Read text version
Let's try this again.
β
Welp, you got a peek behind the curtain earlier today. A real one β the actual skeleton of our newsletter before any content lives in it. If you've ever wondered if we actually use Braze templates, now you know.
We committed every CRM marketer's worst nightmare β and maybe their most universal rite of passage: sending a test email to the full audience. This is what happens when you're training a new employee on Braze and forget to unschedule your pretend-send.
We'll call this an example of radical transparency. Or a very public QA failure. Lesson learned.
Here's what we actually meant to send you.
What we're talking about this month:
π§΅ Your loyalty score is lying to you. Here's what to track instead.
π Braze tips & tricks
π οΈ The most relevant updates from Brazeβs recent releases
πͺ‘ Resources you can use to hone your craft
STITCH IN TIME
Your loyalty score is lying to you β here's what to track instead π§΅
Marketers have spent years chasing real-time data and a real-time view of the customer. We got there. Then we stopped.
Now an entire industry is optimizing around a single number. Health scores, predictive conversion scores, engagement scores β they're all calculated at one moment in time and treated like a relationship summary. Layer in recency, frequency, monetary value, purchase history, tier status, email clicks, app logins, redemption rate. Most platforms call that a complete picture of member health. In a single moment, it is. A relationship is a different thing entirely.
Same score. Different story.
Take two members, both sitting at an 82 out of 100 today. Both land in your loyalist segment. Both get the same communications. One was a 97 last week. The other was a 70. Those members are moving in completely opposite directions, and that score will never tell you that.
The member climbing from 70 to 82 is building something. Reinforce the momentum. The member dropping from 97 to 82 has stopped opening push, missed their usual Friday visit, let their points expire. The relationship is weakening β quietly β and that 82 is actively misleading you.
Because they share a score today, both signals are invisible. That's where loyalty strategies fall flat. The score tells you where someone landed. It says nothing about where they've been or where they're heading.
What we call the Relationship Thread
A member relationship isn't a moment to be scored. It's a thread to be followed. Members don't move through lifecycle stages in a straight line β they live inside ongoing relationships that are constantly strengthening or weakening. Scoring them once a week and slotting them into predetermined segments misses the whole thing.
What you actually need to watch, over time:
Pattern shift β Is this member's behavior changing relative to their own baseline, not your average member's?
Value exchange β Do they feel like the relationship is worth it? Are they actually engaging with benefits?
Presence erosion β Are they quietly pulling back even before purchase behavior changes?
When one of those moves, something is changing. That movement is the signal. Not the score it eventually produces.
Reading relationships, not dashboards
The brands getting this right aren't chasing a better model. They've changed the question β from "what is this member's score today" to "what has this relationship been doing for the past week, month, 90 days."
They're catching drift before it becomes departure. They're showing up in ways that fit where a member actually is, not where a formula placed them.
The good news: most programs already have the data to do this. If you're measuring member health today, start with one addition β track whether the score is moving up or down each time it's calculated. That single layer of context changes what you do next more than any model refinement will.
Over time, build toward the full picture: pattern, value exchange, and presence. Together, those three tell you the story a score never could.
Because a relationship carries everything that happened before this moment. A score doesn't.
- Kayla Siegmeier
Loyalty Specialist & Avid Documentary Watcher
PIN IT
Braze tips & tricks π
Let AI pick the offer!
Manually mapping customer segments to offers? Braze Agent Console enables you to drop an AI agent directly into your Canvas, allowing it to read each user profile as it passes through. Attributes like loyalty status, price sensitivity, and AOV can be analyzed to automatically select the best offer at send time.
The result? One Canvas with infinite personalization and minimal rule management required.
This means less time managing decision trees and more time focusing on strategy. The agent handles the decision-making based on the guardrails you set.
- Devin Howard
Solution Consultant & Ranked Top 500 in the World on Madden (in 2019)
PATCHWORK
Decoding Braze product updates π οΈ
Content Optimizer now available for SMS/MMS/RCS [BETA]
Braze's Content Optimizer agent now works across SMS, MMS, and RCS β not just email. It runs multi-variate tests at scale and learns over time which content combinations perform best, so more of your messages are automatically optimized without manual intervention.
Why it matters: Most teams running SMS are still doing single-variant sends or manually setting up A/B tests. Content Optimizer changes that. It brings the same AI-driven testing infrastructure from email into your highest-engagement channels β and it compounds. The system gets smarter with every send, which means your SMS performance improves without requiring more resources or more manual work from your team.
New Dashboard for Push
A new Push-specific dashboard in the Analytics section of your workspace that aggregates performance across all campaigns and canvases in one view β no more piecing together reports from individual sends.
Why it matters: Push is often the hardest channel to get a clear read on because the data lives fragmented across dozens of campaigns. This dashboard gives you the channel-level view your team has probably been building manually in spreadsheets. If you're reporting on Push performance to stakeholders or trying to spot trends across your program, this is the starting point you've been missing.
Geolocation fields in catalog selections are now GA
Catalog rows can now include a Geolocation field type (latitude/longitude) for things like store locations, restaurants, or events. You can target users within a set radius of any catalog entry, entirely within Braze.
Why it matters: Location-based personalization used to require a data engineering lift to pull location data, build segments, and keep them fresh. This moves that capability directly into your campaign workflow. If you maintain a catalog of store locations, venues, or service areas, you can now show each user the most relevant one based on where they are β no custom code, no complex audience builds, no waiting on the data team.
Partnerships Updates
There are some new partnerships / integrations worth mentioning:
Better Email - Collaborative tool that enables marketing teams to build emails based on an email design system.
Chord - Chord is a CDP specific for eCommerce retailers. With this integration, you can trigger campaigns based on events and updates that happen within Chord directly to Braze
- Cat Mears
Marketing Strategist & Probably Has Multiple Foster Pets Right Now
HONE YOUR CRAFT
Our top resources this month πͺ‘
π§± New Agentic CDP just dropped. Databricks announced a new, agentic Customer Data Platform (CDP) β Databricks CustomerLake β today at the Data + AI Summit in San Francisco. Stitch has been named a launch partner. More to come, but we think this is going to be huge for marketers.
Learn more about CustomerLake CDP >
π©Ί Diagnosing campaign slowdowns. Most CRM teams assume slow campaigns mean they need more headcount β but the real bottleneck is often architectural, not human. Ryan Wilkins, one of our solution leads, breaks down how to diagnose whether your slowdown is a platform problem, an operational model problem, or both β and what to actually do about it.
Read the Martech Cube Article >
π What we're reading/watching
The A/B Test Nobody Asked For - Really Good Emails
Introducing CustomerLake: The Agentic CDP embedded in Databricks - Databricks
LOOSE ENDS
This month's VIP (very important pet) πΎ
Catch up on past editions of Cutting Edge >
Stitch, 830 Massachusetts Ave Floor 3, Suite 9, Indianapolis, IN 46204, USA
Unsubscribe Manage Preferences
β