This paper shows (i) that business sentiment, as captured by survey data, matters for monetary policy decisions in Canada, and (ii) how business perspectives are affected by monetary policy shocks. Measures of business sentiment (soft data) are shown to have systematic explanatory power for monetary policy decisions over and above typical Taylor rule variables. Stronger (weaker) survey results lead to higher (lower) policy rates over the period of study (2001–16). Moreover, we study the effects of monetary policy shocks on firms’ business perspectives using data from the Bank of Canada’s quarterly Business Outlook Survey. The monetary shocks are defined as the fitted residuals of the Taylor rule. Overall, the results are in agreement with the qualitative effects of monetary policy shocks described in the literature. For instance, an unanticipated tightening in monetary policy a year ago (or more) results in firms reporting tighter lending conditions today, as well as slower expected dynamics of future sales, wage growth and output prices. The results are qualitatively similar whether shocks are derived from a standard Taylor rule (hard data) or from an alternative Taylor rule (soft data).