The Hidden Data in Invoices


The data business is one of creativity, and what could be more creative than asking companies to send you their invoices and other types of billing data so you can get them into a database and sell the aggregate results back to them? Now, why would something like this ever make sense? Well, in many industries, there is nothing more useful or important than pricing information. Yet the pricing information that many companies publish (if they publish it at all) is almost always the list price. And in many industries, the list price is close to meaningless, since every customer will have a special deal and varying discount. So how do you develop a database of what companies are really paying for specific products and services? Ask to see their invoices!

There are lots of spins on this intriguing model, so let’s take a look…

The question already on your mind quite likely is, “Why would any company let me look at its invoices?” The simple answer is what I often refer to as “strength in numbers.” A company will happily give up its individual data (properly secured and anonymized) in exchange for access to the aggregate results. And they’ll pay for that access, and that’s exactly the play here.

A great example of collecting, normalizing and reporting out information drawn directly from the internal systems of advertising agencies can be found in a company called SQAD. Its NetCosts product collects data for media purchases from advertising agencies worldwide, generating what may be the only honest look at what broadcasters are charging for media buys, and even what they have charged into the future. You can immediately see how valuable this information can be.

Thomson West has a product called Peer Monitor that does the same thing in a slightly different way: rather than work with the recipients of invoices, it works with the senders of invoices, in this case law firms, to collect similar data to be used in similar ways.

If it sounds like a lot of work, it is, or probably was. That’s because SQAD now receives most of its purchase data digitally, through interfaces with client systems. And while those interfaces were doubtless painful to build, at the same time SQAD has built an almost impregnable franchise, because as long as SQAD doesn’t get greedy, nobody can justify the time, cost and pain to try to compete with them. The same holds true for Peer Monitor.

There’s also what might be called the pre-order model. Here, your objective is to gather RFPs and proposals to get a true look at what companies are proposing to charge for their products. One advantage of picking up information at this stage is that there is often a lot more detail, allowing you to collect even more granular product price data. A great example of this model is MD Buyline, a company that collects price quotes and proposals on medical equipment to build a high-accuracy pricing database.

Lots of variant models, but the objective is the same: gather data on actual prices being charged in the marketplace, the more granular the better. Your job is to aggregate, normalize, and report back to the marketplace, while protecting the anonymity of those who participate.

It’s important to note that the need for pricing data isn’t equally compelling in every market. The dynamic seems to be a reasonably large pool of both buyers and sellers, and a solid tradition of haggling over price. It won’t work for everyone, but it’s certainly worth considering if it would work in your market.

Comment