Some Personal News

#41. The How, What, Why of Donor Data Enrichment + Ad Tech and thinking like a Product Marketer

May 28, 2023

Happy Sunday. To those I spoke to this week or spent time with in-person - thank you. A very warm welcome to all the new subscribers. I’m thrilled to have you as readers and truly appreciate your feedback and support.

Let’s dig in.

Ad Tech: Donor Data, Media & Cookies

The latest regulatory blow to Meta is a $1.2bn fine from the EU - and an order to suspend transfers of user (donor) data to the US.

It said that Facebook, whose European headquarters is in Dublin, had violated rules requiring that transfers of personal data from the EU to the US had appropriate safeguards in place.

It added that changes introduced by Meta Ireland in response to a 2020 ruling by the European Court of Justice “did not address the risks to the fundamental rights and freedoms” of such transfers - even though the transfers largely took place on the basis of contractual clauses endorsed by the European Commission.

I wonder if Meta will have to move away from hyper targeting.

Keep an eye out because I can see Meta developing more media formats (apps, devices, creator formats, etc) that don't require hyper targeting. Zuck must realize by now that the hyper targeting and exploitative practices of early Facebook are now a lost cause.

Don’t forget: user targeting isn’t the only lever a digital media owner has to improve the value of an ad-supported media property. The media owner can also change the quality and content of the media, as well as how (and when and where) it's consumed.

Interestingly, the problems with the current Ad Tech environment drives a new Apple ad - dramatizing how health data is shared without people realizing. Making the same point in a different way recently was an Antiabortion Group who used cellphone data to target ads to Planned Parenthood visitors.

Despite the fact that the world developed a vaccine for COVID-19 faster than Google has been able to get rid of third-party cookies from Chrome… the next stage in the deprecation of cookies is (almost) upon us. Google reiterated this week its 2024 deadline and to get there we’ll see General Availability from July.

This means any developers will have access to the APIs to test. And from Q1 next year a randomly selected 1% of Chrome users will not have cookies, period - and then Google will slowly expand deprecation to more users throughout the year.

Enriching Donor Data

Last week I wrote about running objective-driven growth experiments through the lens of donor-related problems. Quantifying the increase in CVR or engagement on social posts is simple yet understanding the donor’s state of mind, motivation, feelings and readiness to donate is infinitely more difficult.

Data can help. But what type of data exactly? Data Enrichment has been almost as buzzword-y as “CDP” or “AI” recently, so let’s dig into it.

Alteryx is a data and analytics automation platform and defines Data Enrichment as “the process of combining first-party data from internal sources with disparate data from other internal systems or third-party data from external sources.”

Supposedly, “enriched” data becomes more valuable and insightful to the organization. However, first-party data is a confusing term and before you can enrich “it,” you need to know what to enrich.

In a non-profit setup, the following types of 1^st party data can be the most helpful:

Your website behavior data – sessions, and pageviews showing what current and prospective supporters view on your site before making their first (or subsequent) donations.
Your CRM donor PII data – their names, email and physical addresses, and phone numbers.

These types of 1^st party data – when enriched – help answer the following “question groups” respectively:

Content performance, engagement, and likelihood/timeline to convert into a donor or make a subsequent donation based on the content viewed.
Donor profile – who are your donors, what is their capacity to donate, intrinsic motivation, and affiliation to your brand.

Here are three practical examples I’ve been mulling over for each group:

1. Using your website behavior data as a source for enrichment itself.

This use case is arguably the simplest – the goal is to use data from existing donors at the “next stage” of engagement to determine the best next step for those “behind.” Here’s an updated visual I’ve shared several times:

A picture containing text, screenshot, font, number

Description automatically generated

I’ll pick the 3^rd red rectangle above for an example. In your CRM and Web Analytics platform, you already have many use cases where a supporter became a monthly donor after donating several times for the same cause. And you also have many people who are one step behind – they’ve already donated more than once but are yet to turn into monthly donors.

It’s easy to help them > In your Google Analytics (whether older Universal Analytics or newer GA4 version) there’s a “Path Exploration” report. Use this to analyze the behavior of those who already converted and identify the content topics or pages that worked best for them in their last session converting from irregular to monthly donors. Then you can create an audience of those who are one step behind, pass it to Google Ads for a campaign setup, and serve them with a creative that represents those best topics and land them on the highest-converting landing page per your research.

This example isn’t “true” data enrichment – but it helps showcase how 1^st party data can be the enricher itself, helping to enrich the “prospect” data with the “donor” data. A use case like this is free as it doesn’t involve any 3^rd party data – get creative!

2. Donor Profile, Motivation, and Capacity to Donate

This use case is most frequently seen as the only “true” enrichment – finding 3^rd party data qualifiers to deepen your understanding of the donor and, therefore, better inform your targeting for new donors so you don’t boil the ocean reaching out to everybody and anybody. But this one is so broad that I usually like to further break it down into the following three sub-groups:

Community data, providing anonymized insight
Individual data, helping to understand the capacity to give
Individual data, helping to understand the motivation to give

Community data is frequently overlooked, but it’s insanely useful. It’s usually free, or at least significantly cheaper than individual-level - and more importantly in today’s world, it’s privacy-safe. Use publicly available Census data sets as the first qualifier before getting into individual-level donor enrichment. Census data can be used to go beyond zip code level, getting down to individual neighborhoods and, in some cases, blocks – becoming extremely accurate.

If you’re already using Census data or thinking about it - know this - I wouldn’t use it for Demographics, I t’s essentially useless. Instead, use it for psychographics – and to understand the communities your donors are part of. Gone are the days when political party affiliation was enough to determine somebody’s inner beliefs. I’ve found Census data to be most beneficial to highlight donor habits – such as retailers they prefer to shop at, cars they choose to drive, or churches they like to go to. As a marketer, these data points can help you understand the values they’re likely to stand for.

Armed with the above data from the Census database, your BI team would be able to form hypotheses that can be further validated with individual-level data.

But individual data is not created equal, either. To make more sense of it, I like to think of donors in terms of their reason to give versus the means, or capacity, to give.

Capacity is best shown by their existing spending habits. Several well-known data enrichment providers exist – Acxiom, Merkle, Liveramp, Data Axle, Alteryx, and likely a few others. Each of them would be able, given an email, name, phone number and address, to add 100+ dimensions, putting every individual into many buckets of income, beliefs, and affiliations. I find most of that data useless – it’s already modeled based on some black-box algorithms meant to sell me audiences to target down the line.

The instrumental piece is the data they get from the credit agencies and credit card providers – showing donors' spending habits, amount of discretionary income, and other non-profits getting a share of it. So, when signing with any of these companies, ensure you get this data.

3. Understanding a Person’s Motivation to Give

This is the most complex use case. We’re looking for the intangibles here – the experiences that formed them as a person, making them feel connected to the cause of your specific non-profit.

The data sources that are the most helpful are the ones showing your donors in real life – their Facebook feed and posts, their Google search history, or books they buy on Amazon. This data is hard to get in an aggregated way.

Social Listening tools can spit out the topics they most engage with on Facebook and media platforms usually have some “audience composition” report. By uploading your donor lists to Google, Target, or Amazon Ads, you can also get access to those.

Recap

Data enrichment is a complex topic and this ended up being a longer pos, so here’s a recap in a few bullets:

Data enrichment aims to help you better understand your current donors, allowing you to target prospective ones more accurately.
A good starting point is looking at your current donors in aggregate – using Census Bureau data and getting a sense of their habits and communities.
Validate your hypotheses by using individual-level data showing their motivations – the books they buy, the posts they write, and the queries they search for. This data doesn’t exist in aggregate but can be derived from advertising tools offered by websites we all interact the most with – Google, Amazon, Facebook, soon ChatGPT, and others relevant to you.
Motivation is good – but operating in the fundraising world, we need to make sure prospective donors have the capacity to give as well. Use your existing donors’ spending habits and credit bureau data from large Data Enrichment companies to set those rules.
After the above four steps, use your existing 1^st party donor data to determine the content your prospective donors need to see to donate – use your Google Analytics “Path Exploration” reports.

5a. *an alternative step is to skip bullet #4 and target everybody motivated to give but might not yet have the means because of age, income level, or other factors like immediate need/desire. This is precisely the purpose of your branding campaigns if you have a separate budget for this.

Great Messaging <> Great Insight

I received a great question from a reader of SPN last week and so thought I’d share an abridged answer here: how do you ensure that the Marketers you manage are creating great messaging?

Given the noise out there with so many products and orgs competing for eye balls, it's hard to stand out. There are lots of ways to create great messaging, but it all comes from the same place:

Great insight into your donors drives great messaging.

That's it. So then the question is, how do you get there?

We’ve considered the enrichment side of how we could arrive at a solution so let me take a slightly different tack.

This is classic Product Marketing territory. For me it means holistically understanding the environment around your donor, and then being able to connect the most differentiated value of your mission or product to their world.

That means your marketers must have a deep understanding of the donor as well as the demands and challenges they face, combined with a deep understanding of your own proposition, and finally the overall environment (for example, other orgs directly or indirectly competing for their attention and dollars).

When this comes together, messaging must be defensible (meaning, it's true) and must be differentiated (you couldn't slap a peer organization’s name on it and have the market believe it). Finally and most importantly, it must matter. The donor or prospective donor must care about your why's and how's.

Over time as a marketing and fundraising leader you develop a sense if a message meets these requirements, but the only real proof is testing it with your donor base and proxies for the base.

Proxies include mid-major gift officers and good street canvassers - people who are holding conversations with donors. Use them! It’s a good opp for cross-functional collaboration too. And then test with prospective donors who don't know they’re getting the new stuff.

Don't be afraid to get the message out there.

I'm not a huge fan of A/B testing messaging through ads or web pages in the early days. While you may see which one gets more clicks, you won't know why. Having real conversations allows you to ask the follow on questions that give you the insight needed to improve your messaging.

Now onto the fun stuff! >>

Good Reads this Week

IBM thread on rationalizing their content marketing.

Adtech isn't ready for California - referring to the US version of GDPR.

Twitter thread on how Carvana created 1.3 million hyper-personalized videos.

Apple is now publishing a transparency report for the App Store. Lots of interesting data, but a few things to call out: on a weekly average there are 650m visitors, downloading 750m apps, re-downloading 1.5bn, and installing 41bn app updates. And, all the take-down requests come from China.

Meta’s had quite a bit going on recently, including testing on a new ad format that helps users find relevant sales and promotions.

How Addressability Works in Streaming TV.

2023 State of Digital | LUMA Partners - great data and good commentary.

The Verge has a helpful explanation of Google’s plans for AI search (“Google executives … see it as a foundational long-term change to the way people search”).