Scaling e-commerce · 2 guides

Scaling an ecommerce business: affiliates, new countries and cash on delivery

Affiliates, multiple countries and cash on delivery to grow past your first channel.

Warehouse with boxes ready to ship and a worker scanning a parcel

Scaling e-commerce guides

Start here

Scaling an ecommerce business means growing revenue without the cost of each sale growing with it. In practice, your store first needs healthy margins, a stable conversion rate and tracking you can trust. Then it grows through levers: more spend on the channel that works, higher order value, repeat purchases, affiliates, cash on delivery and new countries.

This is the growth hub of the Horsee blog. It covers how to tell whether your store is ready, what each lever requires and in what order to use them. To see the growth tools built into the platform (A/B testing, affiliates, coupons and email with attributed revenue), visit the Horsee growth page.

What does scaling an ecommerce business actually mean?

Growing and scaling are not the same thing. Growing is selling more. Scaling is selling more while keeping, or improving, the profit on each sale. A store that doubles revenue by doubling its acquisition cost has grown, but it hasn't scaled: the owner works twice as hard for the same profit, or less.

That is why the first job of anyone who wants to scale is not finding a new channel. It is understanding the economics of a single order: what it costs to acquire a customer, how much margin they leave on the first purchase and how much on the ones after. With that math in hand, every lever in this guide turns into a simple question. Does it raise margin per order, lower acquisition cost or increase purchases per customer?

How do you calculate the economics of one order?

Unit economics fits in a spreadsheet with a handful of rows. Start from the price the customer pays and subtract, in order:

  1. Cost of goods, including packaging.
  2. Shipping you pay, either all of it or the part you don't pass on.
  3. Payment fees, which vary by method and card type.
  4. Sales tax or VAT you owe on the sale, where it applies.
  5. Discounts: coupons, free shipping, method discounts.
  6. Acquisition cost: ad spend divided by the number of orders it drove.

What's left is the contribution margin of the order. If it is negative, you lose money on every sale, and scaling just digs the hole faster. If it is positive but thin, any rise in acquisition cost wipes it out. The math also shows where to push: sometimes the biggest win is a higher order value or a better payment mix, not a cheaper ad.

Run the numbers with your store's real data, and rerun them every month. Payment fees, shipping rates and ad costs move, and a margin that worked last quarter may not work now.

How do you know your store is ready to scale?

Pouring more budget into a store that doesn't make money per order only speeds up the losses. Before scaling, check five signals:

  1. Positive contribution margin per order after product, shipping, payment fees, taxes and acquisition cost.
  2. Tracking that matches your bank account. The sales your ad platform reports are close to the sales in your store. If they aren't, you will scale the wrong campaign.
  3. Stable conversion. Your site's conversion rate doesn't hinge on a single creative or a lucky day.
  4. Operations that can handle double. Inventory, suppliers, fulfillment and support can absorb more orders without late deliveries.
  5. A checkout that doesn't leak. Cards, wallets and cart recovery working, because every problem there multiplies with traffic.

If one of these fails, start there. The tracking and attribution hub fixes the second, conversion and A/B testing the third, and checkout and payments the last.

Which ROAS should you use to decide where to scale?

ROAS (revenue divided by ad spend) is the number that drives most budget decisions, and the easiest one to misread. The ROAS Meta shows is calculated by Meta, with Meta's attribution. The ROAS that pays the bills is based on revenue that actually landed in your store.

On the Horsee Ads screen, the official ROAS is the real one, calculated from your store's orders, and Meta's ROAS always appears as a secondary number for comparison. Scaling on the real number keeps you from pouring money into a campaign that looks great in Ads Manager and never shows up in your bank account. The Meta Conversions API guide explains how to bring the two numbers closer by sending purchases from your server too.

What are the levers for scaling an ecommerce store?

LeverWhat it growsPrerequisiteRisk
More spend on the channel that worksOrder volumeReliable trackingCost per order rises with spend
Site conversionOrders from the same trafficEnough volume to testCalling test results too early
Average order value (order bumps, upsells, bundles)Revenue per orderA relevant add-on productOffers that distract from the main purchase
Repeat purchases (email, WhatsApp, loyalty)Orders per customerA product people rebuyBurning out your list
AffiliatesChannels with no fixed costMargin to pay commissionLow-quality traffic and fraud
Cash on delivery (COD)Conversion in markets that prefer paying at the doorLogistics that collect on deliveryRefused deliveries
New countriesTotal addressable marketGateway, language and shipping for the destinationOperating and support costs

The order in the table is deliberate. The first levers use what you already have: traffic, customers, your store. The last ones open new fronts, with higher cost and complexity.

Scaling spend: how do you raise the budget without breaking the campaign?

The most direct way to scale is putting more money behind what already works. It also has a ceiling: past a certain point, each extra dollar buys more expensive customers, because the ad reaches people who are less interested. Three habits help:

  • Raise budgets gradually and check real cost per order at each step.
  • Feed new creative often. What usually stalls scale is creative fatigue, not the audience. A queue of variations keeps you from depending on one ad.
  • Set a hard limit. Decide in advance the maximum cost per order your margin can take, and stop raising spend when you hit it.

How does site conversion help you scale?

Every point of conversion makes every sale cheaper, across every channel, at once. If the store converts better, the same ad brings more orders and cost per order drops. It is the lever with the broadest effect and often the cheapest one.

The safe way to improve it is to test. On Horsee, A/B testing runs two versions of the same page on the same URL and picks the winner by conversion measured on paid orders, with no external tool. Ecommerce conversion rate explains how to measure it and which changes to test first.

How do you increase average order value?

A higher average order value means more revenue for every customer you acquire, at the same ad cost. Three tactics work for most stores:

  • Order bump: a cheap, relevant add-on inside checkout, added with one click.
  • Post-purchase upsell: a bigger offer on its own page after payment, without asking for the card again.
  • Bundles and quantity breaks: buy two or three with a tiered discount, which also spreads the shipping cost.

Keep order bumps simple. One sentence, one checkbox, no extra step before payment. The Help Center explains where to set up order bumps and upsells on Horsee.

How do you get customers to buy again?

A customer who already bought doesn't cost an ad to come back. If your product gets rebought (skincare, supplements, food, refills), repeat purchases are one of the highest-return levers, and many stores neglect them because they are busy with acquisition.

The basics are organized contacts and the right message at the right time. On Horsee, customers from orders and leads live in Contacts, Segments split the list by rules (who bought a given product, who hasn't ordered in a while), email Campaigns show the revenue each send brought in, and Automations send messages after purchase or delivery. Loyalty adds points and rewards for customers who come back.

Before repeat purchases, there is an even easier sale: the one that almost happened. Abandoned cart recovery and the cart recovery and CRM hub cover it.

How do you use affiliates to scale?

In an affiliate program, other people promote your product and earn a commission on the sales they drive. The upside is variable cost: you only pay when you sell. The downside is control: the traffic comes from sources you don't manage, and the commission comes out of your margin.

To work, a program needs:

  1. Margin for the commission. The commission replaces part of your acquisition cost, so it has to fit the same math.
  2. Reliable attribution. Each sale must be credited to the right affiliate, by link or by coupon.
  3. Ready-made assets. Pages and creatives affiliates can use without reinventing the offer.
  4. Clear rules. What is and isn't allowed in promotion, when commissions are approved and when they are paid.

On Horsee, the affiliate program is available on nutra stores. The producer creates a network, sets the commission (percentage or fixed, with tiers by order value and a per-order cap), invites affiliates, shares funnels for them to promote and tracks sales and commissions in reports. A network can allow sub-affiliates, who earn on the sales they refer. For any store type, coupons attributed all the way to the order make a simpler partner or creator program possible.

Whatever tool you use, watch for the signs of a program going wrong: a sudden spike of orders from one affiliate with unusually high refund or chargeback rates, coupon codes showing up on deal sites you never approved, or affiliates bidding on your brand name in search ads. Write the rules down before you invite anyone, hold commissions until the return window closes, and remove partners who break the rules quickly. A small group of good affiliates beats a large group you can't monitor.

Cash on delivery: when does COD help you scale?

With cash on delivery, the customer pays when the order arrives, in cash or by card. In US ecommerce it is rare, but in several markets outside the US it is what many shoppers expect, and offering it can win sales that would never happen by card.

The cost is in refusals. A buyer who orders and doesn't pay at the door means shipping both ways and stock sitting idle. COD operations that work have two pieces:

  • Confirmation before shipping. Reach the customer (call or message) to confirm the order and address before dispatch.
  • Logistics built for COD. A fulfillment partner that delivers, collects and reports the status of each order back to you.

On Horsee, Cash on Delivery is a payment method in Settings → Payment Methods. It doesn't go through a gateway: you confirm the payment on the order. For logistics, the N1 Global integration (warehousing and COD fulfillment) sends each order to the store's N1 operation automatically and gets status and tracking back. You connect it in Apps, with the API credentials and the webhook URL the setup screen shows. Give every product a SKU, because N1 matches order items by SKU.

Before you open COD in a new market, ask three questions. Can your margin absorb return shipping on refused orders? Is someone available to confirm orders during the hours they come in? Does the offer state the full amount due at delivery, with no surprises? If any answer is no, start with a small ad budget and measure the refusal rate before scaling spend.

How do you sell to other countries?

Selling abroad multiplies your market, but it is more than translating the store. A shopper in another country expects prices in their currency, payment methods they know, a reasonable delivery time and someone to talk to in their language. Before opening a country, answer:

  1. Currency and gateway. Which gateway charges in that currency, and which local methods matter there?
  2. Language. Are the store, emails and support in the customer's language?
  3. Shipping and delivery time. What does it cost and how long does it take? International shipping changes the margin math.
  4. Duties and rules. What import taxes apply, and which consumer rules hold at the destination?
  5. Returns. Returning a product across borders is expensive. Your policy has to say how it works.

On Horsee, the approach today is one store per market. The AI assistant creates a full translated copy of your store, with products, pages and menus. In the copy, you set the currency in settings, connect a gateway that charges in it (Stripe, PayPal, Ziina or Whop, each with its own currency list, or the Brazilian gateways for BRL), review shipping and policies and, if you like, connect a dedicated domain. The two stores stay independent, which also lets you set different prices and offers per country. If Brazil is on your list, start with how to accept Pix payments.

When should you open a second store or brand?

Sometimes scaling means opening a new front: a new niche, a new brand, a new country. A second store makes sense when the audience or the offer is different enough that mixing them would hurt, for example a supplement brand and an accessories brand, or an English store and a Portuguese one.

On Horsee, multiple stores live under one account, each with its own products, domain, payments and pixels. The number of stores depends on your plan; see pricing.

In what order should you pull the levers?

  1. Make one order profitable. Margin, acquisition cost and tracking that matches your bank account.
  2. Fix checkout. Payment methods, declines, recovery. See on-site checkout vs hosted checkout.
  3. Test conversion. Product page, offer, social proof.
  4. Raise order value. Order bumps, upsells and bundles.
  5. Scale spend on the channel that works. Gradually, with a cost-per-order ceiling.
  6. Turn on repeat purchases. Segments, campaigns and post-purchase automations.
  7. Open new channels. Affiliates and partners.
  8. Open new markets. COD where buyers expect it, new countries when operations can take it.

Skipping steps is the most common mistake. Opening a new country with a leaky checkout just moves the leak to another currency.

Which numbers should you watch every week?

  • Real cost per order by channel and campaign, against the ceiling your margin allows.
  • Real ROAS, based on store revenue, next to the ad platform's ROAS.
  • Conversion rate for the site and for checkout, split by device.
  • Average order value and the take rate of order bumps and upsells.
  • Repeat purchase rate: how many customers buy again within 30, 60 and 90 days.
  • Recovered orders from cart and declined-payment flows.
  • COD refusal rate, if you offer cash on delivery.
  • Delivery times and complaints, the first sign operations are at their limit.

When one of these gets worse while spend goes up, stop raising spend and fix the cause. Scaling is a series of small steps, not a leap.

Mistakes that stall scale

  • Scaling on Ads Manager ROAS. The campaign looks great and the bank account doesn't follow.
  • Ignoring payment costs. Gateway fees, currency conversion and chargebacks eat margin quietly. Always check your provider's official pricing.
  • Affiliate commissions that don't fit. The program brings volume and losses at the same time.
  • COD without confirmation. Refusals climb and return shipping eats the profit.
  • A new country with no support in the local language. Complaints arrive and nobody answers.
  • Operations at the edge. Late deliveries turn into complaints, chargebacks and pricier ads.

Guides in this hub

Dedicated growth guides (affiliates, cash on delivery and selling to new countries) will join this hub as they are published. In the meantime, these launch guides cover the foundations that come before scale:

To see what Horsee includes for growth, visit the growth page. Payment methods for selling at home and abroad are on the payments page.

From the guide to a live store.

Build your store with AI in minutes, with checkout, payments and server-side tracking ready.