How to Increase Online Sales: Find the Broken Number First

The store gets visitors. The ads are running. Orders come in, but not enough of them, and the number has been roughly the same for months. You have tried a discount, a new theme section and a few social posts, and none of it moved anything you could point to. That is the most common situation I see in a store doing under $50,000 a month, and it almost always has the same cause: effort spread across tactics before anyone found out which number was broken.

I read the pages ranking for “how to increase online sales” before writing this. The big ones list fifteen to twenty-five tactics — email, reviews, free shipping, live chat, loyalty programs, influencers — and present every one of them as equally worth doing. The best of them sorts tactics by problem, but assumes you already have enough traffic. None of them tells a small store what to do first, or which of those tactics to skip. That order is the useful part, so that is what this post is.

Why are my sales flat when people are visiting?

Sales are flat when one of three numbers is stuck, and adding tactics does not help until you know which one. Every online store’s revenue is the product of three things: how many people visit, what share of them buy, and how much each buyer spends. Your conversion rate is the middle one — the share of visits that end in an order.

A tactic that improves the wrong number does nothing you can see. A loyalty program improves repeat purchases; if your problem is that nobody finds the store, it has nothing to act on. A traffic campaign brings more people; if the product page loses them, you have paid to make the leak bigger.

Which number is actually broken?

The broken number is the one that has fallen against your own history, or the one that improves most for the least effort. Not the one that looks low against a published benchmark, because those averages mix stores so different from yours that they tell you almost nothing.

Here is the sum, so you can check it. A store with 20,000 visits a month, converting 1.5% of them at an average order of $80, makes 300 orders and $24,000. Improve any one of the three by a tenth and revenue rises by a tenth: 22,000 visits, or a 1.65% conversion rate, or an $88 order all land at $26,400. The levers are worth the same on paper. They are not the same cost to move.

How do I find out in ten minutes?

Open your store analytics, set the range to the last 90 days, and compare it with the 90 days before. If you are on Shopify, the analytics dashboard shows all of this without any setup. Look at four things:

1. Sessions by source. Which channels bring visitors, and which one fell. A drop in one channel is a traffic problem with a specific cause, not a general one.
2. The conversion funnel. Shopify breaks conversion into added to cart, reached checkout, and completed. A low add-to-cart rate points at product pages and price. People reaching checkout and leaving points at shipping cost, payment options or a surprise at the last step.
3. Conversion rate by device. If phones convert far worse than desktops, and most of your visits are on phones, you have found the problem. Google’s own guidance on why page speed matters is written for exactly this case.
4. Average order value and returning customer rate. If the order is small and almost nobody comes back, the fix is in the offer and the follow-up, not in more traffic.

What moves first for a store under $50,000 a month

For a store under $50,000 a month, fix what leaks before you pay for more of what leaks. In the order I usually work through it:

Step 5 is two separate things. Free listings in Merchant Center let your products show on Google Search, the Shopping tab and Images at no cost, and most Shopify stores can turn them on from the same feed their ads would use. Search rankings take longer; Google’s SEO starter guide is the honest place to start, and my Shopify SEO checklist turns it into a list for a store.

What is usually a waste of time at this size?

At under $50,000 a month, most of the long lists are a waste of time until the first four steps are done. Not because the tactics are bad, but because each one needs volume you do not have yet to show a result you can measure.

• Loyalty programs. They reward repeat buyers. If you have few repeat buyers, they reward almost nobody and cost you a monthly app fee and a discount.
• Live chat. Useful when someone is there to answer. An unanswered chat box is worse than none.
• Personalization and recommendation software. It needs lots of traffic and a large catalog to learn anything. A hand-picked “goes with” section does most of the job for free.
• Influencer deals paid up front. Hard to measure, and a single post rarely changes a small store’s month. If you try one, pay per sale with a code.
• Site-wide discounts to fix a slow month. They pull sales forward from customers who would have paid full price, and they teach your list to wait for the next one.
• Redesigning the whole site. The most visible work and rarely the most valuable. Fix the product page and the checkout first; they are where the decision happens.

Should I just spend more on ads?

Spend more on ads only after the store converts the traffic it already gets, because ads buy more of whatever happens after the click. If one visit in a hundred buys, doubling the budget buys twice as many visits that mostly do not.

The reverse is often true. Petticoat Lane cut ad spend 32% year over year at a better return. Less money, spent on the traffic that was already converting, beat more money spread thinly. Whether ads are the right next step for your store at all is the question in is Google Ads worth it, and what you can afford to pay for each new customer is worked through in customer acquisition cost for ecommerce.

How long before I see anything?

Checkout and shipping fixes usually show within a few weeks, because they act on visitors you already have. Email to past buyers shows as soon as it sends. Product page changes need enough orders to read, which at a few hundred orders a month means a month or more. Search rankings take months, and I have written separately about how long SEO takes because the honest answer is longer than most owners are told.

Change one thing at a time, and compare the same number before and after. That is slower than doing ten things at once, and it is the only way you find out which one worked.

Questions people ask