Featured Project
BOWLING & ENTERTAINMENT
·
LIFECYCLE · REACTIVATION
3.7 Million Guests Went Quiet. I Brought Them Back for the Holidays.
An always-on reactivation program that reintroduces dormant guests across three brands in twelve weeks, timed to the busiest season of the year, without putting the inbox at risk.
§ The Challenge
The holidays are when this client's guests book parties, gather families, and buy gift cards. Yet about 3.7M opted-in guests had gone quiet. The brand only mailed contacts who had clicked in the last 120 days, so everyone else had dropped out of view: millions of past customers the brand had already paid to acquire, sitting idle going into the most valuable weeks of the year.
Reaching them again carried real risk. A sudden surge of mail to unengaged addresses can trigger bounces, complaints, and spam filtering, and the client was already rebuilding its Gmail sender reputation. A misstep would hurt the active guests too.
§ The Approach
The client's first idea was to export each brand's segment and manually carve it into 15% slices every week. It would work on paper, but it meant weekly manual effort, a high chance of error, and no guarantee a guest wouldn't be mailed twice or skipped.
Instead of splitting lists by hand, I made the platform do the throttling. One entry segment feeds three brand branches, and each branch releases a new, capped batch every seven days, with creative testing built in from week one.
Client
National bowling and entertainment operator with three brands
Audience
~3.7M inactive, opted-in contacts
Timeline
12 weekly batches, early October through late December
Constraint
Weekly inactive sends capped at 15% of each brand's active volume
Deliverable
Program architecture, volume model, creative test plan, build options, and launch
§ Where Creativity Came In
Same goal. A completely different build.
What was requested
× Export three brand segments every week
× Hand-cut each into 15% slices in a spreadsheet
× Re-import and schedule each slice manually
× Risk of double-mailing or skipping guests
What I built
→ One entry segment, three brand branches
→ The platform does the throttling automatically
→ A new batch released every 7 days, hands-free
→ Every guest mailed exactly once
I explored several routes before landing on the final design. A one-time ranking field would have given exact weekly counts, but it needed a manual export and import, and the platform has no native last-engagement date to rank by. Engagement tiers solved the ordering problem but created uneven tier sizes. Each idea sharpened the brief.
The final design uses tools already native to the Automation Center: segment nodes, A/B split nodes, and staggered wait nodes. No exports, no spreadsheets, and every guest is mailed exactly once.
§ The Plan I Delivered
One entry. Three brands. Twelve weekly releases.
The program runs entirely from native nodes, so it needs no weekly exports. Each brand's branch releases a new batch every seven days through staggered wait nodes, and the creative test runs across all three brands for the first four weeks.
§ The Volume Model
A straight 12-week split works for Brand A, but pushes B and C past the guardrail.
Brand
Weekly active sends
Inactive contacts
Batch over 12 weeks
Share of weekly send
Batches needed for ≤15%
Brand A
~670K
~1.2M
~100K
14.9%
12
Brand B
~280K
~585K
~49K
17.4%
14
Brand C
~430K
~1.9M
~158K
36.8%
30 (over 26 limit)
Each batch is sized against the brand's own weekly active send from the prior week. Batch counts round up so no week exceeds the cap. The A/B split node supports a maximum of 26 branches.
§ Creative That Earns Its Spot
The winner is locked in before Black Friday.
Two versions per brand, six emails in total, are copied into each weekly batch. After four weeks, the version with the stronger click rate becomes the creative for every remaining batch.
§ Two Paths for the Split Limit
A clear choice, not a mid-build surprise.
Twelve batches fit comfortably inside one A/B split node. The question was Brands B and C, which need more batches than the 12-week window allows to stay under 15%, and in Brand C's case, more than a single split node can produce.
Path 1 · Speed
Accept higher weekly volume
A · 12 wks
B · 12 wks
C · 12 wks
All three brands run 12 batches over 12 weeks and finish together before year-end, entirely within the holiday window.
Trade-off: Brand B rises to ~17% and Brand C to ~37% of its weekly send, adding deliverability risk during the busiest weeks.
Path 2 · Safety
Add a second split layer
A · 12 wks
B · 14 wks
C · 30 wks
Nesting a split beneath Brand C's first-level branch breaks past the 26-branch limit and holds every batch at or under 15%.
Trade-off: A larger build, and a program that continues past the holidays for Brands B and C.
◆ Why it matters for the holidays
Launching in early October puts 12 weekly batches in market before year-end, reaching hundreds of thousands of lapsed guests per brand while they're planning parties and buying gifts. Because volume rises steadily instead of all at once, the client reconnects with them without putting holiday inbox placement for its active guests at risk.
Results
Launched on schedule, right into the busiest season of the year.
I designed, built, and launched the reactivation automation ahead of the holiday window. It now works around the clock to bring dormant contacts back onto the client's active lists, where they can be converted into repeat customers during the weeks that matter most.
3.7M
Dormant contacts back in play
On a defined path back to the active audience, instead of sitting unused.
1
Always-on automation
Replaces a weekly export, split, and re-import cycle across three brands.
12
Weekly releases
Spanning October through the December peak, with no manual upkeep.
§ From dormant contact to repeat customer
01
Dormant
Inactive 120+ days, invisible to regular sends
02
Reintroduced
Released in a capped weekly batch
03
Re-engaged
04
Repeat customer
Books, throws parties, and buys gift cards
Revenue reach at peak season
Hundreds of thousands of past guests per brand are back in front of holiday party, booking, and gift card offers.
Hours back every week
The team no longer exports, slices, or re-imports lists. The platform handles the throttling.
Deliverability protected
Brand-level volume caps, an exit check, and a gradual ramp keep active guests' inbox placement safe, particularly at Gmail.
Smarter creative, automatically
The test winner rolls out to every remaining batch before Black Friday, with no extra build.
§ How Success Is Measured
Revenue that was already on the list.
Every reactivated guest is a customer the brand already paid to acquire. Success is tracked week over week on four signals.
01 · Reactivation
Guests who click and rejoin the active audience
02 · Engagement
Click rate by brand and creative version
03 · List health
Hard bounce and complaint rates held within safe thresholds
04 · Inbox
Stable inbox placement for active guests, particularly at Gmail
