Many document indexing has been developed since the emergence of the Internet for local, regional, or global authors. Some local document indexing has also been developed for Indonesian authors like Garuda, Neliti, IPI (Indonesian Publication Index), and Sinta (Indonesian Science and Technology Index). The most phenomenal among Indonesian document indexing is Sinta because it is used to score authors that are taken into account when the researcher apply for research grants and promotion. It is also to rank authors, institutions, and journals. Unfortunately, there was no research about Sinta to convince authors that Sinta database is reliable. The purposes of this research are to propose a technique to evaluate document indexing services such as Sinta and to propose some improvements. We used focused web crawler for data gathering. The data then processed by using bibliometrics and information retrievals methods. The results show that there were noises in Sinta's database like duplicate accounts, foreign authors, foreign articles, invalid Scopus articles, difference citation count, and retracted corrigendum articles. This research also revealed a side effect of Sinta that is unethical ways of publication by Indonesian authors. We also showed that some part of the design of Sinta was inappropriate.