The Value of Data
Data indicating consumers' purchase behavior is very desirable for those advertisers targeting conversions. In this sense, search data is the most accurate. Search data is useful to advertisers as a source of purchase intent and, as a result, it is the most valuable data in the advertising market as a whole. Previous purchases combined with current intent signals result in high-level intent data indicating whether a consumer is close to a conversion (i.e. a purchase or other desired action). This data type is near the bottom of the marketing funnel and is highly valued.
Slightly more removed from data related to immediate purchases are data points that indicate consumers who are in-market, i.e. who have demonstrated a strong intent towards a product by navigating to a product page, adding a product to their cart, or filling out a quote request.
Even further away in the marketing funnel is interest-based targeting, i.e. consumers who have demonstrated some level of interest in a product or idea but not strong enough to assign them to the in-market category. Examples of this behavior are consumers who are reading blogs, articles or product reviews, who are surfing a hobby or fansite, who are reading industry news, etc.
Demographics data related to a consumer's general income, region (e.g. rural or urban), or industry type is of similar value as low-level interest data.
The value of location data may vary significantly. Broad-based location data, such as a postal code, is helpful to narrow down the gap of desired consumers. However, location can also be very specific (e.g. Wi-Fi- triangulated data within a shopping mall or barometric pressure that might indicate the exact floor within a mall at which the customer finds itself). Based on such data, advertisers can target consumers who are in the immediate vicinity of their stores. Such data is as valuable as high-level intent data described above.

How relevant is data accuracy?

Since buyers have no good way of ascertaining data accuracy, currently this isn't very relevant.
Since performance marketeers have some methods to gauge ROI (e.g. how many purchases resulted from this campaign), they overcome this by buying a small volume first to test the waters.
Brand marketeers, on the other hand, have little ways of understanding the impact of their campaign. It's important that the right audience is targeted and that they reach their volume goal, but since awareness isn't easily measurable, they tend to take more care with data accuracy — but the only way they have to do this is through dealing with reputed data sources.
Offering a verification layer for data accuracy might be a valuable service to provide for brand marketing purposes.

Related reading

Why We Need to Rethink the Price of Data - Adotas
How to Price Audiences When Selling Data | Lotame
Last modified 7mo ago