Roadmap: Creating a simple marketing data science roadmap
AIDA is also known as the conversion funnel. It is the process that marketers use to nurture customers through the journey with the end goal being a purchase/conversion/action.
Awareness — This is usually some form of brand marketing with the strategic goal of raising awareness about the brand to potential customers. e.g. Display ads for prospecting
Interest — At this stage there are a set of customers who’s attention has been captured. e.g. Someone clicking on a display ad and landing on the site
Desire — Once a customer has shown interest then the task is to showcase the benefits of the service/product on offer. It’s quite normal to use language that would resonate with the customer
Action — This is the process of converting the customer through a seamless checkout process. e.g. a one click online checkout using PayPal which saves the customer having to fill in name, address, payment details etc.
What does this mean for marketing data scientists?
If you are collaborating with colleagues in marketing then understanding the marketing plan, the intentions and how data can be used to make informed choices is a must. Enhancing that further with models that are relevant adds further value. For example, if the goal of the marketing team is to raise awareness then using an XGBoost to find customers to remarket to is good but not relevant. Instead, using clustering techniques such as k-means can help identify groups of customers which can then be used for lookalikes or regression/drivers modelling using survey data to see which group of customers see’s the greatest increase in awareness as a result of advertising. As a result, it is important for both the marketing and data science function to connect and understand what’s on the radar for the marketing plan, where data science can be used and what other data science capabilities can be drawn on to drive marketing effectiveness and efficiency.
The roadmap is a living and breathing document which extends as a the business evolves. Some capabilities can be used in more than one scenario, other times using multiple capabilities may be necessary to maximise the impact. As mentioned in my previous post, the three key benefits are:
- Its a north star
- It shows the business what the team will be working on, the dependencies and expected due dates.
- It gives the team focus
So what does a data science roadmap look like?
Above we have a very simple data science roadmap that shows the organisations measures of success/KPIs and the data science models and capabilities needed. Once the capabilities are scoped out, the data science team would need to create a backlog of work and liaise with the scrum master to undertake sprint planning etc. The roadmap would keep on extending as the business evolves and the challenges change. Keeping a track of what has been achieved prevents duplication of effort. The team should try and maximise the capability by using it in as many places as possible where relevant. This will help grow the ROI of the data science project.
