CRM for Ecommerce: What You Need and What You Don’t

Search for a CRM for ecommerce and you get a ranked list of ten tools. Half of them are sales pipelines built for a rep chasing a lead. A few are email platforms. One is a helpdesk. They are presented as competitors, scored on the same axes, as though a store owner could pick any of them and get the same job done.

They are three different products. Buying the wrong one is how a store ends up paying a monthly seat fee for a deal pipeline that nobody opens after week three, while the abandoned cart flow that would actually pay for it never gets built.

This piece sorts the category out. What each product type is really for, the signals that mean you genuinely need a CRM rather than a marketing platform, verified entry pricing with the billing unit named, and the pricing model that quietly punishes whichever way your business grows. TopTut sells none of these tools and takes nothing for naming them.

What a CRM for ecommerce actually does

A CRM for ecommerce keeps one record per customer and hangs everything else off it: orders, emails sent and opened, support conversations, notes, and whatever custom fields you add. That single record is the whole idea. Everything else the category sells is a consequence of having it.

The classic CRM was built around that record because someone was going to work it by hand. A rep opens the account, sees the last three conversations, notes the objection, moves the deal to the next stage. Pipeline stages, task reminders, deal values and forecasting all exist to help a human being manage a queue of relationships.

Retail ecommerce usually has no such human. Nobody calls the customer who bought a candle. The work is done by automation reacting to behaviour, and the tool that does that well is not a pipeline, it is a lifecycle marketing platform holding the same customer record with order data attached. Same idea, different machinery on top.

The three products people keep confusing

There are three distinct product categories sold under the CRM banner, and each one has a different job, a different failure signal and a different billing unit. Get the category right and the shortlist writes itself.

Category	Built for	Signal you picked wrong	Typical billing unit
Sales CRM	A person contacting a lead. Deal stages, tasks, quotes, forecasting, call logging.	Every record sits in one stage forever. Nobody has moved a deal in a month. Pipeline reports show one column.	Per seat, per month
Ecommerce marketing and lifecycle platform	Automation reacting to behaviour. Segments, flows, campaigns, browse and cart events, product data.	You are paying for contacts who have not opened anything in a year. You built nine flows and only one sends volume.	Per contact or active profile, sometimes plus send volume
Customer service helpdesk	Answering people. Shared inbox, tickets, macros, order lookup inside the reply window.	Support lives in a personal inbox and two people answer the same question. Or you are paying per ticket for a store with twelve tickets a week.	Per ticket, per resolution, or per agent seat

Plenty of tools straddle two categories. HubSpot sells a sales CRM and a marketing platform that share a contact database. Klaviyo has added service features. That is fine. What matters is which category you are actually buying into, because that determines what you pay for and what will sit unused.

The signals that mean you genuinely need a CRM

You need a real CRM when a human being is involved in the sale or in the relationship after it, and that human needs a queue. If no person ever contacts a customer directly to move money, a pipeline is decoration. Here are the signals worth acting on.

• High average order value with a person in the loop. Furniture, machinery, bespoke jewellery, anything where a buyer emails first and someone answers before the card is charged. That exchange is a deal, and deals need stages.
• B2B or wholesale sitting alongside retail. Two buyer types, two price lists, purchase orders and net terms on one side. The wholesale side is a sales function wearing a store’s clothing.
• Quotes and negotiated pricing. The moment a price is proposed rather than published, you have a document that needs versioning, an owner and a follow-up date. Marketing platforms have none of that.
• Long consideration cycles. Weeks between first contact and purchase, with several touches in between, means someone has to remember where each conversation stopped.
• Account management after the sale. Reorders, contract renewals, a named contact who expects to be called. That is an ongoing relationship with a person responsible for it.

If none of those apply, buy the marketing platform instead. A store selling forty pound consumables to consumers who never speak to anyone needs segments, flows and clean order data, and would get nothing from a deal pipeline except a monthly invoice. Pair that platform with solid organic acquisition, which is what our ecommerce SEO checklist is for, and you have covered the two things that actually move revenue at that size.

Seven tools, sorted by what they actually are

Below are seven commonly shortlisted tools with the category they belong to, the billing unit, and the entry price verified on each vendor’s own pricing page in August 2026. Note how few of them are competing for the same job.

Tool	What it actually is	Billing unit and entry price	Ecommerce integrations	Standout strength	Honest caveat
Klaviyo	Ecommerce marketing and lifecycle platform	Active profiles. Free up to 250 profiles and 500 emails a month; paid tiers priced by profile band on the pricing page.	Deep native Shopify, plus WooCommerce, BigCommerce, Magento	Product and browse level data in segments without engineering help	Cost climbs with the list whether or not the list engages
Omnisend	Ecommerce marketing and lifecycle platform	Contacts. Free covers 250 contacts and 500 emails a month; Standard from 16 USD a month for 500 contacts, Pro from 59 USD for 2,500 with unlimited sends	Shopify, WooCommerce, BigCommerce	Email, SMS and push in one flow builder at small-store prices	Reporting is shallower than Klaviyo once you want cohort views
Mailchimp	General email marketing, ecommerce bolted on	Contacts plus a send multiplier. Free under 250 contacts capped at 500 emails a month; Essentials 13 USD and Standard 20 USD a month at 500 contacts	Shopify and WooCommerce via connectors	Familiar, cheap at the bottom, everyone can use it	The send cap scales with your contact tier, so frequency is rationed
HubSpot	Sales CRM plus a separate marketing product	Per seat. Free CRM covers 1,000 contacts and 2 users; Sales Hub and Marketing Hub Starter list at 20 USD per seat per month	Shopify and WooCommerce connectors, strong API	The free CRM is genuinely usable for a wholesale pipeline	Marketing Hub also meters marketing contacts, so you pay on two axes at once
Zoho CRM	Sales CRM	Per user. Free for 3 users; Standard 14 USD, Professional 23 USD, Enterprise 40 USD per user per month billed annually	Via Zoho Commerce and third party connectors, thinner than the marketing tools	Cheapest credible pipeline for a wholesale side of the business	Not an ecommerce marketing engine; do not expect Klaviyo style flows
Drip	Ecommerce marketing and lifecycle platform	People on the list. From 39 USD a month for up to 2,500 contacts with unlimited sends, 14 day trial, no free plan	Shopify, WooCommerce, Magento	Unlimited sends at every tier, so frequency costs nothing extra	Smaller ecosystem, so a niche app you use may need custom work
Gorgias	Customer service helpdesk	Ticket volume, with AI resolutions billed only when a conversation is resolved. Helpdesk plans scale from 50 to 5,000 tickets a month	Shopify, WooCommerce, BigCommerce with order actions in the reply pane	Refund and reorder from inside the ticket, no tab switching	Not a marketing tool at all; it belongs next to one, not instead of one

Two things stand out. First, only two rows in that table are the same kind of product as each other in more than a loose sense. Second, the cheapest way into a genuine sales pipeline is free, because HubSpot’s free CRM covers 1,000 contacts and two users with no time limit, per its published pricing. If your wholesale side is small, that is your answer and it costs nothing.

The billing-unit trap that decides your bill in year two

Every pricing model punishes one specific kind of growth, and the trap is choosing the model that punishes the growth you are planning. Price per contact and your list becomes a liability. Price per send and frequency becomes rationed. Price per seat and hiring costs you twice. Price per order and success bills you directly.

Billing model	What it punishes	What it ignores	Fits a store that
Per contact or active profile	List growth, including contacts who never open anything	Send frequency, team size, revenue	Has a small engaged list and sends often
Per contact plus a send multiplier	Frequency. The cap rises only when your contact tier rises	Team size	Sends a few campaigns a month and no more
Per seat or per user	Team size. Every new hire is a recurring line item	List size, send volume, order count	Runs a big list with two or three people touching it
Per order, per ticket or per resolution	Success. Costs rise in step with volume	List size and team size	Wants costs to track revenue rather than lead it

Worked example: the same store under two billing models

Assumptions, stated plainly: one store, list grows from 500 to 2,500 contacts over twelve months, three people need access to the tool, and the team wants to send four campaigns a month on top of automated flows. All prices are the verified list prices above.

• Contact-based with a send multiplier. Mailchimp Standard is 20 USD a month at 500 contacts and allows twelve times your contact count in monthly sends, so 6,000 emails at that tier. Four campaigns to 500 people is 2,000 emails, leaving 4,000 for flows. Comfortable. At 2,500 contacts the cap becomes 30,000, four campaigns cost 10,000, and the monthly fee has moved up several tiers with the list.
• Contact-based with unlimited sends. Omnisend Standard is 16 USD a month at 500 contacts with a 6,000 email allowance; Pro at 2,500 contacts is 59 USD with no send cap. Same growth path, but the tenth campaign in a month costs nothing extra.
• Per seat, for contrast. Zoho CRM Enterprise at 40 USD per user per month billed annually costs 120 USD a month for those three people whether the list is 500 or 25,000. Growing the list is free. Adding a fourth person costs 40 USD a month forever.

The arithmetic makes the point better than any recommendation can. Under the first two models, a five times list increase drives your bill; under the third it does not move it at all. Under the third, hiring drives your bill; under the first two it is free. Pick the model that is blind to the thing you plan to grow.

List hygiene is therefore a direct cost lever, not housekeeping. Klaviyo’s billing documentation states that suppressed or unsubscribed profiles are excluded from the billable profile count once processed. Suppressing 6,000 dead contacts from a 24,000 record list cuts the billed base by a quarter, and on contact pricing that is a quarter off that line item every month, forever. Most stores never do it because deleting contacts feels like going backwards.

The data that makes any of this work

Four data streams determine whether your tool can do anything useful: full order history, product-level line item detail, browse and cart events, and a lifetime value figure derived from the first two. Without those, you own a newsletter tool with a customer list glued to the side.

Integration depth matters more than the feature list, and it is the thing comparison posts never check. Two platforms can both claim a Shopify integration while one syncs every line item with SKU and variant and the other syncs an order total and a date. The first can build a segment of people who bought a specific scent and are due a refill. The second can only tell you someone spent 42 USD in March.

Ask the vendor two questions in writing before you pay, and treat a vague answer as a no.

1. Does the integration sync line-item product data, including product ID, variant and SKU, or only order totals?
2. Does it backfill historical orders when I install it, and how far back, or does it only capture orders placed after connection?

The second question is the expensive one. A platform that starts from zero on install means your win-back and replenishment segments are empty for months, and lifetime value is wrong until enough history accumulates. On Shopify you can sanity check what a tool is receiving by looking at which webhook topics it subscribes to under Settings, then Notifications, then Webhooks.

Topics that matter for lifecycle work
  orders/create        new order, the backbone of everything
  orders/updated       refunds, edits, cancellations
  orders/fulfilled     triggers shipping and review requests
  checkouts/create     powers abandoned checkout
  customers/create     new profile

Inside the order payload, check for:
  line_items[]         product_id, variant_id, sku, quantity, price
If line_items is absent, you have order totals only.

On a self-hosted store the same logic applies through the store’s REST endpoints and webhooks rather than a hosted app catalogue, which is one of several reasons platform choice constrains tooling later. If you are still deciding where the store itself lives, our guides to Shopify alternatives and to building a WordPress ecommerce site both cover the integration consequences.

The flows worth building first, and the segments behind them

Build abandoned cart first, then post-purchase, then replenishment, then win-back, then welcome, and stop there until one of them is producing consistent revenue. Abandoned cart goes first because it catches people who have already chosen the product and stalled, which is the shortest distance between an automation and a completed order in most stores.

Build in this order

1. Abandoned cart or checkout. Highest intent audience you will ever message. Two or three emails, the first within an hour or two, product images pulled from the cart.
2. Post-purchase and review request. Sets delivery expectations, cuts where-is-my-order tickets, and asks for the review once the product has actually arrived rather than the day it shipped.
3. Replenishment, if you sell consumables. Trigger on a typical consumption window for that specific product, not a blanket thirty days. This one needs line-item data, which is why the integration question above matters.
4. Win-back. Aimed at people whose gap since last order has passed the point where they normally reorder. It doubles as your list hygiene mechanism, since non-responders can be suppressed.
5. Welcome. Useful, but last in the queue because it reaches people who have not bought yet and it competes with the discount that got them to subscribe.

That ordering reflects common practice across the ecommerce marketing world rather than a measured result from any particular store. Your own numbers will differ, sometimes sharply, and the only honest way to rank your flows is to look at revenue attributed to each one in your own reporting after ninety days.

Segmentation that is worth the effort

Four segments cover almost everything a small store needs. RFM in plain language means recency, how long since they last bought; frequency, how many times they have bought; and monetary, how much they have spent. Score each on a simple scale, and your best customers are the ones scoring high on all three.

• First-time versus repeat buyers. Different messages entirely. One needs reassurance, the other needs a reason to come back.
• Product-category affinity. Someone who only ever buys one category should not receive campaigns about the other three.
• Engaged in the last 90 days. Your default campaign audience, and the single best protection for deliverability.
• High RFM. Early access, restock notices, the occasional personal note.

Most stores over-segment and under-send. Twenty segments and one campaign a month is a worse business than four segments and eight campaigns. Segmentation is a way to send more without annoying people, not a substitute for sending.

Migration, lock-in and the sending domain problem

Contacts and order history export cleanly; almost nothing else does. A CSV of email addresses, names, custom fields and past orders will move between any two platforms in an afternoon. Plan for that part to be easy and for everything valuable around it to be rebuilt by hand.

• Flow logic does not export. Branching, timing, conditional splits and the templates attached to each step get rebuilt in the new builder. Screenshot every flow before you cancel anything.
• Engagement history usually does not follow. Opens, clicks and per-message history mostly stay behind, so your engaged-in-90-days segment starts empty and your suppression logic has nothing to work with on day one. Export a suppression list separately and import it first.
• Deliverability reputation does not transfer. It is attached to the sending infrastructure and the domain, not to your account.

Mistakes that cost real money

Five mistakes account for most of the wasted spend in this category, and four of them are decided in the first week of ownership.

• Buying a sales CRM for a pure retail store. You pay per seat for pipeline features nobody uses, and you still need a marketing platform, so you now run two tools and pay twice.
• Paying for unengaged contacts. Under contact pricing, a contact who has not opened anything in eighteen months costs the same as your best customer, and drags your engagement rates down while doing it.
• Importing a purchased or scraped list. Spam complaints and hard bounces from a list that never opted in damage your sending reputation, which then suppresses delivery to the people who did opt in. The damage lands on your good contacts.
• Building twelve flows before one works. Twelve half-finished automations is twelve things to debug and no way to tell which one is earning. Ship one, watch it for a month, then build the next.
• Treating the CRM as the source of truth for inventory. It is not, and it will happily send a back-in-stock email for something you sold out of an hour ago. Your store platform owns stock; the CRM reads it.

There is a sixth, quieter one: buying a platform for a feature you saw in a demo and will not configure for six months. The tool you will actually set up this week beats the better tool you will get around to. That principle runs through most of the tooling advice in our roundup of content marketing tools as well.

The verdict by store profile

Pick by profile, not by feature count. Here is what I would do in each case.

• Retail store under about 1,000 contacts, no B2B. Omnisend or Mailchimp at the entry tier, or a free plan while you build the first two flows. No CRM. Revisit when the list passes a few thousand.
• Growing retail store, thousands of contacts, product data matters. Klaviyo if you want the deepest segmentation and can live with profile-based pricing, Drip if you intend to send frequently and want unlimited sends at a fixed contact tier. Still no CRM.
• Retail plus a wholesale or B2B side. Two tools. Marketing platform for retail, HubSpot’s free CRM for the wholesale pipeline until it outgrows two users, then Zoho CRM Standard or Professional per seat, which is the cheaper way to add people.
• High-value considered purchases, quotes, long cycles. A real CRM is the primary system here, with email second. Zoho for cost control, HubSpot if you want the marketing side on the same contact record and accept paying on two axes.
• Drowning in support tickets. That is a helpdesk problem, not a CRM problem. Gorgias or a comparable helpdesk, priced on ticket volume, sitting alongside whatever marketing platform you already run.

The one I would avoid for a normal retail store is any per-seat sales CRM bought on the promise that you will grow into the pipeline features. You will not, and the invoice arrives every month regardless. Spend the money on the flows and the list instead, and keep the rest of your acquisition work honest with the fundamentals covered in our guide to internet marketing strategies.

Pricing here was checked in August 2026 against each vendor’s own pages: Omnisend, Mailchimp, Zoho CRM and Gorgias. Vendors change tiers often, so confirm before you commit.

Frequently asked questions