46. Some Personal News

How to Use AI in Your Analysis (Fundraising and Marketing) + The Good, Bad & Alternatives to GA4 and Finding a Competitive Edge in 2H 2023

Jul 02, 2023

Happy Sunday. For readers in the US it’s a longer weekend, so I thought I’d share a longer edition of SPN. 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. Wishing you all a very productive and healthy start to the second half of 2023!

Let’s dig in!

Changing of the Guard: UA → GA4

The end. After multiple delays and countless “reminder” nudges in every medium imaginable, for those of us in the data every day - we get to rid our lives of this image. Finally.

Google Analytics 4 and Universal Analytics: key differences

GA4 is the only (Google) analytics and measurement option now, and in light of not having much choice, most marketers will adopt it fast. For those who haven’t done it yet – for the most part, you have a pleasant surprise awaiting.

So much resistance to the migration is no surprise – Universal Analytics has been around, virtually unchanged, since 2013. And before that, there was essentially an identical Classic Analytics.

That makes 20 years total of using the same product almost every day – an entire career for most marketers who learned the ropes of SEM, Programmatic, Product Analytics, Attribution, and over time, Leadership and Coaching – all while using this one tool. That’s what it was for me, at least.

It wasn’t a good tool in the last couple of years. New/Return users’ statistics were off for years due to inaccurate cookie measurement; Cross-Device attribution was a joke since day 0; and, the ability to do meaningful content engagement analytics was non-existent.

As I explored GA4 over the last few weeks I felt like unboxing a new iPhone after using my Blackberry for years. Granted when I first got the iPhone 3 I went right back to my Blackberry for a while - it took a lot of getting used to - but I began to like it more by the minute.

GA4 works much faster. Maybe this is because the servers are not overloaded yet – but it’s simply more pleasant to interact with. I find myself waiting for the report to load much less than I used to.
There are new use cases. In the context of the Donor Journey, behavior on the website is as essential as paid media impressions – and this goes well beyond page views previously available in UA. As I was exploring what Events are, I kept thinking that it should now be possible to build custom definitions of Donor Journey Stages and use them to create Audiences for marketing and reporting. And all of this – across devices.
Takes care of the legalities. Google was recently forced to update its privacy terms for certain products, ridding them of the “Special Service Provider” status and putting more pressure and responsibility on the Advertisers. I support increasing customer data controls and privacy regulations – but Universal Analytics was built for the 2005 version of 3^rd party cookies when the motto was to “track everything everywhere all at once.” UA was, therefore, acting “against” you in compliance matters. Marketers are not legal experts and using GA4 will automatically prevent lots of exposure.
Built for AI. The Insights section of GA4 is still far from functional – but it’s already much better than it used to be in Universal Analytics. The prospect of having a ChatGPT-like model built into the Web Analytics tool is not what I expected – but it makes so much sense.
Well supported. Google has made a bit of a fool of themselves by pushing the migration back as many times as they did. But the good part is they have no choice now – all their eggs are in the GA4 basket. There’s no hedging. Moreover, given that the enterprise version of GA4 is two times cheaper than what UA360 used to be and noting the customer attrition in the last year, Google will be desperate to acquire lots of new, smaller organizations onto the platform to maintain the revenue. This should result in better support and faster updates to the product.

GA4 could be better. Some of the metrics still make little sense. “Bounce Rate,” which is now defined differently from what it used to be… and it seems to change every time you change the dimensions, making it useless. Several popular functions – like Attribution Models - have been killed off. And the learning curve with the new UI feels steep.

But it’s not bad, either. Different, somewhat innovative, and more up to date with the reality of marketing in 2024 and beyond.

One caveat, however. What is undoubtedly wrong is the experience with GA4 if you allowed Google to migrate your properties automatically. These automatic migrations combine the worst of both worlds – vastly inaccurate data, carrying over the structure from your legacy UA setup, and the new interface that already makes it hard to find the reports you need, even if they are accurate.

If you’ve waited until now and are currently using the automatically-migrated version of GA4 – put aside time to re-implement it from scratch ASAP and find an external partner to do so if you don’t have at least two web development experts in-house to be devoted to this full-time.

This is also a great chance to align your data structure with the Donor Journey your organization wants to implement. You can’t manage what you can’t measure!

And if you feel GA4 is not for you – there are alternatives. Adobe and Google are not the only players in the field. I’ve found the following tools to be handy and competitive:

Amplitude – a great option if your primary use case is Mobile App or Product analytics, with a focus on helpful engagement metrics.
Snowplow – the most democratized open-source platform, allowing for easy Cloud integrations and custom setup. The closest option to building an Analytics platform in-house while still having guides, best practices, and community to support.
Mixpanel – a personal favorite from the closest, suite-like alternatives and focus on the Audiences.

2nd Half 2023 > Competitive Edge

As we turn a corner into the second half of 2023, I’m looking for a competitive edge. Where I’ve landed? Double down on data and conversion tracking.

Our ability to measure and drive real business results (aka qualified, prospective donors and advocates, donations and hand-raisers, and actual ROAS) is what’s going to set our Orgs’ performance apart.

It’s surprising how often I see suboptimal tracking setups.

Whenever I see conversion tracking best practices applied, there’s an average uplift of 8-20% of conversions measured:

• Use the Google Ads Tag for primary conversion goal(s)
• Use enhanced conversions and consent mode to track more
• Use data-driven attribution to attribute better
• Use Events from GA4 as backup (secondary) conversion goal(s)

And this is only the bare minimum.

How many of the following things have you heard about this year, but haven’t really tested yet?

• Offline Conversion Tracking (OCT)

• Profit on Ad Spend (POAS)

• Conversion adjustments

• Conversion modeling

• Server-side tagging

These advanced configurations allow you to measure more and drive REAL business results (aka qualified prospective donors, donations and actual ROAS), not just sub-optimal donations from low value donors (pMax, anyone?).

What’s your Org’s biggest competitive edge in H2 2023?

It’s understanding how to measure more and drive real business results.

Once you have the foundation in place to measure and drive real business results, you can fully leverage automation to maximize your ROAS with:

• Advanced consolidated campaign structures

• Advanced Performance Max strategies

• Advanced broad match testing setups

• Advanced DSA page feed setups

• Advanced RSA tests

Now’s the time to begin amping up testing and discovery. Q4 is less than 12 weeks away. Make it happen!

How to Use AI in Your Fundraising and Marketing Analysis

I’ve been exploring how AI can improve analytics and augment operations as a non-profit.

Contrary to some fears, AI isn’t here to take away jobs - it’s here to enhance them. The challenge (and opportunity) is for non-profit operators like us to assess impact areas and use this technology responsibly.

AI for analytics isn’t one-size-fits-all. So I’m going to take a fresh look at each type of non-profit analytics and surface how AI can enhance each. Plus, I’ll make it more concrete with some real-world examples.

Recap: The 4 Pillars of Business Analytics

There are four main types of business analytics: Descriptive, Diagnostic, Predictive, and Prescriptive. Each one has its unique role, challenges, and opportunities. And, each one can benefit from the right application of AI.

Descriptive Analytics answers “What happened?” by providing a clear view of past events.
Diagnostic Analytics goes a step further, asking “Why did it happen?” by analyzing data across variables to reveal underlying patterns.
Predictive Analytics uses historical data to predict future events, answering the question, “What’s likely to happen?”
Prescriptive Analytics ties everything together, using insights from the other three types to answer the question, “What should we do?”

AI can enhance each type of analytics but its role varies depending on the specific requirements of each stage.

I’m going to break down each type - 2 today and 2 next week. Buckle in for today’s Descriptive and Diagnostic ride.

AI in Descriptive Analytics: The Pelvic Floor

You’d use Descriptive Analytics to get a clear view of past events and to lay the foundation for more advanced concepts. It’s a basic yet critical step. Surprisingly, AI can be extremely useful right from the start.

To make it more concrete, suppose we’re dealing with a common issue - donor churn. We have churn rate data from the past six months as follows:

This table looks innocent enough but depending on the data it could have been a lot of trouble to get here.

Here’s how AI could help us in this context:

Visualization: This table is a good start, but wouldn’t it be nice to see the trend over time? Generative AI can suggest and create appropriate visualizations based on the data we have.
Exploratory data analysis (EDA): For more advanced descriptive analytics, Generative AI could suggest a structure for our EDA and even write out the code for this.
Q&A: Instead of just showing a table, we could use NLP to ask questions about our data like “What was our churn rate last month?” - far more entertaining!
Data Narration: To communicate our findings, we can use NLP to create a precise narrative of the observed insights. This is a great help – writing out table summaries is tedious.
Unlocking new data: Using computer vision or speech-to-text services, we can turn unstructured data such as donor call transcripts into actionable data or maybe we needed to process some cancellation letters to create the report above?
Data access: With the help of Generative AI we could write, debug and document complex SQL queries that were perhaps necessary to get this data.
Data preparation: Generative AI can help us to create data preparation scripts, or suggest suitable data cleaning methods.

Key AI Archetypes for Descriptive Analytics

For descriptive analytics, potentially the following AI Archetypes are at work:

Generative AI
Computer Vision
Speech Recognition
Natural Language Processing (NLP)

Next, let’s move to Diagnostic Analytics.

AI in Diagnostic Analytics: Reduce Time to Insight

Diagnostic Analytics takes us a step further by answering “Why did it happen?”.

Continuing with our example, you’ll have noticed the upward trend in donor churn rates. But we’re not content with just knowing the “what” - we want to know the “why.” So, let’s dig deeper and enrich our churn data with more donor information, such as age groups.

Here’s what we find:

This table shows that younger donors are more likely to churn and this trend increased in Q3.

So how could AI have assisted in this diagnostic process?

Pattern Detection: We could use tools like Auto ML to check for important “features” (i.e. key influences) that drive churn. In this case, “donor age” might have emerged as the most important attribute among a plethora of other variables.
Q&A: Just like above, NLP could provide us with more intuitive ways to interact with our data, e.g. “What was the churn rate for donors below 30 years”.
Data Narration: NLP could help to automatically generate explanatory text to communicate our findings.
Methodological support: Generative AI could help guide us through analysis frameworks (e.g., 5-Why, Issue Tree) so we can do more effective number crunching. It could also help us design and set up experiments like A/B tests for further data collection.

Key AI Archetypes for Diagnostic Analytics

For Diagnostic Analytics, we’d primarily leverage:

Automated Machine Learning
Generative AI
Natural Language Processing (NLP)

Next week we’ll leave the past behind and get into Predictive Analytics e.g. churn prediction.

AI is not about replacing us - it’s about augmenting our capabilities and allowing us to operate at a far more effective and efficient level - to the benefit of our respective missions, and by extension our fundraising and program efficiency ratios.

Now onto the fun stuff!