Customer knowledge databases

Customer Knowledge is a generic term for information about the customer. It includes transaction data, browsing histories, CRM and sales systems, and customer service records. It can also include linking social media, external data and modelling and analytic scoring. A database for customer knowledge as a whole links all these elements together for both analysis, but also to present a single view of the customer to sales or service staff.


Although integrated web-tracking and customer databases are on the horizon, most customer knowledge databases are normally used by businesses with hundreds or thousands of accounts where the business wants to be able to target and track the customers not just with transactional data and CRM-style data, but also pooling external data such as news reports, press and PR and more general customer information from the field. These type of databases are used to provide sales and service teams with information about each customer that they target to help with bids and sales activity, and to allow analysts to run cross-analysis in order to improve targeting and communications.

Such customer knowledge databases need to be able to draw in and summarise transaction data both at the account level and across customers, and to allow for the collection and integration of other structured and unstructured information such as financial reports, firmographics, customer press releases and coverage in the media to summary details about the company’s structure and objectives. This formal type of information needs to allow for sales and service teams to add and share news, account reports and account plans.

As such a customer database has to be much more flexible than at traditional transaction based database, though it must also have connections into the transaction databases to process orders, product supply and invoicing. And for those in a B2B world, the database needs to be flexible enough to support complex business relationships where there are potentially multiple contacts, multiple customer sites, and groups and networks of businesses.

The ideal of a customer knowledge database is not just about pooling existing data sources. Ideally it also allows information to be captured by field staff to allow notes and information to be shared by sales and non-sales teams. This means that access is most easily managed with a web-based interface with appropriate access control.

As the data requirements move and change and flexibility is required, a customer knowledge database has to be open to change and to a wide variety of data types and formats without the tightness of a standard relational database. At the heart of a ck-database is often a more loosely defined key-value or noSQL database that can grow over time, but that allows for structures and standards to be applied to the data. These tight-loose approaches are elements we have pioneered in Notanant (http://www.notanant.com) our web-based customer knowledge solution.

Data collected into the database also then needs to feed into sales and marketing planning software for instance in order to set appointments, schedule follow up calls, or use event-triggered marketing to return and recontact the customer. CRM solutions such as Salesforce.com allow for external connectors to legacy systems, and to bring in data from systems like SAP or Oracle.

Analysts will need to be able to pull off data in bulk, run queries and extract data for modelling or deeper analysis and then reapply the data as flags, classification and scores back to the main database.


Online behaviour and customer knowledge

The integration of website traffic data into customer knowledge databases is a developing field. Web traffic is normally anonymous or semi-anonymous unless the customer logs in, or has some other form of tracking. Web site data is currently mostly handled separately from customer knowledge data as the data streams are both large and difficult to operationalise within an account management environment, being more suited to modelling behaviour back to online marketing.

This separation is likely to reduce as it becomes possible to use online behaviours to flag for the possibility of contact. For instance, if a customer visits a support site with a problem, this might be an opportunity for a proactive service call to provide assistance. However, this then involves attaching much larger data-streams to customer data, again reinforcing the need for flexibility.

Similarly, adding social media handles to the customer database then allows for items such as tweets, or public likes to be collected and attached to the customer feed and can be used to trigger sales communications in real time.


Customer knowledge databases for analysis

Most databases for customer knowledge are based around creating an individual view of the customer for the purposes of sales, service or marketing activity. Analysis requires pulling the data together into a form that can be analysed statistically and interrogated for patterns and trends. Though the customer knowledge database is already a pool of data, analysis normally involves drawing extracts from the database and tools like data-warehousing in order to link together the multiple dimensions of data available.

As customer knowledge includes both structured and unstructured data, mechanisms and methods for classifying, sorting and filtering the unstructured data are necessary including tools such as text analytics.


For help and advice on building customer, competitor or marketing knowledge systems contact info@dobney.com