Did you know that in 1960, the world of marketing was dealing with an average of 5 marketing channels and in 2013, that numbers has increased to 60+?! No wonder integrating online and offline data has become such a pressing need yet an increasing challenge for almost all marketing organizations.
As integrating online and offline channel mixes becomes more commonly practiced in overall campaign strategies, the marketing impact on business performance can no longer be determined by metrics like email opens/clicks, ad impressions, lead conversion, etc. An integrated database is the foundation for a holistic view of our customers and their interaction with the brand.
Here are key practices and procedures to apply to database projects:
1. Identify topline business objectives and benefits.
While the notion of having an integrated database or data warehouse is obvious, it is still critical for you to outline the topline business objectives and benefits before you start building the database. It will help you prioritize the data sources and data attributes that you need to integrate, and it will also have a direct impact on your project scope, timeline, and budget.
For example, if your topline objective is to monitor word-of-mouth or customer referrals on brand awareness, then social media content and review content may be critical. If your topline objective is to understand and optimize campaign attribution, then you would want to integrate as many online and offline channels as your communications apply. If your topline objective is to prove marketing impact on ROI, then transaction data is just as equally important as both your online and offline campaign information.
2. Evaluate your current data assets and identify gaps.
So, what data do you have access to (campaign, lead, opportunity, transaction, etc.), where are they stored (enterprise data warehouse, marketing automation system, sales automation system, personal computer, vendor hosted, etc.), what formats are they in (CSV, Excel, document, video, text, etc.), what types of data are they (structured, multi-structured, unstructured, etc.). Make a list and fair assessment because this will help you communicate effectively with your IT counterparts (or, if you are not technical enough, they can help you develop this list as well).
Once you have the list, it will also help you be clear what data you don’t have. Interestingly enough, while offline channels have a much longer history than online channels, many times it’s the offline data that marketers are challenged to acquire.
3. Create a plan to start tracking or acquiring the data you need:
Sometimes it is easy enough to acquire the data you don’t currently have, such as demographics and firmographics information, market sizing, email addresses, DMA (Designated Market Area), etc., from third party resources. Other times, it is not so easy, especially if the data is unique to your business (e.g., contact history, marketing spend details, survey information, etc.).
If the data is not obtainable elsewhere, create a plan to start tracking or acquiring it. This may involve necessary changes in management and processes, investment in technologies and people, cross-functional alignments and negotiations, as well as holding your advertising agencies and campaign execution vendors accountable for tracking and reporting back campaign results. I recently had the chance to talk to several media agencies, and they are highly motivated to improve their tracking capabilities and ability to provide such data back to their clients.
4. Decide the proper database structure:
Before you turn to Oracle, SQL Server, or Teradata, let your data help you make the decision. The types of data sources, formats, volume, and, more importantly, what do you want to do with the them, will dictate what type of database structure you should build. Most mainstream databases, like the ones I mentioned above, are built on a relational data structure and work best with structured data.
If you are planning to bring multi-structured and/or unstructured data into your database, you should examine the extended features for the integration of structured and unstructured data in your database system or explore other database solutions. The database technologies for multi-structured and unstructured data are still fairly new and can be over our heads as business owners.
My best suggestion: start educating yourself by talking to your IT friends/colleagues and attending some big data technology conferences.
5. Keep up driving data quality:
Online data or offline data, structured data or unstructured data… honestly, there are many tools and talented professionals out there today to support the effort of data capturing, managing, reporting, and analysis. So don’t worry too much about it. You may have world class systems and talent, but if your data quality sucks, it will be really hard to make the magic.
How many times have you heard your data people (or maybe you have experienced it yourself) complain about missing data, dirty data, untruthful data, incomplete data, and so on? Make sure you have good control of data quality, and it will make the data integration and ongoing management much easier.
6. Get the right people on the project team:
In the SaaS (software as service) dominated world, it’s all about empowering business people to do things themselves and minimize dependency on IT supports. But, if there’s one area in which I would not shy away from asking for IT help, it would be building a database.
A database is the foundation of a business, and it is critical to build it right from the beginning. It requires in-depth information gathering, project planning, solution evaluation, data design, and much more. And getting the experienced subject matter experts on your project team is key to successful implementation.