The first kitchentop price calculator running on a 2D bin-packing algorithm — turning a manual, expert-only quote into an instant, self-serve one.
Product
Website
Domain
Construction & renovation
Client
Role
Service designer, Web developer
Activities
Metrics
Countertop prices were traditionally calculated by hand, by a trained person in a CAD environment. The goal was to cut that manual labor without limiting product specs or feature variety. Two extra complications: people shop very differently depending on their renovation stage (curious "early-bird" browsers vs. determined price-seekers), and the audience spans both B2B furnishers and B2C homeowners with very different product knowledge — so one calculator had to stay easy for everyone.
I developed my own version of a lean-cutting (2D bin-packing) algorithm as a replacement of a human stonecraft engineer. The result: a three-column layout serving both user flows, with material discovery (filters, direct search, "find similar"), countertop configuration (I/L/U shapes, parts, edge, cutouts), auto-updating price, save-to-email, and a hidden cutting-layout map for B2B clients.
Calculations went ×25 faster: median time to a price dropped to ~91 seconds vs. ~38 minutes the old messenger way. Average B2C check grew +34% (time-conscious clients who valued the clarity), and B2B sales rose ~+10%. It remains the first and only kitchentop calculator pricing in slabs via a 2D bin-packing algorithm.
Roleplay calculations with the owner; Excel low-fi prototype to reproduce the math.
Mapping the B2B purchase path and specific B2B needs.
Feature comparison across 24 competing calculators (direct and indirect competition).
Getting statistics from 78 real B2C leads (Instagram ads) on most-wanted features.
Combining external lean-cutting tools with an Excel calculator allowed to test the concept with stakeholders. Some formula adjustments were made after.
Try in GoogleSheets
1 B2C persona, 3 B2Bs + the client (stonecraftsman) persona. Each with their distinct needs and fears.
The B2C map shows that people don't trust anyone anymore at this stage of home renovation :)
Doesn't matter what persona is it, based on their renovation stage they search either starting from dimensions or from visuals.
Identified core feature set and visual style that fits the product.
I ended up with three-column layout, precisely supporting both user flows: the early-bird inspiration hunters and determined price seekers. The mobile layout was hard to deal with, but based on customer statistics I prioritized inspiration hunters flow.
Check it live
Filters
Help in discovering materials based on Brand name, Color, Pattern and Surface finish.
Direct material search
For the situation where a certain stone color needs to be found quickly.
Find similar
Allows to discover similar stone colors across the brands and find more affordable ones.
Product shape & size
Supports I, L, U shapes — which are 90% of all the orders. For more complex ones suggests to send the sketch to the company.
Add parts
Allows to add extras: kitchen island, bar counter, vertical supports, backsplashes.
Edge shape
Offers 2 thickness options and several decorative facets which impacts the price.
Cutouts and sinks
Offers a variety of cutouts for different types of appliances: undermount sinks, integrated basins etc.
Name & specs
Shows Brand name, Color name, Surface finish (in case it is non-standard). Labels mark the availability of exact stone colors, and highlight the most popular ones.
Price
Displays the final cost of the countertop based on its dimensions and features. Updates itself automatically when sizing changes.
Save the offer
Sends the relevant item card to the client's email.
Secret feature: cutting layout
Shows the precise material cutting map when clicked. Bulk clients are told about it privately, others aren't supposed to check it.
visually-driven B2C clients want live photos the business can't supply at scale (500+ stone colors).
Potential: an AR try-on for color/thickness in the client's own interior.
small orders work in the current UI; bigger ones (e.g. 6 windowsills) don't.
Potential: a windowsill product tab with adjusted inputs.
by local business design, deals close on-site only.
Potential: a conventional purchase flow once proper remote sales infrastructure exists.
