Want to learn more about personas? Check out What they are and how to use them
How many personas will I have?
- The algorithm determines the number of personas based on your customer data. When clustering for personas, Faraday applies a version of the k-means clustering algorithm to sort your enriched customer data into groups based on a variety of consumer attributes from the Faraday Identity Graph.
Are all persona clusters the same size?
- Nope, the size all depends on how the k-mean algorithm grouped your customers, by finding similar groups based on the clustering attributes for your industry.
Can I choose the number of personas?
- Yes and no...sure you could but you would be implementing your own personal bias to determine how many groups you want, versus allowing the data to determine the groups.
Can I create personas for each product?
- Faraday recommends creating personas based on all of your customers, from there we can add a post hoc analysis to include a breakout by product type to easily know which personas prefer specific products. This will allow you to easily manage who buys what and see changes over time.
Can I choose the fields that are used to cluster/create personas?
- Our data science team has spent a lot of time curating and reviewing the data fields chosen to use as clusters. These are ones that help provide distinct groupings as well as high coverage fields to allow 100% assignment to your customers, new customers, leads, and more.
- We do not recommend selecting fields outside of the pre-selected options chosen by our team.
Can I include first-party data to be used to create the clusters?
- Similar to choosing your own clusters, we cluster using FIG data. This allows 100% persona assignment to your customers, new customers, leads, and more. By requiring a specific action such as purchasing a product to be a part of the clustering, most new customers and leads will not be able to be assigned to persona until that action is taken.
- We do not recommend including first-party data as clustering but leverage post-hoc analysis to further understand the personas.
Will my personas change over time?
- They certainly can! Our methodology for creating personas allows us to apply personas to all of your new customers over time combined with leveraging our dashboard, you can easily monitor shifts in persona groups growing or changing, even reflect on growth in groups to recent marketing efforts.
How often will we refresh/re-run the persona clustering?
- It depends, usually we recommend once per year, but if there have been drastic changes to your customer base or product line, this may represent a being a good time to re-run the clusters.
- You will always be able to monitor personas in your dashboard
There are too many personas, how do I activate all of them?
- It is common for our algorithms to come up with 5 or more personas which at times can seem a bit overwhelming. The easy answer is you don't have to! Start taking a closer look at the persona groups, which are of higher value for you, can you focus on 1 - 2 creative variants? Are some similar enough that you could treat the same way and not need to come up with 5 creative variances?
- Work with your AM to help discuss strategies!
Faraday Personas are quantitatively developed using your first-party customer data, our third-party consumer data, and unsupervised machine learning (ML). At a high level, the ML algorithm sorts your enriched customer data into distinct groups, which are ultimately used to define your personas
When clustering for personas, Faraday applies a version of the k-means clustering algorithm to sort your enriched customer data into groups based on a variety of consumer attributes from the Faraday Identity Graph.
The distinct personas/output of the clustering analysis. The personas are called groups or clusters.
Post hoc analysis
After clustering, Faraday can apply first party or FIG data to further analyze the already determined clusters/groups/personas
A frequently-visual description of the relative numbers of times each possible outcome is observed or expected to occur