Real-time bidding(RTB)advertising is a promising business model emerging with the integration of big data and online marketing.Based on analysis of Web users Cookie data,RTB has the potential of identifying target audiences and delivering best matched ads in real time,and also enabling advertisers bid for each matched ad impression via auction-based programmatic buying paradigm.In RTB markets,publishers are suppliers of ad impressions(or ad inventory),and typically they have multiple channels to sell these ad impressions.In this paper,we analyze three commonly-used strategies for publishers to allocate the ad inventory,i.e.,the offline guaranteed contract,the online RTB channel,and the hybrid channel.We explicitly take the quality of ad impressions into consideration,since it is a key parameter for publishers decision when choosing a proper channel to sell these ad impressions.We establish models to optimize the reserve price for maximizing the revenue of those publishers choosing the online RTB channel and the hybrid channel.We also compared these three strategies in pursuit of finding the proper strategy for publishers when selling ad impressions with different qualities.