Marketing Automation Tools: Pick by Category, Not Rank

Four pricing models — and the migration cost nobody budgets for.

The commonest expensive mistake here is buying a platform three tiers above the problem.

A team sending one newsletter a month does not need a journey builder. It needs a clean list and something worth writing.

If you are still deciding whether to automate, our explainer on what marketing automation is and where it fits in a funnel covers that.

This is not a ranking. Rankings sell software; categories tell you what to shortlist.

The four categories of marketing automation tools, side by side

Read the last column first. The prerequisite decides whether the purchase works, and it is the column every vendor demo skips.

Email and sequence tools: per contact, or per email sent

This category sends things and remembers who opened what. It is where most Indian SMEs should start, and where many should stay.

The pricing model splits two ways, and that split matters more than any feature list.

• Per contact stored. Mailchimp prices in contact-count tiers with a monthly send allowance, and is free below a small contact threshold (read 7 October 2026). Zoho Campaigns also prices by contact tier and will bill an Indian entity in rupees — its pricing page carries an INR option (read 7 October 2026).
• Per email sent. Brevo works on sending credits — one email sent deducts one credit — and does not charge for storing contacts, with a free tier capped by daily sends (read 7 October 2026).

Per-contact billing punishes a dormant list. Per-send billing punishes high frequency. Pick the one that is wrong in the direction you are not growing.

One billing detail matters before you import. Klaviyo's billing documentation — the page shows "Updated Mar 25, 2026" — counts as an active profile any profile that can be emailed, regardless of consent status. On a profile-counted model an uncleaned list is a line item, not just a liability.

This category will not design your sequences. That work lives in email marketing automation and in how you structure a drip campaign.

CRM plus automation: priced per seat

If the gap is "leads arrive and nobody follows up", you need a CRM with workflow rules, not a marketing platform.

Pricing is almost always per user, per month, so cost follows headcount rather than list size. Zoho CRM is per user per month, bills in rupees as well as dollars, includes workflow automation on its free tier, and gates full journey orchestration to Professional and above (read 7 October 2026).

That gating pattern is what to look for: basic rules are cheap, journeys are the upsell.

Choosing the CRM itself is a separate exercise — our guide to a CRM for a small business covers it.

Full platforms: seats, contacts and an onboarding fee

A full platform earns its cost when several people run campaigns across several channels against a funnel everyone agrees on. Below that you are paying for governance you do not need.

The pricing model is hybrid, and that is where budgets break. HubSpot's Marketing Hub pricing page bills on both axes — core seats and marketing contacts — and states that its Professional and Enterprise tiers carry a required one-time onboarding fee (read 7 October 2026). ActiveCampaign combines contact bands with included seats and sells extra seats as an add-on (read 7 October 2026).

Two companies on the same plan name can therefore pay very differently: one with a small team and a huge list, one with the reverse.

At the app-first end, pricing stops being public. MoEngage describes itself as a customer engagement platform spanning push, email, SMS, WhatsApp and in-app messaging, and publishes no price (read 7 October 2026). Quote-based pricing signals the buying process, not a red flag — expect scoping and procurement.

The connector tier: per task, per operation, per execution

For many small teams this tier replaces a platform: you keep the email tool and the CRM you already pay for and wire them together.

The billing unit is the whole decision: it decides whether complexity or frequency costs you more.

• Zapier bills tasks, and every step in a Zap consumes them, with a small free monthly allowance (read 7 October 2026). A twelve-step workflow costs roughly twelve times a one-step one.
• Make bills credits per module action, with a free tier limited by credits, active scenarios and scheduling interval (read 7 October 2026).
• n8n bills monthly workflow executions regardless of step count, includes unlimited users, and publishes a self-hostable community edition (read 7 October 2026). Flatter for complex flows, harder on high-frequency triggers.

Estimate runs per month and steps per run before you buy — that sum tells you which billing unit is on your side.

What has to exist before automation is worth buying

Automation multiplies a process. It does not create one. A team with no segmentation, no content and no owner ends up with a more expensive version of what it already had.

Five things have to be real before any of these tools pays back:

1. A named owner. One person whose job includes the platform. Shared ownership means nobody rebuilds the broken flow.
2. Segments that exist. If you cannot name three groups in your list and what each should hear, start with customer segmentation.
3. Content for every branch. Four paths need four sets of copy. Most stalled implementations are content shortages in a software costume.
4. A trigger your data captures. "When they view the pricing page" works only if that event reaches the tool. Check before the demo.
5. One agreed success metric. Pick it from your marketing KPIs and write it down.

Automation does not manufacture demand either. If the top of the funnel is thin, a platform nurtures a trickle faster; fixing the trickle is organic and paid work, and organic typically takes three to six months to show movement.

The migration cost nobody budgets

The subscription is the cheap part. Switching costs you weeks of someone's attention, and that rarely appears in the business case.

• Data cleanup. Duplicates, dead addresses, fields that meant different things in different years. On a per-contact model you pay for every row you fail to delete — a first-party data problem before it is a software one.
• Re-permissioning the list. Consent does not always travel cleanly between systems. Treat anything you cannot evidence as a re-opt-in job.
• Rebuilding every sequence. Flows do not export between vendors usably, and the rebuild surfaces every undocumented rule the old system was quietly running.
• Deliverability warm-up. A new sending setup has no reputation. Expect a ramp, and read up on email deliverability first.
• The weeks both systems run. Two subscriptions, and someone watching for double sends.

Budget the overlap and the rebuild as part of the purchase. A migration costed honestly is survivable; one assumed free is where teams abandon a half-configured platform.

Buying in India: dollar invoices and messaging consent

Two things change the maths for an Indian entity, neither of them a feature.

The invoice currency. Most of these vendors price in US dollars, so your rupee cost moves with the exchange rate at every renewal, and imported software is taxed differently from a domestic invoice. Finance should sign that off, not marketing. Where a vendor can bill in rupees — Zoho's pricing pages carry an INR option, read 7 October 2026 — the comparison gets simpler.

Messaging channels carry their own rules. Commercial SMS to Indian numbers runs through the telecom industry's registration regime: sender identities and message templates are registered and approved before they will deliver, and subscriber preferences are enforced at the network, not inside your tool. A platform that shows you an SMS node has not necessarily solved that registration — ask who does, and verify current requirements with your operator or registration partner. Our note on SMS marketing in India covers the channel. None of this is legal advice.

WhatsApp needs separate pricing attention. Meta's WhatsApp Business Platform pricing documentation bills per message, with rates varying by template category and recipient country — that page shows no last-updated date, so treat this as read on 7 October 2026. A subscription that "includes WhatsApp" still sits on top of Meta's per-message charges. Digital Hangover does not run WhatsApp broadcast marketing.

If paid campaigns generate leads nobody works, the fix is upstream — that is what our performance marketing team handles, alongside the lead nurturing the tool will run.

Five questions to answer before a demo

Answer all five in writing first. A vendor call where you know these takes thirty minutes and produces a decision.

1. What is the one workflow that has to work in month one? Name it specifically. If you cannot, you are not ready to buy.
2. Which billing unit are we exposed to? Contacts, sends, seats, tasks or executions — and which is growing fastest.
3. Who owns this on Monday morning? A name, with hours protected.
4. What exists already — segments, content, captured events? List them, then make the vendor demo against your list, not their sample data.
5. What does leaving look like? Export format, contract term, who rebuilds the flows. Ask before you sign.

Where to go from here

Work from the problem to the category, then to two tools, then a demo. That order keeps you out of the tier above your problem.

Under a few thousand contacts, shortlist an email tool and a connector and stop. A full platform is defensible once several people run campaigns across several channels — and the onboarding fee, the migration weeks and the content backlog are part of its price.