How much does lifecycle marketing cost? The honest breakdown

Ask a vendor what lifecycle marketing costs and you get a subscription price. That price is usually the smallest of four numbers. The bill splits into software (the platform that stores your users and sends the messages), people (whoever designs, writes, builds and reviews the programs), production (templates, design, render testing), and integration (getting clean event data into the platform in the first place).

For most teams the order of size runs people first, integration second, software third, production fourth. Teams budget in exactly the reverse order, which is why lifecycle projects so often stall six weeks in with a paid platform, no events flowing into it, and nobody assigned to build anything.

Email platforms bill on one of three shapes, and the shape matters more than the sticker price because it decides how the cost behaves as you grow.

Contact-priced. You pay for the number of people stored in the database, in tiers. Klaviyo, Mailchimp and HubSpot's marketing tiers all work broadly this way. The cost steps rather than slopes: you sit comfortably inside a tier, then one import pushes you over the boundary and the bill jumps. The trap is that you are billed for contacts whether or not you ever email them, so an unpruned list is a standing monthly charge for people who will never open anything. That makes list hygiene a direct cost lever, not only a deliverability one.

Volume-priced.You pay per thousand messages sent. This is how sending infrastructure works — Amazon SES, Postmark, Mailgun, SendGrid's relay tiers. Per-send costs here are the lowest in the market by a wide margin, because you are buying delivery and nothing else. No campaign builder, no segmentation, no automation, no reporting worth the name. Teams who pick this path are buying an engine and building the car.

Contract-priced.Braze, Iterable and Salesforce Marketing Cloud sell annual contracts negotiated on monthly tracked users, data points, or channel entitlements, with no public price list. Cost is set by what you commit to, not what you use, and the commitment is annual. If you don't know your growth curve, you will either over-commit and pay for headroom you never reach, or under-commit and renegotiate from a weak position mid-year.

The practical rule: contact-priced platforms punish list size, volume-priced infrastructure punishes send frequency, and contract-priced platforms punish uncertainty. Match the shape to whichever of those three you have most control over. The ESP comparison guide covers fit; this is the money half of the same decision.

Whoever builds the programs is the largest line in almost every lifecycle budget. Three structures, each with a different failure mode.

In-house.A salaried CRM or lifecycle manager. You pay a full salary plus on-costs regardless of how much work exists in a given month, and the knowledge stays in the building. This works once there is genuinely a full-time job's worth of program work — several live programs, a testing roadmap, real reporting duties. Before that point you are paying full-time for part-time output, and the hire spends their first two quarters waiting on engineering for the events they need.

Agency.A monthly retainer, usually scoped in deliverables — this many emails, this many flows, this much reporting. Retainers buy capacity, which is exactly right when the constraint is production hours. They are a poor fit when the constraint is judgement, because a retainer priced in email volume gives the agency no reason to tell you that three of your five programs shouldn't exist. Agencies also rarely own your data model, so the integration work lands back on your engineers anyway.

Fractional.An experienced operator for a fixed number of days a month. You buy seniority without buying a full salary, and the person is incentivised to reduce your program count rather than inflate it. The trade-off is real: nobody is in the building daily, execution is slower, and if you don't write the decisions down, the knowledge leaves when the engagement ends.

Our recommendation, stated plainly: below a full-time workload, take fractional senior time over an agency retainer. Retainers are priced against production hours, and production hours are the part of this job that has compressed fastest — you end up paying a premium for the cheapest input in the stack. The cost of that choice is velocity, so buy it back with a documented data model and a template system the fractional operator hands over rather than carries in their head.

Production is the smallest of the four buckets and the easiest to control, because it is mostly a build-once decision. A modular template system — a header, a hero, a body block, a footer, all sharing one set of brand tokens — is built once and assembled per email afterwards. Design an email from scratch every time and you convert a capital cost into a recurring one, and you pay it again on every send for as long as the program runs.

Render testing sits in the same bucket and is genuinely worth paying for. Email clients disagree about almost everything, and the failures are silent: Gmail clips a message once the HTML passes 102KB, hiding your unsubscribe link below the fold of a "view entire message" link. Dark mode inverts colours your designer never tested. You either buy a testing tool or you discover these in production, and discovering them in production is more expensive than any subscription. The 102KB clipping guide covers that specific failure.

The integration bucket is where budgets actually break. Every platform quote silently assumes that clean, well-named events already arrive from your product — signup, activation, purchase, cancellation — attached to a stable user identity. For most companies that data doesn't exist yet in usable form, and creating it is engineering work: instrumenting events, resolving identities across web and app, deciding whether a warehouse and a sync layer sit between your product and your ESP. That work is invisible on the invoice and it is frequently the longest item on the timeline. The CRM versus CDP decision is really a question about how much of this tax you want to pay up front.

Before signing anything, price the stack as a system rather than a set of subscriptions. The Martech Stack Audit skill runs that assessment against how data actually moves between your tools, which is where redundant spend and missing pieces both show up.

Point figures would be dishonest here — pricing moves, currencies differ, and a number invented for an article gets quoted in somebody's budget deck six months later. At each stage, though, the structure of the bill holds steady. Find yourself in this table and you know which line to interrogate.

Two patterns worth naming. First, the software line only becomes the dominant cost at the top row, and by then it is negotiable — annual commitments, multi-year terms and channel bundling are all levers. Second, the surprise line moves down the stack as you grow: it starts as setup, becomes production, and ends as data engineering. Budget for the next row's surprise, not the current one's.

A genuinely free lifecycle stack exists, and it is more capable than most people assume. Authentication costs nothing but DNS records and an afternoon — SPF, DKIM and DMARC are configuration, not products, and skipping them is the single most common reason a small sender's email lands in spam (the setup is here). Google Postmaster Tools reports your Gmail reputation for free. Most contact-priced ESPs run a free tier capped on contacts and monthly sends, which is enough to run a welcome flow and a win-back before you pay anything. Orbit is free, and Claude with Orbit handles the strategy, the copy, the Liquid and the pre-send QA that a small team would otherwise buy in — the honest comparison of what that changes is worth reading before assuming it replaces a specialist.

You pay for the free stack in time and coverage instead. You will hit the free tier's contact ceiling exactly when the program starts working. You have no render testing, so client-specific breakage reaches your list. And nobody is watching your deliverability daily. Those are acceptable costs at small volumes and unacceptable ones past the point where a bad send damages a real revenue line.

AI changes the shape of the curve rather than flattening it. The work that has genuinely compressed is production: drafting variants, writing templating logic, building segmentation queries, running pre-send checks, producing the documentation nobody had time for. The work that has not compressed is judgement — deciding which programs deserve to exist, choosing what to measure, reading a deliverability signal correctly, and reviewing output before it reaches a customer. So the cost per email falls sharply while the cost of running the program properly falls much less. The practical consequence is that spending shifts away from production capacity and towards fewer, more senior people who review more and build less.

If you take one thing from this: price the whole system, not the subscription. The platform is the cheapest thing you will buy, the people are the most expensive, and the data plumbing is the item most likely to be missing from the plan entirely.