Spreadsheets are frequently used in industry to support critical business decisions. Unfortunately, they also suffer from error-proneness, which sometimes results in costly consequences. Experiments in the field of program education have shown that programmers tend to make fewer errors and can better focus on the logic of a program if they use a block-based language instead of a textual one. We hypothesize that a block-based formula editor could support spreadsheet users in a similar way. Therefore, we develop XLBlocks and conduct a think-aloud study with 13 experienced spreadsheet users from industry. Participants are asked to create and edit several formulas, using our block-based language. We then ask them to evaluate this editor using the Cognitive Dimensions of Notations framework. We found that for all dimensions the block-based formula editor received a better evaluation than the default text-based formula editor.