A full understanding of your data assets is important both in underpinning day-to-day operations and in developing business strategy. This applies not only when you’re using external IT systems but also in your internal work processes.
It’s tempting to assume that somebody somewhere has got this covered. But when you try to track down something definitive in writing, you may find that much of the big picture is in various people’s heads.
A good way to document everyone’s knowledge and ensure a common understanding of your data types and sources is to produce a data discovery report. This doesn’t need to be a long, complicated document, but it should answer a number of key questions.
1. Which systems are being used?
Most publishers use several different systems to manage their customer and product data. These could be either off-the shelf products or internal databases and spreadsheets.
2. What information does each system hold?
This might be individual contacts, institutional records, article submissions, book, journal and article sales, or usage and denials data.
3. How much data is in each system?
For customer data, what is the estimated number of contacts? For sales data, how many transactions does it cover?
4. How reliable is each data set?
Data quality issues (such as duplicate records, typos and out-of-date information) may be fixable, but a first step towards this is understanding the type and extent of such problems.
5. What is each data set used for?
What information do users need from each system? Bear in mind that the users the system is primarily designed for may not be the only ones to find the data useful; for example, subscriptions data can help editorial staff to find new authors from subscribing institutions.
6. What issues are there with the data?
Do users trust that the data is reliable? Can institutions be unambiguously identified via standardized IDs? Is the information stored GDPR-compliant, and does it support you in meeting changing open access requirements?
7. What further opportunities could the data provide?
Could you be making more use of the data, either as it stands or if it was cleaned or enhanced in specific ways?
We hope this checklist will prove useful in creating your own report. But if it all sounds a bit overwhelming and you’re not sure where to start, or what questions to ask, why not take advantage of our expertise in this area? Simply get in touch to find out more about commissioning a data discovery report from DataSalon.
The DataSalon Blog 