Harrah's mined its own data and courted its average customers instead of the whales

the problem

Every casino chased rich whales into glitzier properties; the ordinary customer nobody watched paid the bills

background

what everyone would do

The orthodox play was to match the arms race: build a bigger, more spectacular property, court the same high-rollers with comped suites and no-limit credit, and out-glamour the neighbours. It fails because whales have all the leverage and a competitor next door, so the capex buys no durable edge, and nobody in the industry had measured whether the VIP was even the profitable customer.

what they saw

The industry settled who matters by argument, but the answer sat unread in its own transaction data; Loveman treated customer value as measurable, grown by experiments, not a segment to court.

the move

Harrah's made the counterintuitive call to stop chasing high-rollers and instead sink its capital into a data-analytics team (Loveman's 'propeller heads'), a customer loyalty program (Total Rewards) that tracked every player's actual behaviour across all properties, and a pool of controlled experiments on offers and comps. The measurements showed the profitable engine was not the VIP but the unglamorous middle — ordinary repeat and local customers — and the company directed its spending to lift their value.

why it works

The consensus belief that high-rollers drive profit was a doctrine nobody checked, so capital drained into glitzier properties chasing customers with maximum leverage and no loyalty. Replacing it with measurement — Total Rewards tracking actual behaviour across properties and controlled experiments on offers — let Harrah's redirect each dollar to the segments where its own data showed the largest, most repeatable return: the loyal middle. That spending compounded across visits, producing revenue growth the capex arms race could not, and because the answer derived from Harrah's internal data, competitors running on intuition could not copy the finding until Harrah's had already built the loyalty lead.

the payoff

Against a weak economy Harrah's revenue rose 50% and stock and profits about doubled in 1999, beating rivals and catching Wall Street's eye.

where it breaks

It fails where there is no repeat-transaction stream to mine and cultivate: a purely one-shot purchase, project-based deal, or market without the mechanics of return visits leaves no 'average customer' whose lifetime value can be measured and grown, so the data yields nothing. It also erodes as loyalties commoditize once every competitor runs a lookalike rewards program, and it depends on actually running experiments and acting on the answers — an organisation that measures but argues anyway gets only the cost of the analytics. The capital must also go to the measured segment, not the one that still 'feels' important.

what came after

The case became a canonical Harvard Business School teaching story, and Total Rewards-style data-driven loyalty was copied across the gaming industry; combined with the analytics culture that produced it, the approach let Harrah's widen its gap over competitors who still trusted intuition over empirical data before they belatedly imitated the program.

references

1. [1]HBS Faculty & Research — Harrah's Entertainment Inc. (Case 502-011)Harvard Business School, 2001hbs.edu
2. [2]The Case Centre — Harrah's Entertainment, Inc (9-502-011)The Case Centre, 2001thecasecentre.org

keep it

same kind of clever

• 1020 · Abu Rayhan al-BiruniAl-Biruni measured the Earth's size from one mountaintop instead of crossing a desert
• 400 BCE · Classical Athens (antidosis procedure)Athens made rich citizens prove their wealth by betting their entire estate on it
• 1797 · Henry CavendishCavendish weighed the Earth by hiding from his own experiment in the next room