Insights Through Data

I’m introducing a different type of post this week: still a tips and tricks style of post, but it stands against additional business context than usual, to be more relevant and applicable.

Sequence Analysis techniques and concepts originate from biology research, but they are actually very relevant in the field I am more familiar with: Marketing. Since a lot of Marketing analytics revolve around the understanding of customer behavior, it only seems logical that the sequence of the observed behavior comes to prominence. For instance, in most research scenarios, a customer who purchased a music album on iTunes after watching a video of the artist on YouTube, should be segmented differently than the one who purchased that same album and then proceeded to watch the same YouTube video: the sequence reveals most certainly a different causality between those actions.

Now, unlike biology where sequencing applies to strings of nucleic acids, sequencing in marketing implies trying to identify trends within large amounts of transactions, using reference time/date stamps. Here are some examples:

• Combing through weblog data such as Google Analytics, to understand the paths taken by visitors among the pages or actions on a web property, to optimize their experience and drive them towards a favored outcome, such as purchasing in e-commerce.
• Analyzing Marketing Automation data such as Marketo, Eloqua, Pardot to identify the combined impact of marketing campaigns in sequence. Is a free trial campaign performing better after or before a webinar to introduce the offer? Is there a Golden Path of campaign touches that will convert a prospect to a customer in a minimal amount of time at a minimal cost? It is also interesting to note that the effectiveness of an automated marketing campaign can vary depending on the industry in which it is launched. For example, in the finance sector, using marketing automation tools can be a fantastic way to curate a personalized approach to marketing involving custom emails and text messages for instance. In other industries, alternative approaches might have more of an impact.
• Monitoring and maintaining Customer Relationship to prevent churn in subscriptions using predictive models for instance, using CRM data (SFDC, SAP, Microsoft Dynamics, Oracle…)
• Integrating your marketing solutions together (e.g. finding a way to integrate SMS for Microsoft Dynamics) to help improve your marketing efforts and getting people to pass through the customer journey funnel as quickly as possible.
• Recommendation Systems to evaluate what is the most relevant next purchase for a customer, using predictive models here again, feeding from past transaction sequences, like seen on Amazon, the uncontested pioneer in that field.

Continue reading →