Diagnosis comes first, and it ignores what the company says about itself. An orientation is visible in artifacts: which function wins budget fights, what gets a manager promoted, what question the CEO asks first when numbers dip ('cut costs?' is production thinking, 'push harder?' is selling, 'what changed for the customer?' is marketing). Audit those artifacts and the real orientation confesses. Usually it's a philosophy the market outgrew a while ago, still enforced by the incentive plan.
A shift is then a sequencing problem, because the old orientation funds the company and therefore defends itself. The workable order: prove the new philosophy inside one protected unit with its own P&L and stage-appropriate metrics. Change incentives before communications, so the announcement lands on people already paid to believe it. Then scale, while guarding against relapse, because the first bad quarter is when the organisation reaches for its old reflexes.
The research gives the destination a concrete shape. Kohli and Jaworski defined market orientation as three organisational behaviours: intelligence generation (systematically learning what customers need now and next), intelligence dissemination (that learning crossing department walls), and responsiveness (the company acting on it within a competitive timeframe). Narver and Slater added interfunctional coordination: the customer is everyone's job, not marketing's department. The point of the constructs is that each behaviour can be audited, measured, and managed like any other capability.
The mature nuance: orientation is a portfolio decision, not a company-wide monolith. A commodity division can rationally stay production-oriented while the consumer arm runs fully marketing-oriented. The skill is running both without one culture strangling the other: separate scoreboards, separate talent models, one honest map of which business is which.