Secrecy versus Transparency in Sales of Network Goods

In many of today's business settings, a firm faces the problem of deciding whether to commit to sales volume disclosure when selling network goods that are subject to demand uncertainty. If the sales are kept a secret, customers are uncertain about the purchase decisions of others when they make their own decisions. When the sales are transparent, customers are willing to pay more for a massive turnout, but may forgo the purchase if the turnout is found to be low. We identify two countervailing effects of sales volume disclosure: (i) a pro-transparency Matthew effect: that is, the firm has an intrinsic tendency, driven by positive network externalities, to expose itself to sales volume uncertainty, because the benefit of a realized large sales volume tends to outweigh the loss of a realized small sales volume; and (ii) a pro-secrecy saturation effect: that is, for a sufficiently large expectation of network benefits, customers would make a purchase anyway even without knowing the exact sales volume realization, but may be turned away if observing a small sales volume realization. We quantify the tradeoff between these two effects under three pricing strategies: exogenous pricing, endogenous pricing and contingent pricing. For the exogenous pricing problem, we show that transparency is dominated (resp., dominating) when the network benefit is relatively strong (resp., weak). For the endogenous pricing problem, the tradeoff depends on the distribution of customer valuations and the expected network benefit: if the customer valuation distribution is bounded (resp., has a heavy tail) and the network benefit is sufficiently strong (resp., weak and bounded), maintaining secrecy (resp., transparency) is optimal. Under contingent pricing, transparency becomes always optimal. Lastly, we incorporate forward-looking customer behavior into our model and find that such strategic customer behavior may be detrimental or favorable to the firm who releases sales information.