Our published rules
Ranked by fit, never by who pays.
Home Hunt uses a deterministic matching function, not AI. The same answers and seed produce the same order. Payment is not an input, and a founding flag is not a boost.
What we ask
The eight screens cover the side of your move, broad areas, property type, bedrooms and must-haves, budget range where relevant, timing, preferred language, availability, seven lifestyle sliders, two home-style sliders and optional favourites. Sellers and landlords describe their property instead of a search. Agents complete a longer profile with the services, areas, experience, price or rent bands, availability and the same style questions.
We leave financing, down payment, exact address, income, household details and whether you have already signed with someone for a later conversation. We never ask or score protected grounds: ethnicity, religion, country of origin, marital or family status, age, sex, orientation or disability. Preferred language is the only culture-related preference.
Filters, then sections
Before scoring, we keep profiles with a shared service, an overlapping area and an accepting status. A selected region includes the places inside it. We then calculate each section from answers that both people supplied.
| Section | Weight | How it works |
|---|---|---|
| Lifestyle sliders | 25% | 100 minus the distance between positions. |
| Favourites | 10% | Shared pick in each answered row. |
| Experience fit | 30% | Property types, situations and matching price or rent band. |
| Home style | 20% | The two home-style sliders. |
| Availability | 15% | Any shared time window. |
If favourites are skipped, lifestyle sliders become 35%. For every slider, both choosing Neither scores 100; only one choosing Neither scores 50. If you choose a preferred language other than English, language counts for 20% of the score and all other sections scale to 80%; an agent either speaks that language (100) or does not (0).
Shortlists and rotation
We show three to six profiles. Agents within three score points rotate in a seeded order, so a near-tie is not permanently decided by a tiny difference. If fewer than three local profiles qualify, we widen from core areas to secondary areas, then a surrounding region, then Toronto. When at least three qualifying profiles exist, the page never returns fewer than three.
Payment never changes a score. You message first; an agent cannot cold-contact you through this matching flow. Agents pay a membership to be listed (some founding members have a free introductory period). Matches only include members. Paying never changes a match score.
Worked example
A fixed example, calculated at build time
This example uses a downtown/east-end buyer who prefers Mandarin, evenings and weekends. It is calculated with the sample profiles above when this page is built.
Lina Park Sample profile
99%Lifestyle 98 · favourites 100 · experience 100 · home style 99 · availability 100 · language 100
Celeste Wong Sample profile
66%Lifestyle 83 · favourites 67 · experience 83 · home style 78 · availability 100 · language 0
Dev Singh Sample profile
51%Lifestyle 80 · favourites 0 · experience 50 · home style 71 · availability 100 · language 0