Hi everyone, I'm fairly new to marketing automation and just helped my team roll out Braze for our user base. We're about 50,000 users now, which feels huge to me! I'm here to learn from your experiences.
We followed the setup guides, but we definitely hit some snags when we actually turned things on for real people. I was hoping to hear what common pitfalls you've seen. For us, the biggest issue was with our customer data sync from Salesforce. Some custom fields for user preferences didn't map over correctly, so a segment got a campaign totally meant for another group. 😅
Also, our initial email deliverability took a hit in the first week. We warmed up the IPs, but I'm wondering if there's something specific about the initial Braze onboarding that we might have missed? Any guidance on checking the CRM sync reliability or workflow gaps early on would be so appreciated. I want to make sure our foundation is solid as we grow.
Welcome, and congrats on the rollout. It's a big step, and encountering these hiccups is more common than you'd think.
The data sync issue you described is a classic one. Even with the guides, the devil is often in the details of your specific data schema. For a foundation check, I'd suggest creating a small test segment of known users and running validation campaigns that don't go live. Use them to verify that the field values Braze shows match your source of truth. It's a manual step, but it catches mapping errors before they affect 50k people.
On the deliverability dip, warming the IPs is correct, but with a new Braze account, your sender reputation starts from scratch, too. A sudden spike in volume from a new sending domain can trigger filters, even with warmed IPs. A more gradual ramp-up in sending volume, maybe starting with your most engaged users, can help. Did you notice if the hit was primarily with one ISP, like Gmail or Outlook? That can point to specific authentication issues.
Let's keep it real.
Oh yeah, the Salesforce custom field mapping is a notorious trap door. We had a similar mess where a single-select picklist synced as a text field, which meant all our conditional logic in Braze workflows just... didn't work. The real killer was it *looked* fine in the dashboard until you saw the actual values.
For a sanity check, I'd pull a report of, say, your top 20 most important custom fields from Salesforce and then use Braze's user export to compare. Do it for a sample of 100 known users. You'll spot mismatches in data type or formatting immediately. It's boring but it saved us from sending a loyalty campaign to people who'd just churned.
On deliverability, warming the IPs is one thing, but did you also gradually increase the volume of your sends? Even with a warmed IP, hitting 50k users right out of the gate can look like a blast to ISPs. We staggered our first campaigns over two weeks, starting with our most engaged users first to build positive engagement signals.