This content, written by Rich Taylor, was initially posted in Looker Blog on Mar 11, 2016. The content is subject to limited support.
Arguably the most important measurement problem for marketing departments is campaign attribution or understanding which marketing campaigns and activities drive pipeline and revenue. I have help build four at four different companies. If your data is centralized, you can build marketing attribution models in hours not weeks.
Why should you centralize your data and what does that mean?
Marketers have data and programs spread across too many places. hiefmartec.com lists 1,876 companies as part of the and on the marketing technologies every company must use. Centralizing marketing data is crucial to understanding if campaigns and programs are delivering value to the organization.
So how can you centralize your data? I’ve seen this happen in two stages. Stage 1 is to push all campaigns into your CRM. With the Force.com platform from Salesforce.com (SFDC) and the ability for marketing software to push data into SFDC, it is fairly easy to centralize your data into a CRM. Marketing automation platforms like Marketo, Hubspot, Pardot and Eloqua can automate many processes into your CRM as well. Centralizing data in a CRM then allows marketers to have a single source for all marketing campaigns and programs, but you are stuck with the structure that your CRM gives you. You are also limited to only using software that will push data into your CRM. It is a great first step, but usually leaves you with a desire for more.
Stage 2 involves moving data into a database and using a data analytics platform to understand your marketing metrics and KPIs. Centralizing marketing data in a database can provides the most flexible option for your data, reporting and analytic needs. Buyer beware that not all data analytics platforms are created equal and many people have been disappointed by both the length and inflexibility of the platforms they choose. Marketers find that either the data analytics is hidden behind a gatekeeper and takes too long to get what they need or they have all the self-service capabilities they need but not enough access to the data they desire.
Is there a data analytics platform that provides governance and self-service?
Yes, Looker provides both. I have a personal experience to illustrate why choosing the right data analytics platform can make data modeling project like marketing attribution, take hours instead of weeks.
While working at a previous company I suggested we create a marketing attribution model the first month I was there. While most of our marketing campaign data was in Salesforce.com, we didn’t want to use a marketing attribution tool like Convertro, BrightFunnel and Bizible. We wanted to build our own. The tool of choice was data visualization software that was helpful for self-service. It took 3 months to get our data into a database, another 2 months to build a model and 2 months to get buy in from executives in other departments. We decided on one attribution model to use as building other attribution models was cumbersome in the tool we chose. Once there was buy-in, we had to change our infrastructure to allow more people to get access to the data. The cost of the project went way up and while we could now measure marketing attribution to pipeline, it didn’t feel like a successful project.
When I joined Looker I saw a similar issue, we needed to measure marketing attribution. I mentioned this one Monday morning and by Friday we had 3 different marketing attribution models built on our data (first touch, last touch and modified first touch). With only a few hours of work we had all the models we needed to get started. One of our marketing managers built a dashboard from the models the next week. The following month another data analyst caught wind of what we were doing and built a multi-touch attribution model in a couple days in his spare time.
The difference was that our data was centralized in a Redshift database and the data platform (we use Looker at Looker :), allowed data analysts to model our data easily. The result is dashboards built by marketers that have access to the data we need.