Two companies of the same size, in the same sector, with what looks like the same data, can come back with valuations an order of magnitude apart. The difference is almost never the subject matter. It is these six things.
The single biggest factor, and the one you cannot fix. A buyer needs enough history to test a signal against outcomes they already know. Eighteen months of data cannot be back-tested against a downturn that happened three years ago. Ten years can. This is why a slightly boring dataset with a decade behind it beats an interesting one that started last year.
A dataset that updates once a quarter is a report. One that updates daily is a feed. Feeds are bought as subscriptions and priced annually. Reports are bought once and haggled over. If your systems can deliver daily and you are delivering monthly, you are being paid for the wrong product.
Aggregated data has usually had the useful part removed. A monthly total tells a buyer what happened. Individual transactions tell them why, and let them cut the data in ways you have not thought of. The more you summarize before you sell, the fewer buyers you have.
Data that comes straight from the source is worth more than data that has been cleaned, corrected, reconciled or restated. Every adjustment is a decision somebody made, and a buyer has to model that decision before they can trust the number. If more than a fifth of your data is touched after capture, expect questions.
Geographic and market coverage widens the buyer pool rather than raising the price per buyer. A dataset covering one country has one set of buyers. The same dataset covering twelve has several sets, and several buyers bidding is what moves a price.
The hardest to judge from the inside. Most companies assume their data is ordinary because they look at it every day. Uniqueness is not about the subject, it is about the combination: this activity, at this granularity, over this period, from this vantage point. A buyer will pay for a view they cannot assemble anywhere else, and they are usually better placed than you to know whether it exists.
Two of the six you can change. Refresh frequency is usually an engineering decision rather than a data one. Granularity is often a matter of what you export rather than what you hold.
The other four are what they are. Which is the point of putting a number on it before you commit to anything.
Xferdata values a company's data, builds the auction pack, and runs the auction to buyers globally.
If you own the data, run a valuation and find out what yours is worth.
If you buy data, sign up and tell us what you are after. You will see packs as they come to market.