Professional terminology management systems are expensive, but the first step towards an efficient solution is collecting the content. This doesn't necessarily require high tech applications. Common office tools like Microsoft Excel or its free open source equivalent Calc, part of the OpenOffice.org suite, are perfect for the job.
I've recently created a sample spreadsheet termbase in Google Docs. It's free and it's web-based, so it allows me to easily share my dictionary. Moreover, this is a perfect input file for a conversion to SDL MultiTerm.
Although spreadsheets are primarily intended for numbers, formulas, charts and financial data, they include a few features that can be very convenient for glossaries: you can easily move columns around, insert new columns or hide existing ones, sort your data alphabetically, export the content into tab-delimited exchange files etc.
In spreadsheet format, you just need to respect the basic principles of database design. One such principle is that each column represents a field (English, French, Definition, Gender, ...), and each row represents a concept.
Furthermore, it is not a good idea to mix different types of information within the same field. If you do enter for instance both a term and its gender in the same field, it will be impossible to easily retrieve the term without the extra information. This is a critical issue once you want to migrate your data to a more sophisticated termbase solution.
As your termbase grows, you will at some point have a need for more power and flexibility. Once you require advanced search and filtering functions, integration in translation memory tools, collaborative workflows and dynamic publishing solutions, professional terminology management systems like MultiTerm are the next step forward.