1. Prioritise your choice of selection
The range of choices in business data tends to be much more limited than for consumer lists. It's a useful exercise therefore to prioritise them in terms of the 'need to have selections' and then follow up with the 'nice to have' where possible.
2. Check the provenance of your data carefully
Business data changes very quickly, and using poor quality data risks you losing much more than the price of goneaways. It's very important that you get clear answers about the source and ownership of the data, how it was built, how it is maintained - and its average age. Keep asking until you feel satisfied with the answers. Ensure that the database is screened using TPS, FPS and MPS, and that the owner complies with data protection law. Preferably get it in writing.
3. Base your choice on relevance, not size
Size isn't the most important thing with a business database. There are ways that numbers of records in selections can become confused, so direct comparisons are difficult.
Base your decision on whether your brief is answered exactly. If necessary provide each data owner with a spreadsheet to ensure that you are comparing like with like. Remember that your needs may not necessarily be satisfied from one source alone.
Ask for a sample of the data to the selection you need and ensure that you make the selection - not the data owner. Call a random selection of the data to make your comparison.
4. Pick the right licensing deal
It rapidly becomes cost effective to license data for 12 months' multiple use - particularly when you know you may want to reuse the list, or your activity is strategic rather than tactical. Remember that it's very difficult to build intelligence about key target groups if you've only paid for single use.
Some data owners will allow you to purchase the data outright, but first think carefully about the price of maintaining that data - and the management headaches associated with it. Why not leave responsibility for maintenance of the data with the data owner ?
5. Ask for discount, but pay for quality
Always drive for a bargain, but be prepared to pay for good quality data. The data owner's position on price may well confirm or confound your findings to the questions you raised in point two, above.