Overview
In ServiceTitan, this playbook is for marketing managers who already have multiple Marketing campaigns running and want to make them work harder together. It covers four decisions that come up once a schedule is full: how to refine segments without breaking live campaigns, how to test variants, how to plan seasonally on the Campaign Calendar in Marketing Pro Enterprise, and how to compare campaigns against each other in Marketing Analytics and Atlas.
For the click-by-click procedures referenced below, see Build and manage multiple campaigns in ServiceTitan Max. This article focuses on the frameworks and judgment calls — not the navigation.
Who uses this feature
Marketing managers, business owners, and other marketers on your team
Primarily benefits Residential Service and Replacement business types and Commercial Service and Replacement business types
Applies to all trades
Feature configuration
This article assumes you have launched at least one Autopilot campaign and have multiple campaigns running. If you haven't, see Build your first automated campaign in ServiceTitan Max and Build and manage multiple campaigns in ServiceTitan Max first.
The Campaign Calendar requires access to your Enterprise Hub account.
Things to know
The supported pattern to test changes to campaigns is to clone a Pro Campaign, change one variable, and compare results in Marketing Analytics.
Only Pro Campaigns can be cloned. Autopilot campaigns become Pro Campaigns once launched, so they can be cloned from that point forward. Tracking Campaigns cannot be cloned.
Every active campaign adds inbound volume that lands in Contact Center, online bookings that flow through Scheduling, jobs that Dispatch assigns, and estimates that Field generates in the home. The frameworks below assume those connections — measurement, capacity planning, and decisions all read across the suite.
Atlas attributes results across the ServiceTitan Max suite. Comparison questions like Which of my campaigns is generating the most revenue? are how you measure across campaigns rather than within one.
Best practices
Refine before you add. When a schedule is already full, the optimization move is tightening existing segments — not adding more campaigns. New campaigns on top of overlapping segments compound the problem.
Change one thing at a time. When you clone a campaign to test a variant, change only one variable. Two changes mean you can't attribute the result.
Plan the year, not the week. Open the Campaign Calendar in Enterprise Hub at the start of each quarter and look at the next three to six months. Seasonal campaigns work better when they're sequenced against the always-on automated ones, not slotted in last-minute.
Compare campaigns, don't review them. Marketing Analytics tabs and Atlas are most useful when you ask "which of these is winning" rather than "how is this one doing." Cross-campaign comparison is the optimization view.
Let data accumulate before deciding. A cloned variant needs enough time for the campaign sequence to complete at least once before the comparison is fair. Pausing too early kills variants that would have won. The ServiceTitan recommendation is 4 weeks of time.
Use cases
A residential HVAC marketing manager runs an always-on Unsold Estimates automated campaign year-round and overlays seasonal one-time campaigns — Spring Tune-Up in March, Fall Furnace Check in October. The optimization works each quarter: open the Campaign Calendar, confirm seasonal sends don't collide with always-on follow-ups, and tighten the Unsold Estimates segment's Last Email Date exclusion if it does.
A commercial plumbing contractor wants to know whether a new subject line outperforms the current one on a Memberships Expiring campaign. The marketing manager clones the campaign, changes only the first email's subject line, gives both four weeks to run, and uses Marketing Analytics > Booked Jobs plus Atlas to determine which version generated more revenue before keeping one and stopping the other.
A multi-trade business operating across four regional tenants uses quarterly Campaign Calendar reviews to coordinate sends. The marketing lead spots a November gap in one tenant's schedule, launches an Enterprise Hub campaign to fill it, and staggers two overlapping Memberships Expiring campaigns in two other tenants so customers don't receive both in the same week.
Refining segments without breaking live campaigns
When the schedule is already running multiple campaigns, segment changes carry risk. The framework below decides what to do based on what else depends on the segment.
For the procedural steps — opening Retention Segments, adding exclusion filters, saving the segment — see Build and manage multiple campaigns in ServiceTitan Max, Step 2.
Pick the right action for the segment
For every segment that needs a change, choose one of three actions based on what's live:
Reuse the segment as-is when the targeting is still right for every campaign using it.
Refine the segment in place only when no other live campaign depends on it. Editing an active segment changes who receives content from every live campaign using that segment — this is rarely the right call when more than one campaign is running.
Clone the segment and modify the clone in every other case. Cloning lets you make the change you need without disturbing the live campaigns pointed at the original.
Caution: Editing an active segment changes who receives content from every live campaign using that segment. Clone the segment instead, then point the new campaign at the clone.
Use exclusion filters to prevent over-messaging
When several campaigns target overlapping groups, the same customer can land in two or three audiences in the same week. Exclusion filters on the segment are how you prevent that. The three patterns that come up most:
Last Email Date — Exclude customers who received marketing email in the last 7 to 14 days. Use this when an always-on automated campaign and a seasonal one-time campaign target similar customers.
Scheduled Job Date — Exclude customers with a job on the books in the next 30 days. Use this when you're following up on estimates or memberships and don't want to message customers who are already booked.
Last Call Date — Exclude customers who recently called in through Contact Center. Use this when you want re-engagement campaigns to skip customers your CSRs have already spoken to.
Use the right date filter type
Match the filter to the campaign type:
Dynamic date filters for automated campaigns, so prospects enter the segment as soon as they qualify.
Static date filters for one-time campaigns where you want a fixed list at send time.
Tip: Atlas can refine a segment for you in plain language. From the segment or campaign builder, click Edit with Atlas and describe the change — for example, Make this segment residential customers only or Exclude anyone who booked a job in the last 14 days.
Testing campaign variants
Marketing does not include a testing feature. The supported pattern is to clone a working Pro Campaign, change one variable, and compare the original against the variant in Marketing Analytics.
For the procedural steps — opening Campaign Manager, clicking More > Clone, naming the clone, saving — see Build and manage multiple campaigns in ServiceTitan Max, Step 4.
Decide what to test
A useful test changes one thing the data can resolve. Pick a variable based on the question you're trying to answer:
Test audience filters when you want to know whether narrower or broader targeting drives more booked work. Example: clone an all open estimates campaign and narrow the clone to open estimates over $5,000.
Test subject lines or email copy when open rates or click-through rates are flat. Change only the subject line of the first email in the sequence — not the body, not the second email — and keep everything else identical.
Test channel mix when you want to know whether SMS adds incremental bookings on top of email, or whether direct mail performs against email for the same audience.
Test tracking numbers when you want Contact Center to attribute inbound calls separately to the original and the variant. This is what makes a side-by-side comparison possible on the Calls tab in Marketing Analytics.
Rules that keep the test fair
Change one variable. If you change the audience and the subject line at once, neither result is attributable.
Leave the original running. Don't edit or stop a working campaign while you test against it. Editing a live campaign limits what you can change and risks disrupting the customers already enrolled.
Give the test time. Let the cloned variant run long enough for the campaign sequence to complete at least once. For a single-email campaign that may be days; for a multi-step Unsold Estimates sequence it may be several weeks.
Use the same stop trigger on both. If the original stops when the estimate is sold or dismissed, the clone should too. Different stop triggers change who completes the sequence and skew the comparison.
Tip: When cloning to test a new variant, leave the original running. Compare performance side by side in Marketing Analytics before deciding which to keep.
Decide the winner
Once both campaigns have completed at least one sequence, compare them on the metric the test was designed to answer:
A subject line test is decided on the Booked Jobs tab if booking is the goal, or by asking Atlas which version converted more opens to clicks.
An audience filter test is decided on the Revenue tab — a narrower audience often has lower volume but higher revenue per send.
A channel mix test is decided by comparing total booked jobs across both campaigns, not per-message engagement.
A tracking number test uses the Calls tab to attribute inbound volume to each variant.
Once a winner is clear, stop the losing campaign and let the winner continue.
Planning seasonally on the Campaign Calendar in Marketing Pro Enterprise
The Campaign Calendar in Marketing Pro Enterprise is where seasonal planning happens. The framework below applies to opening it once a quarter and using it to coordinate the next three to six months, not to fill in this week.
For the procedural steps — signing in to the Enterprise Hub, opening Marketing Pro Enterprise, navigating to the Campaign Calendar, applying filters — see Build and manage multiple campaigns in ServiceTitan Max, Step 6.
Layer always-on and seasonal campaigns
Multi-campaign schedules work best with two layers:
Always-on automated campaigns in the base layer — Unsold Estimates, Memberships Expiring, Aging Equipment, Idle Account, Recurring Service. These run continuously and enroll customers as they qualify.
Seasonal one-time campaigns layered on top — Spring Tune-Up, Fall Furnace Check, Holiday Promo, Storm Response. These send once at a specific moment when demand is highest.
The Campaign Calendar's monthly view shows both layers together. The planning question each quarter: where do the layers collide, and where are there gaps?
Look for overlap
On the monthly view, overlap shows up as multiple campaigns scheduled in the same week targeting overlapping segments. When you see it:
Add exclusion filters to the segments involved before the seasonal campaign launches. Last Email Date on the always-on automated segment is the most common fix.
Stagger the sends. If two seasonal campaigns are scheduled the same week, move one to the following week. Seasonal demand windows are wider than a single week — staggering rarely costs revenue.
Decide which campaign owns the customer for that window. When two campaigns legitimately target the same customer at the same time, pick the higher-revenue campaign and exclude its audience from the other.
Look for gaps
Empty weeks on the calendar — particularly months with no seasonal campaign overlaying the always-on layer — are planning opportunities. When you see a gap:
Check whether the always-on automated campaigns are still doing the work. If Unsold Estimates and Memberships Expiring are generating enough revenue for that month, the gap may not need filling.
Plan a seasonal campaign for the gap when the month aligns with a trade-seasonal demand pattern. Build the campaign now and schedule it; don't wait until the month arrives.
Fill cross-tenant gaps with Enterprise Hub campaigns. If three of four tenants have a campaign running in November and one doesn't, an Enterprise Hub campaign can fill that tenant's gap without local effort.
Coordinate before launching
Each new campaign on the calendar adds inbound volume that lands in Contact Center, online bookings through Scheduling, and jobs Dispatch assigns. Before publishing a new seasonal campaign:
Tell CSRs the campaign is going live so they're staffed for the call volume.
Tell dispatchers so capacity is planned for the resulting bookings.
Tell technicians so they know which campaign is driving the work they'll see in the field.
Tip: Ask Atlas "What's my campaign schedule for next month?" to get a plain-language summary of what's scheduled across your tenants.
Comparing performance across campaigns
Marketing Analytics and Atlas are most useful when you ask them to compare campaigns against each other, not to review one in isolation. The framework below picks the right tab for the decision you're trying to make.
For the procedural steps — opening Marketing Analytics, navigating tabs, opening Atlas — see Build and manage multiple campaigns in ServiceTitan Max, Step 7.
Match the metric to the decision
Decision you're making | Tab to use | Why |
|---|---|---|
Which campaigns to keep, which to pause | Revenue | Attributed revenue is the bottom-line outcome of the entire ServiceTitan Max suite working together. |
Whether a cloned variant beat the original | The metric the test was designed to answer (see "Decide the winner" above) | A fair comparison uses the metric the test was built around, not the one that looks best. |
Capacity planning conversations with dispatchers | Booked Jobs | Booked-job counts predict the work Dispatch will need to assign. |
Whether bookings convert to sold work | Sold Jobs | Sold-job counts show which campaigns are bringing in customers who actually convert. |
CSR staffing for inbound volume | Calls | Call counts from tracking numbers route through Contact Center — these predict CSR workload. |
Ask Atlas the cross-campaign questions
Single-campaign questions ("How is my Unsold Estimates campaign performing?") are useful in isolation. The optimization questions are the ones that read across campaigns:
Which of my campaigns is generating the most revenue?
Which of my campaigns is generating the most booked jobs?
Are any of my audiences overlapping?
Which campaign should I pause first?
These are the questions that turn a list of campaigns into a ranked schedule.
Apply what Atlas recommends
The Your Campaigns, Supercharged by Atlas panel at the top of Campaign Manager surfaces recommendations on live campaigns. When a recommendation makes sense, confirm it in the Atlas chat and Atlas applies the change directly.
Tip: Atlas attributes results across the ServiceTitan Max suite — sold estimates from Field, online bookings from Scheduling, jobs assigned by Dispatch, and calls routed through Contact Center all roll up into the campaign performance you see in Analytics. That's what makes a multi-campaign view possible from a single question.
Deciding what to keep, pause, or refine
After enough time has passed for cloned variants and seasonal campaigns to complete at least one sequence, the comparison data in Marketing Analytics resolves four kinds of decisions. The framework:
Keep winners running. A cloned variant that outperformed the original on the metric the test was designed to answer should stay; the original should stop. A seasonal campaign that exceeded its expected revenue for the season is a candidate to repeat next year — clone it now while the configuration is fresh.
Refine campaigns that produced volume but not revenue. A campaign generating high call volume or booked jobs but low sold revenue usually has a segment problem, not a content problem. The audience is too broad. Tighten the segment's filters (see "Refining segments" above), then clone the campaign with the refined segment and let the original wind down.
Pause campaigns whose moment has passed. One-time seasonal campaigns finish on their own. Always-on automated campaigns aimed at a seasonal pattern — a Holiday Tune-Up automated campaign, for example — need to be stopped manually once the season ends. Stopped email campaigns can be resumed; stopped direct mail and SMS campaigns cannot — clone them instead to build a new but similar campaign next season.
Plan the next cycle on the Campaign Calendar. Once the current cycle's decisions are made, return to the Campaign Calendar in Marketing Pro Enterprise. The next quarter's seasonal layer is built on what the current quarter taught you.