The purpose of this paper is to determine whether the findings and the results of the text mining of the collected textual inquiries of the already acquired profitable clients can be used for the improvement of the pay-per-click campaigns and, accordingly, attract a greater number of profitable clients. In the first part of the paper, using the text mining conducted on the textual inquiries of profitable clients, certain conclusions have been made about the communication manner of the profitable target audience through the most commonly used terms and n-grams when requesting a certain service. In the second part of the paper, these findings and conclusions are applied to the communication in ads in the pay-per-click campaigns. The testing period, which lasts for a month, is then used to examine if the ads of the campaign created on the basis of the previously mentioned text mining can target a larger number of profitable and perspective clients more accurately than the ads created on the basis of expertise. The testing of the new ads is performed in the Google AdWords account of the company Identalia. The results were then compared using the marketing metrics (impression, click-through rate, conversion rate, profitability and perspectiveness of the potential client) which will provide an insight into the results. The open source software Orange was used for text mining, and Google AdWords platform and the CRM software of Identalia were used for the analysis of the results.