Matching hotels to buyers, in a market with almost no ratings

The problem

JLL is one of the world's largest commercial real-estate services firms. When a hotel came to market, matching it to the buyers most likely to want it was manual work done by expensive people, and it could take up to three months — during which the property sat.

The asset they had was history: tens of thousands of properties, roughly two hundred thousand potential investors, and twenty-five years of transactions. The question was whether that history could rank, for any newly listed property, the hundred buyers worth calling first.

The sparsity problem

Recommender systems are a solved problem when your users rate things. Commercial property buyers do not. On the modeling subset — 492 properties against 10,351 users — the ratings matrix was 99.35% empty. Sixty-five percent of users had rated one item or fewer. Ninety percent had rated six or fewer.

Almost every published collaborative-filtering result assumes a matrix an order of magnitude denser than this one, so the signal had to come from behavior instead of ratings: what buyers responded to, what they bid on, what they had said they were looking for.

Four input feeds

Property records, responses to marketing campaigns, bidding history, and each investor's stated investment preferences.

Two models

A buyer model built from investment criteria — 408 features across 10,351 users — and a property model over location, type, tenure, brand, room count and asking price.

Preprocessing that mattered

Z-score normalization to remove per-user rating bias, and binarization to stop heavily skewed behavior from dominating the similarity measures.

The algorithm bake-off

Several algorithm families were implemented and compared. User-based collaborative filtering outperformed everything else, and cosine distance beat the alternatives for both the user-based and item-based variants.

A plain popular-items baseline outperformed item-based collaborative filtering, which ruled that variant out.

The headline-optimal nearest-neighbor value was nn = 10. We shipped nn = 20, because nn = 10 gave a worse true-positive rate across the higher false-positive range — the range this system operates in, since a broker works down a shortlist of calls.

What was delivered

A hybrid framework that builds several candidate models — collaborative filtering, content-based, vector-space and nearest-neighbor — evaluates them against each other, and serves recommendations from whichever currently performs best, instead of hard-coding the winner of one comparison.

The output is the hundred buyers most likely to want a given property, ranked, produced the moment it's listed. The process it replaced took up to three months.

Result

The recommendation list is generated instantly once a property goes to market, and the platform was adopted across JLL globally. The client reported an increase in hotel sales of over $13.2 million.