speed as a product. Because you can now generate concept directions in hours, you can sell a tiered concept package: a Standard tier (one refined direction), a Plus tier (three directions plus a revision round), a Premium tier (six directions, mood boards, a walkthrough render). The client chooses how much exploration to buy. You've turned AI's option-generating superpower into a menu - and clients routinely pay more for choice and speed, the two things AI gives you in abundance.
Notice you're not hiding the AI. You're charging for the outcome it makes possible: more options, sooner, with your judgement curating which ones are worth showing.
Keep most of the gain, share a little, and never race competitors to the bottom
You don't have to keep every rupee AI saves - and trying to looks greedy and invites undercutting. The smart play is a split: keep the majority of the efficiency gain as the reward for the investment and skill that made it possible, and pass a visible slice to the client as competitive advantage - a faster turnaround, an extra concept option, a sharper price than the studio still doing it the slow way.
What you must not do is let AT race you to the floor. If everyone in your city drops fees because 'AI made it easy', the whole profession devalues its judgement - and judgement is the one thing the plausibility machine can't supply. The studios that thrive frame it honestly: 'AI lets us explore more for you, faster - here's the package', not 'AI made it cheap, so we're the lowest bidder'. The first sells value. The second sells your margin.
For context: Indian architectural fees commonly sit in single-digit percentages of construction cost, and interior fees as fixed packages or per-square-foot rates. AI doesn't change those benchmarks - it changes how profitably you can deliver against them, if your model captures the gain.
Your percentage-of-construction-cost fee already decouples income from hours - lean into it. AI's time savings flow straight to margin and capacity: more projects, better feasibility studies, richer optioneering at no extra fee cost. Where you do bill stages or hourly (feasibility, consultations), shift to fixed-scope packages so AI's speed becomes profit, not a discount. And price the _judgement_ explicitly: code verification, structural review and the stamped drawing are professional deliverables AI never touches, and they anchor your fee no matter how fast the renders come.
Move off per-hour and onto packaged concept tiers - this is where AI pays you back hardest. Sell a Standard / Plus / Premium concept menu built on AI mood boards and recolours, where the higher tiers buy more directions and revision rounds. Clients love choosing how much exploration to fund, and you stop giving away six options for the price of one. Keep your sourcing, FF&E and site work priced separately and firmly human - that's the buildable value AI can't deliver, and it protects your floor.
As a solo studio, AI is your leverage to compete with bigger firms on output - but only if you don't price like you're desperate. Resist the instinct to win work by being cheapest because 'AI made it quick'. Instead, offer more value: more concept directions, faster turnaround, a tighter package than a firm three times your size. Set a fixed concept-package price, deliver visibly more for it, and keep the efficiency gain to fund your own time and learning. Racing to the bottom with AI just means working harder for less.
Studio Matrx Moodboards (Style Explorer)
Concept-option generator you can package
Apartment and villa moods across rooms and styles - the engine behind a tiered concept package, since you can produce many client-ready directions fast. Limitation: it's for exploration and alignment, not a dimensioned or costed deliverable you can bill as documentation.
Veras (EvolveLAB / Chaos)
BIM-integrated render speed
Renders off your real geometry inside Revit/SketchUp/Rhino, collapsing viz time so more visualisation fits inside a fixed fee. Limitation: render speed is a margin and capacity gain, not a reason to discount - the fee should track the building's value, not the minutes saved.
ChatGPT (OpenAI, GPT-5 era)
Proposal + rate-card drafting
Strong at tabular output - draft tiered fee schedules, scope tables and proposal language fast. Limitation: it will invent benchmark fee figures and 'standard' rates that sound authoritative; verify every number against your real costs and local market, never trust its pricing 'facts'.
“AI makes design work faster, so it's only fair - and good for winning clients - to charge proportionally less for it.”
Faster effort is not less value. The client needs the same outcome - convincing options, a buildable design, your judgement - and AI delivering it sooner doesn't shrink its worth; it often grows it. Billing proportionally less hands your entire productivity gain to the client and pay-cuts yourself for being better. Price the outcome (value-based or fixed packages), keep most of the efficiency gain, pass on a competitive slice - never race the whole profession to the floor.
Workshop — build a value-based concept package rate card
You'll convert one service you currently bill by the hour into a tiered, value-based package that captures AI's speed instead of leaking it. By the end you'll have a three-tier rate card you can put in your next proposal.
A blank doc and the template below. Bring one real service you currently bill hourly (e.g. interior concept design, or feasibility studies).
CONCEPT PACKAGE RATE CARD (value-based) SERVICE: ____________________ OLD MODEL: hours ____ x rate Rs ____ /hr = Rs ________ (what AI just quietly cut to a fraction) NEW MODEL - sell the outcome, in three tiers: STANDARD Rs ______ 1 refined direction, 1 revision PLUS Rs ______ 3 directions, 2 revisions, mood board PREMIUM Rs ______ 6 directions, walkthrough, full board set WHAT STAYS HUMAN + PRICED SEPARATELY: sourcing / FF&E / dimensioned docs / site MARGIN SPLIT: keep ____ % pass on ____ % (as speed / extra options)
- 1Write down the old hourly maths for one real service, honestly - including how few hours AI now takes. Stare at how much fee you've been silently deleting.
- 2Define three value tiers by outcome, not hours: how many concept directions, revisions and deliverables each buys. Use the Standard / Plus / Premium frame.
- 3Price each tier on the value to the client and your market position - anchor on the worth of the choices delivered, not the minutes spent.
- 4Separate out the human-only, buildable work (sourcing, FF&E, dimensioned docs, site) as its own line that AI never discounts - this is your floor.
- 5Set your margin split: decide what share of the efficiency gain you keep and what visible slice (faster turnaround, an extra option) you pass on as competitive edge.
- 6Pressure-test the card against one real client: would they happily pay the Plus tier for three curated directions in two days? If yes, the value framing works.
You’ll walk away with
A three-tier, value-based concept rate card for one real service - pricing the outcome AI makes possible, ring-fencing the human-only work, and an explicit margin split you can drop into your next proposal.
A quick reframe, if you have five minutes.
- 01Take your last AI-assisted project and calculate what you'd have billed on the old hourly model versus a fixed-package model. The gap is the margin your pricing was leaking.
- 02Ask ChatGPT to draft a tiered proposal table for your service - then delete every fee figure it invents and replace them with numbers anchored in your real costs and market.
AI shrinks your effort, not the client's value - so never let faster collapse into cheaper. Shift from hourly to value-based or packaged pricing, sell AI's option-generating speed as a tiered product, and keep most of the efficiency gain while passing a visible slice on as competitive edge. Price the judgement and the buildable work that AI can never deliver, and you protect both your margin and your profession.
Billing time means AI's speed becomes the client's discount and your pay cut. Move to value-based or fixed packages; sell concept directions in Standard/Plus/Premium tiers; ring-fence human-only buildable work as your floor; keep most of the gain, pass a slice on, and never race the market to the bottom.
Should I charge less for design work because AI makes it faster?
No. The client's outcome - convincing options, a buildable design, your judgement - is worth the same or more when delivered sooner. Billing proportionally less just hands your productivity gain to the client and pay-cuts you for being better. Move from hourly to value-based or fixed-package pricing so faster delivery becomes margin and capacity, then pass on a visible slice (speed, extra options) as competitive edge rather than a blanket discount.
How do I price AI-generated concept options?
Package them as a tiered product rather than billing the (now tiny) hours. A Standard tier might buy one refined direction, Plus three directions plus revisions, Premium six directions with a walkthrough. Clients pay for choice and speed - exactly what AI gives you - and you stop delivering six options for the price of one. Keep sourcing, FF&E and dimensioned documentation priced separately as human-only work.
What are typical design fees in India, and does AI change them?
Indian architectural fees commonly sit at single-digit percentages of construction cost; interior fees are usually fixed packages or per-square-foot rates. AI doesn't move those benchmarks - it changes how profitably you deliver against them. A percentage or fixed-package model already decouples your income from hours, so AI's time savings flow to margin and capacity instead of leaking out as a discount the way an hourly model would.
Pricing the value of AI-augmented work assumes your team can actually deliver it - and verify it. Next: building the AI skills, and the new roles, that make the speed real without letting craft atrophy.