Not all wealth models predict which prospects would be just as likely as your existing customers or donors to spend or give. Our models fall into two buckets: descriptive models and predictive models. 

Both descriptive and predictive models allow you to gain greater insights into what makes your customers and donors unique, but there are key differences that should influence how you read your modeling results. 

            Note: any key sidenotes and/or related links to further dive into above stated details.

 

This article will cover:

  • Descriptive Modeling
  • Which models are descriptive?
  • Predictive Modeling
  • Which models are Predictive?
  • In Summary

 

 

Descriptive Modeling

A descriptive model takes a list of your best people and tells you what qualities set you apart from everyone else. You can then use that to identify others who resemble them!

Our descriptive models allow us to take the data you already have and show you commonalities among the individuals in your database. More than commonalities among your customers or donors, we’re showing you all of the qualities that make them unique. That ‘it’ factor that separates them from the rest of your customers and donors. So, we can help you find hidden gems in your database who you haven’t pursued yet or even under-performers who have the propensity and capacity to contribute more. 

Let’s say you’ve screened the existing donors in your database and you’re ready to identify which prospects to engage. The question is: how do you find donors most likely to contribute to your organization? Using a Look-Alike Model you can identify unique qualities among your best. So, using these analytical capabilities, you may find that a majority of your best donors are women over the age of 40 who like dogs. All you have to do here is upload a list of your donors. Then, we add our data, analyze it, and create a model of your list. 

So, once you’ve gained a greater understanding of what makes your donors unique, using a descriptive model, you’ll be able to identify and engaging other prospects who have those same qualities. 

            Note: any key sidenotes and/or related links to further dive into above stated details.

 

Which models are descriptive?

Look-Alike Model 

This is the most common type of model, and is the only one that you are currently able to build within the normal WE Platform.

Donor 360 Model

            Note: For more information on Look-Alike modeling, check out our  Guide [hyperlink].

 

Predictive Modeling

Predictive models, on the other hand, are custom algorithms that are created by our data science team. Using these unique algorithms, made just for you, we can predict who’s most likely to give or spend with your organization or business. This will help you evaluate who, within your database (and beyond) are likely to give or spend. Not just in general, but to your organization specifically.

Our data scientists will employ WealthEngine data along with yours to create a custom algorithm. This model will help you rank your prospect lists by order of similarity to your best. 

 

"A descriptive process model describes what has happened in a software development. It is often developed as the result of a postmortem analysis. A predictive model describes what should be done during software development, including responses to error situations."

            Note: any key sidenotes and/or related links to further dive into above stated details.

 

Which models are predictive?

Custom 4-pack models???

  • Major Gift Model
  • Planned Giving Model
  • Likelihood to Give Model
  • Next Gift Amount Model

Enterprise Models

 

For our models (custom, 4-pack, and enterprise), users can upload a file of the donors and prospects they want to be screened. From there, our data scientists prepare the information they’ve provided and use machine learning to create a custom model of the given data. Once the model scores are generated, results are delivered to our users with results and a detailed analysis of our findings.???????

            Note: any key sidenotes and/or related links to further dive into above stated details.

 

In Summary:

Both descriptive and predictive models allow you to gain greater insights into what makes your customers and donors unique. However, a descriptive model tells you that a given prospect fits the description of your top donors (which can certainly provide meaningful correlations), while a predictive model identifies that their behavior is similar, meaning it is able to better predict future behavior. 

 

A descriptive model will tell you that your lead looks like Samuel L. Jackson – but they will not necessarily behave like him; a predictive model will tell you that your lead behaves like Samuel L. Jackson – even if they don’t actually look like him. 

There will frequently be strong correlations between Samuel L. Jackson’s individual attributes and his behavior; often people who strongly resemble him are more likely to behave similarly than people who don’t. However, this is certainly not a guarantee and therefore cannot be truly predictive at the level of a model that measures behavior. 

            Note: any key sidenotes and/or related links to further dive into above stated details