Business Intelligence is a highly customer-facing job, producing reporting and analytic systems for customers. It has something of a barrier to entry as you can't really practice it in a vacuum. You can only really meaningfully experiment with it in the context of somebody's live data set.
If you have a job offer, take it unless it looks like there might be issues with the job situation. The tooling can vary, but a good working knowledge of SQL and databases is a core skill. Most of the other tooling is fairly easy to learn with the possible exception of OLAP cubes - MDX has something of a paradigm hump to get over if you're used to SQL.
There is something of a view that it isn't real progamming, but you tend to find yourself writing little utilities to do various things. For example, in recent times I've written:
A cube partition cloning utility in
C#
A binary BCP file row-counter in C
(for speed)
A BCP control file generator that
used introspection from the system
dictionary in Python.
A generic type-II slowly changing
dimension handler.
Quite a few code generation tools
of one sort or another
A java-doc like tool to extract structred
comments from a SQL DDL file and produce
a data dictionary document. This tool was
subsequently adapted to create a documentation tool
for a report model that extracted the element
descriptions.
Ungodly amounts of T-SQL stored
procedure code.
From that perspective It's not fair to say that Business Intelligence is not about 'Real' programming. While there is a prevalence of 'Tools Guys' in the industry, the tools themselves are by and large fairly easy to learn. Having programming ability gives you something of a differentiator in that you can get out of a tight corner when the tools run out of steam (which happens quite frequently on even moderately complex data).
In many cases (about half of all DW systems in financial services are done this way as an explicit design decision) Stored procedure code is the preferred means of writing a complex ETL routine. From this point the programming skill is a central capability.
One issue is that recruitment agents tend to match on buzzwords and will exclude you from consideration based on lack of prior experience with a tool that would take you a few days to pick up. However, if you've already got an offer, that's the first hurdle overcome.
Also, you will find that cultivating relationships with the business and getting together good business analysis skills is a big win in this industry. If you find yourself working in an industry for any length of time, the domain skills will be a real bonus when negotiating for rates or applying for more senior jobs. Good customer-facing analysis skills are always a differentiator in any IT-related development work.
In London, a business intelligence contractor can earn on the high side of £100,000/year before tax.