# Email Marketing for SaaS Companies

> A start to finish email setup for SaaS teams: data plumbing, which sequences to build first, tool choice by stage, reporting and a 90 day rollout plan.

Source: https://saas-marketing.net/guides/email-marketing-for-saas-companies/
Topic: SaaS Email Marketing
Type: guide
Published: 2026-09-11
Last updated: 2026-09-11
Publisher: SaaS Marketing (saas-marketing.net)
License: CC BY 4.0. Quote or republish with attribution and a link to https://saas-marketing.net/guides/email-marketing-for-saas-companies/

## Short answer

Build a SaaS email program in this order: data plumbing first (events, traits, identity resolution, authenticated sending domains), then the five sequences ranked by revenue per hour of work, which are trial expiry, onboarding, dunning, activation nudges and expansion. Choose the tool by list size and event volume rather than feature lists. Expect a working program in 90 days with a team of one, provided you skip the newsletter until month three.

## Key takeaways

- Dunning beats the newsletter. A failed payment recovery email is the highest revenue per send email a SaaS will ever write.
- Wire events, traits and identity before choosing a tool, or you will pick the wrong one and migrate within a year.
- Separate marketing and transactional sending domains. One bad campaign should never take password resets with it.
- At 5,000 profiles expect $0 to $100 a month; at 100,000 the same feature set runs $1,500 to $6,000.
- Build five sequences in 90 days, measure revenue per send, and report that single number to keep the program funded.
- Most SaaS email programs fail on data quality, not copy. Budget engineering time, not agency time.

---

If you are the first marketer at a SaaS company and someone has asked you to "do email", the instinct is to write a newsletter. Resist it for eight weeks. The newsletter is the lowest revenue per hour thing on the list, and building it first tends to lock you into a tool that cannot send on product events.

This is the build order I would follow, written for a team of one with a few days of engineering time to spend.

## Week one: the data plumbing nobody wants to do

Before you evaluate a single tool, get four things in place. Every problem later in the program traces back to one of them.

**Events.** A named list of product actions worth emailing about. Keep it to twelve for now. Typical set: signed_up, activated, project_created, invited_teammate, hit_usage_limit, subscription_started, payment_failed, subscription_cancelled, feature_x_used, exported_data, integration_connected, logged_in_after_dormancy. Name them in past tense with snake case and never rename one after launch.

**Traits.** Attributes that describe the person and the account: plan, seat count, ACV, signup source, role, account created date, last active date. Keep person and account separate from day one, because retrofitting an account object into a tool that only has contacts is the most painful migration in this category.

**Identity.** One stable user ID from your database, sent with every event, with email as a secondary key. Email as a primary key breaks the moment someone changes theirs, and someone always does.

**Sending domains.** Marketing on `mail.yourdomain.com`, transactional on `notifications.yourdomain.com`, each with SPF, DKIM and a DMARC policy. Google and Yahoo now enforce authentication and one-click unsubscribe for bulk senders, so this is not optional hygiene, it is a delivery requirement.

Sharing one domain between campaigns and password resets means a badly targeted announcement can put receipts in spam. Separate the subdomains before your list crosses 5,000 people. Doing it afterwards means warming a new domain while your reputation is already established elsewhere.

## Which five sequences should you build, and in what order?

Rank by revenue per hour of build time, not by funnel position. That ordering produces a counterintuitive list, and it is correct.

### Why dunning comes before the newsletter

A dunning email goes to someone whose card just failed, who already decided to pay you, and who mostly needs to update an expired card. There is no persuasion involved. Recovery rates of 30 to 50 percent of involuntary churn are commonly reported, and involuntary churn is often a third of total churn at self serve companies.

It takes about a day to build: four sends over 14 days, escalating tone, each linking directly to a card update page that does not require a login. Stripe and most billing providers expose the events you need with no custom instrumentation at all.

Compare that to the newsletter. Weeks of ongoing effort, indefinite payback, and a real risk of list fatigue if the cadence outruns your ability to say something useful. I would not write a single issue before the five sequences above are live.

**30-50%** Share of involuntary churn typically recoverable by a well built dunning sequence

### The trial expiry sequence

Highest intent audience in your entire database, and a deadline you control. Five sends: day 3 value check, day 7 mid-trial nudge based on whether they activated, day 11 case or feature relevant to their usage, day 13 expiry warning with the specific things they will lose, day 15 post-expiry recovery with a short extension offer for active users only.

Branch on activation. A user who never created a project needs a different day 11 email than a user who created nine, and sending both the same message wastes the strongest segmentation signal you have.

## Which tool at which stage, and what it actually costs

Choose on event volume and data model, not on feature checklists. Every tool in this category can send an email.

| Stage | Profiles | Sensible options | Typical monthly cost | Watch out for |
| --- | --- | --- | --- | --- |
| Pre-seed to seed | Up to 5,000 | Loops, Userlist | $0-100 | Limited branching depth |
| Seed to Series A | 5,000-25,000 | Loops, Encharge, Customer.io | $200-900 | Event volume overages |
| Series A to B | 25,000-100,000 | Customer.io, HubSpot | $900-3,000 | HubSpot contact tier jumps |
| Series B and beyond | 100,000+ | Customer.io, Braze | $1,500-6,000+ | Braze annual contracts and implementation cost |

A few opinions on the specific tools. Loops is the fastest to get running and the easiest to hand to a non technical founder, with a genuinely simple event API. Userlist was built for SaaS specifically and has the cleanest account-versus-user model at the small end, which matters more than people expect. Encharge sits in between with stronger branching than Loops at a similar price.

Customer.io is where most product-led companies land by Series A, and it is the one I would pick if I expected to still be using the tool in three years. Its pricing scales on profiles and events, so model both before signing. HubSpot makes sense when the same team runs the CRM and marketing email and nobody wants two systems, but its product event handling is weak and the contact tier pricing gets expensive fast. Braze is a genuinely different class of tool with an implementation cost to match, and below roughly $30M ARR it is usually overbuying.

Self hosting is a real option if you have engineering capacity and a data residency constraint, covered in [Mautic for SaaS](/guides/mautic-for-saas/). For a wider field including tools not listed here, see [the best email marketing software for SaaS](/guides/best-email-marketing-software-for-saas/).

Every vendor demo shows the campaign builder. Spend your trial sending 500 test events through the API with realistic payloads instead, and check how the tool handles a nested object, a late arriving event and a user identified after the fact. That is where tools in this category actually differ.

## What reporting keeps the program funded?

One table, monthly, with revenue in it. Open rates have not been trustworthy since Apple Mail Privacy Protection started prefetching images, and reporting them to a CFO invites a question you cannot answer.

The view that works:

| Sequence | Sends | Clicks | Holdout gap | Revenue effect | Per send |
| --- | --- | --- | --- | --- | --- |
| Dunning | 310 | 41% | +38 pts recovery | $41,000 recovered | $132 |
| Trial expiry | 2,140 | 9.2% | +2.1 pts conversion | $63,000 new ARR | $29 |
| Expansion | 890 | 7.4% | +1.4 pts upgrade | $22,000 expansion | $25 |
| Onboarding | 6,300 | 11% | +6 pts activation | Indirect | n/a |
| Newsletter | 14,200 | 2.8% | Not measurable | Indirect | n/a |

The holdout column is what makes this credible. Hold back 5 to 10 percent of eligible recipients from each triggered sequence and compare. Without it, every number in the revenue column is a claim that a user who was going to convert anyway converted because of your email, and a skeptical finance team will say so.

Revenue per send is the single number to lead with. It ranks sequences honestly, it explains why dunning gets built first, and it survives translation to people who do not care about email. Model your own figures with the [SaaS email revenue calculator](/calculators/email-revenue/) before you commit to a target, and check your rates against the [SaaS email benchmarks](/research/saas-email-benchmarks/) rather than against generic cross-industry averages.

## The 90 day rollout

**Ninety days from nothing to a working program**

What to skip in month one: the newsletter, any preference centre beyond a single unsubscribe, segmentation beyond plan and activation status, and A/B testing. Testing subject lines on 400 sends produces noise, and the time is better spent shipping the next sequence.

## Where these programs actually fail

Not on copy. On data.

The most common failure is events that fire inconsistently. A `project_created` event that also fires on template duplication will send congratulations emails to people who did nothing, and by the time you notice, trust in the whole program is damaged internally. Audit event definitions with the engineer who wrote them, monthly, for the first six months.

Second failure: nobody owns it after launch. Lifecycle email is not a project, it is a surface that decays. Sequences reference features that got renamed, links rot, and a branch that was correct in March is wrong in September. Budget two hours a week forever, and put a review date on every sequence. Common errors worth pre-empting are collected in the [SaaS email marketing mistakes](/guides/saas-email-marketing-mistakes/) write-up.

Third, and most expensive: choosing a tool before the data model exists. Teams pick based on a demo, discover six months later that the tool has no account object or cannot branch on a numeric event property, and migrate. That migration costs roughly six weeks and everyone's goodwill.

## What to do next

Write the twelve event names down today and get them in front of an engineer this week. That single step unblocks everything else, and it is the one most teams postpone because it does not feel like marketing.

Once events exist, build dunning first, then trial expiry, then the rest in order. If your motion is sales-led rather than product-led the emphasis shifts, and [B2B SaaS email marketing](/guides/b2b-saas-email-marketing/) covers that variant. For the mechanics of firing on product actions, the [behavioral email trigger](/glossary/behavioral-email-trigger/) definition is the short version, and [drip campaigns for SaaS](/guides/drip-campaigns-for-saas/) explains the narrow cases where a fixed schedule still beats an event. Once the five sequences run, expansion is where the next meaningful revenue sits, covered in [expansion revenue email campaigns](/playbooks/expansion-revenue-email-campaigns/) and the wider [SaaS email marketing](/saas-email-marketing/) hub.

## Frequently asked questions

### What email sequences should a SaaS company build first?

Trial expiry, then onboarding, then dunning, then activation nudges, then expansion. Rank by revenue per hour of build time. Trial expiry touches users with the highest intent and the shortest decision window, and dunning recovers revenue you have already earned, which makes both far more valuable per hour than a newsletter.

### How much does SaaS email marketing software cost?

At 5,000 profiles, $0 to $100 a month across Loops, Userlist and Encharge. At 25,000, roughly $200 to $900 depending on event volume. At 100,000, roughly $1,500 to $6,000, with Braze and enterprise HubSpot tiers sitting well above that. Event volume drives cost more than contact count in product-led tools.

### Do you need a customer data platform to run SaaS lifecycle email?

No, not under about $10M ARR. A well designed event schema sent directly from your application to your email tool covers most needs. A CDP earns its cost when three or more systems need the same events and you are maintaining parallel integrations for each.

### Should marketing and product emails send from the same domain?

Use the same root domain for brand recognition but separate subdomains for marketing and transactional sending, each with its own authentication records. That way a poorly received campaign cannot damage delivery of password resets and receipts, which must arrive.

### How long does it take to build a SaaS email program?

Ninety days for a working program covering five sequences, assuming one marketer and a few days of engineering help for event instrumentation. The instrumentation is the long pole. Teams that try to build sequences before events exist spend the same 90 days and end up with date based emails.

### What should you report to keep an email program funded?

Revenue per send and revenue influenced by sequence, with a holdout group. Open rates stopped being trustworthy after Apple Mail Privacy Protection, and click rates alone do not survive a CFO question. One table showing recovered and accelerated revenue per sequence is what keeps the budget.