The other side of the card equation is to drive employees to put more on their cards -- corporate or p-cards, to be exact. The extreme of this is to mandate purchases (with limited or no exception) on cards. There are multiple reasons for this -- not the least of which are the air, hotel and cash-back program ideas the story suggests. Perhaps the most important thing the article overlooks is that card-based programs not only can help drive vendor compliance (if you set up your controls properly), but overall spend reporting and visibility, at least on a basic level. It's unlikely a small business will ever invest in a spend analysis package from the likes of Ariba, BravoSolution, SAP, Emptoris, Zycus, Spend Radar, etc. Yet basic card-level information (especially if you can get down to invoice-level detail) can provide a simple means to track spending trends and volumes, showing where to focus your sourcing efforts in the future. Card-level reporting data may not let you do a perfect 80/20 analysis of your overall spending, but if you can drive the bulk of your purchases onto a card, you'll get data you can work with. Also, another suggestion here is to check out Mint.com as well as the integration offered with Mint.com and various T&E, travel and related packages.
I've come to believe the concept of p-cards is more than sound (Oxygen Finance sold me on the philosophical concept of closed-loop contracting, buying and payment networks). Yet the execution of card-based approaches (with high bank APRs for suppliers) quickly yields diminishing returns for certain classes of spend in larger organizations. Clearly, one size does not fit all. In the future, I suspect we'll see the rise of non bank-based p-card-like networks -- even buying co-ops/GPOs that focus on payment mechanisms and controls -- but until then, even though the issuers are making obscene margins off of our plastic, the visibility and compliance such programs can provide to a small business are sound.