Regardless of how the shares of FactSetResearch Systems (NYSE:FDS) have performed over the past couple of years, the business itself has been pretty consistent. This is not an accident. The company's subscription-based business model leads to a fairly predictable revenue stream and allows the company to plan its expenditures accordingly. Of course, it doesn't hurt that the company has a product that the financial community wants and needs.

The company's third quarter was no exception -- revenues were up nearly 25%, and earnings per share were up a strong 30%. Part of this strong performance is due to the company's acquisition of JCF in late 2004. As comparables with JCF are in place, I expect that the growth numbers will slow a bit. That said, the company also turned in another quarter of positive free cash flow, which it also does on a regular basis and over the years has led to large cash hoard, which the company often uses to repurchase shares in the open market and from founders.

From here, it looks like more of the same for FactSet, though I do expect that the growth rates will slow down a bit. Simply put, its clients need the type of data that FactSet and competitor Thomson (NYSE:TOC) provide. However, the company's consistent ability to retain more than 95% of its clients -- note that is different from subscribers, which may vary in a downturn -- means that you can pretty much count on the business to perform well in all seasons. The company also pays a small dividend, which it has been increasing.

If there has been one knock on the business, it is the revenue from soft dollars. Soft dollars are payments from customers for the company's data services via commissions to FactSet's brokerage subsidiary FactSet Data Systems. For FactSet's customers, this means that the cost of research is now paid for by the customer's clients in the brokerage fees on the transactions instead of in a management fee.

While FactSet is doing nothing wrong here, there has been concern in the past that perhaps FactSet's customers wouldn't use FactSet's services or as much of them if they had to account for the cost in their fee structure. Since then the fervor over this issue has died down, but should it return, I noticed a couple of years ago that the revenue from commissions at FactSet is a declining portion of the pie. Should the issue come back as a hot button, I'd double-check FactSet's 10-K, because the impact to the business may not be as large as is assumed.

For related Foolishness, check out:

Interested in other dividend-paying stocks? Why not take a peek at Motley Fool Income Investor? A free trial is just a click away.

Nathan Parmelee has no financial interest in any of the companies mentioned. You can view his profile here. The Motley Fool has an ironclad disclosure policy.