1. Backup your file in SA 2004.
2. Select Maintenance, Database Utilities, Compact Database (only because you are using a version from before 2008).
3. If there are no errors, then install Sage 50.
4. After installing the original Sage 50, install the update if it was not part of the original install.  If you do not have the latest update for your version or cannot find it on the website, then open the sample file and it should download for you.  It will normally be downloaded to C:\Users\Public\Documents\Sage Software\Simply Accounting\Download and the latest file is called SA_20161CPUP1.exe but if you are not running Sage 50 2016, this file will not be the correct one.  You need to use the one from your version.   The "20161" part indicates 2016.1, the C indicates Canadian, and PUP is short for Payroll or Program Update Plan or something like that.  There is always a 1 at the end.  So if you are installing 2015, then the latest release is 2015.3.
5. Once the update is installed open your Simply 2004 files and allow the program to convert.

Once you update, the sdb file will be converted to an sai file and you will open the sai file from now on.  There are variations on the above so it is not a set-in-stone procedure, but it works for me when I do this for clients.  If you are upgrading to 2016 for example, some people will install 2016.0 and upgrade the data file, then find out after that the 2016.1 update for the program is available and then upgrade the program, then go back and upgrade the data file again.  I prefer to update a data file once with only the latest version available.

If your data file is stored on a server computer, you need to install the Server components of the Connection Manager on the server before you can access the data file on that server.  This was not needed back in 2004 but it is now.

On a moderately sized file, the above process should be roughly about  a half hour but it does depend on whether you were given the latest release to install or have to download it separately, how big your data file really is and if you are running an older and slower computer.

Hope this helps