Marketing Challenge: Our Gmail audience--50% of our database--were having difficulty receiving communications.
Our sending practices were aligned with Gmail’s Bulk Sender Guidelines, as well as Oracle best practices:
- Target audience built off engaged contacts
- Branding & Deliverability properly configured
- Proper throttling when volume ramped up
Our campaign/email metrics were on par with our previous marketing automation platform’s benchmarks (30% Open Rate, 3% Click Through Rate), so from an aggregate perspective, there weren’t red flags.
Unbeknownst to us, we were having a deliverability issue with Gmail, which didn’t come to surface until numerous Gmail contacts reported they hadn’t been receiving their communications.
Again, looking at aggregate level reporting (Email/Campaign Analysis Overviews), things looked fine. We needed deeper insight into our data for our Gmail audience.
Step 1) Checked Reputation
Step 2) Leveraged Insight reporting for deeper analysis.
We customized the email analysis report to now include email domain as a column (Insight > Catalog > Shared Folders > Email > Email Analysis Overview - click Edit).
Under the Criteria tab, locate Email Address Domain field under the Contact folder/category. Drag field over next to Email Name. You can further customize, adding additional fields, removing fields, moving columns, if desired.
IMPORTANT: when ready to save this report, select SAVE AS (floppy disk with pencil icon), as you would not want to overwrite the master report.
After running the report, Gmail’s activity metrics were drastically lower than other inbox providers. This told us one of two things, either A) our Gmail audience is collectively uninterested (which historical benchmarks prove otherwise), or B) Total Delivered did not equate to communications physically landing in one’s inbox.
Step 3) Developed an Action Plan
With option B being obvious, we put forth an action plan to re-establish our reputation with Gmail, which included:
- Temporarily halting outbound communications to our Gmail contacts.
- Further filter our Gmail audience to only market to our most active contacts.
- Introduce flow logic at the campaign level to route Gmail contacts to dedicated email steps, which we customize with extensive throttling.
- Increased domain-level report frequency to identify red flags earlier.
Also, we configured our sending domains on Google’s Postmaster Tools due to Gmail not offering a traditional feedback loop via Oracle’s Branding & Deliverability configuration. If you have a Gmail account, there are two options:
A) Submit a My Oracle Support ticket to have the Oracle Deliverability Team configure for you. You’ll need to supply domains/subdomains, your account short name, and the email address/es of those needing view rights.
B) Configure Postmaster Tools yourself (it’s super simple, see below):
- Navigate to https://gmail.com/postmaster/ and click Get Started.
- Enter the domain/sub-domain you use to authenticate your email.
- Copy the TXT record and add it to your DNS configuration. Once configured and propagated, click “VERIFY.”
- Once verified, give it a couple of days of high-volume sends in order for data to be reflected in Postmaster Tool dashboards.
Step 4) Rolled Out Action Plan
Delivering to our most engaged Gmail contacts, as well as throttling as to not set off Gmail’s algorithm alarms helped re-establish our baseline reputation. As we saw our deliverability metrics increase, we slowly began to expand our time span of engaged Gmail contacts, e.g. started with 0-2 months, increased to 3-4 months, etc. Anytime we saw Gmail react negatively, we’d pump the brakes. When reaction was positive, we’d step on the gas a bit.
The custom email overview with domain level breakdown was key as we ramped up our volume. As our volume increased, we were able to better leverage Google Postmaster data, including these insightful dashboards (not all-inclusive):
Spam Rate - highlights the volume of user-report spam vs. email sent to inbox (DKIM authentication required)
Domain & IP Reputation - insight as to whether Gmail spam filters might mark emails from Domain or IP as spam or not.
Feedback Loop - shows average spam rate across all identifiers flagged on a given day and the number of unique identifiers flagged by FBL per day (when applicable) over time.
Results/Conclusion: It took us roughly 4 weeks to get back into Gmail’s good graces. We’ve continued to monitor the reports/dashboards above, in addition to automating other Insight reports, Spam Unsubscribe and Bounceback with Messaging, to assist with our on-going reputation efforts.
It’s not foolproof, but we’re proactively putting ourselves in the best position to land in our audience’s inbox. As the adage goes, “it takes years to build a reputation and just minutes to ruin it.”
Cloud Course Influencers:
- Eloqua Fundamentals: Fundamentals of Segmentation/Campaign Canvas
- Profile & Target: Data Cleansing
- Insight for Reporters/Analyzers