Persona Mapping: The data science behind identity
Before I joined Idealspot my fund used data science to identify trends in the municipal bond market, focusing in particular on the “retail persona.” Of the 40,000 trades per day, 95% were small lots of ~$70k invested and the other 2,000 trades were $5mm+ invested. Everyone focused on the big trades, very few people (when I launched Troubadour) were focusing on what we called the “long tail.” In 2013 there was extreme value in the long tail. Mining the data and buying habits of this “persona,” became my mission. Focusing on this persona was the core of my strategy, the tactics being to focus on areas other people were not.
Fast forward to 2019, in my opinion an even larger opportunity exists in the commercial real estate market.
- Reason #1: it’s a larger asset class than municipals ($16 trn vs $4 trn)
- Reason#2: there are a bunch of legacy CRE “bros” that made their money off of relationships and back of the envelope math, but the good ole boy network is dying
- Reason#3: very few of the participants in this market are paying attention to data, especially the “CRE bros”
In the land of the blind the one eyed woman is king.
The holy trinity of data science is Identity + Intent + Location. Persona is the identity component of data science. It’s the “who.” Who are these people? Categories include, gender, age, income, ethnicity, household size, marital status, etc. Starting with the persona from scratch can be overwhelming unless you have an array of your own proprietary data. There are really cool ways to incorporate GIS mapping to pull out persona and more narrowly define it. There are also creative ways to create a synthetic identity by mining the data of your competitor’s locations.
One example, using mobile data IdealSpot has the ability to look at macro statistics around points of origin. We call these migration patterns. If you can capture where people come from that go to your business, or where they go to during the day if you own say a multi-family building, you will know a lot more about their persona by using data science.
Over the next several weeks I will write more about persona mapping and then get into figuring out Intent and Location using data.