How to Increase Sales Online: Check for a Leak Before You Add Anything

Something that used to sell stops selling, or a number that used to climb goes flat, and the instinct is to add: a new channel, a discount, a bigger budget. I read the pages ranking for “how to increase sales online” before writing this, and they all answer that instinct — fifteen or thirty tactics to add, each treated as equally worth doing. Not one of them opens with the cheaper question: is something you already built quietly broken? Fixing a leak is faster and cheaper than adding a tactic on top of it, and it is also the reason the new tactic you are about to buy might not work either.

These four break silently more often than most owners check, in roughly the order I would look at them.

Why do sales stall even when nothing obviously changed?

Sales stall without an obvious cause most often because something that was working stopped, not because demand disappeared. A theme update can quietly change a shipping threshold. An app update can turn off an email flow without an announcement. A supplier delay can take your best‑selling size or color out of stock for three weeks while every other product still shows as available. None of these show up as a dramatic error. They show up as a number that used to be higher.

The ranking guides treat “add a tactic” and “fix what broke” as the same kind of work. They are not. A new tactic takes weeks to show a result and competes for attention with everything else on the list. A broken flow, found and fixed, recovers revenue that was already yours within days.

Are your bestsellers actually in stock?

Pull up your top ten products by revenue over the last 90 days and check current stock on every one, including every size and color variant, not just the parent product. A store page can show as “available” while the specific variant most people actually buy — the mid‑size, the popular color — is the one sitting at zero.

If a bestseller has been out of stock for more than a few days, that alone can explain a flat month better than any traffic or conversion story, and no amount of added spend fixes it — spend just buys more visits to a page with nothing to sell.

Is the abandoned-checkout flow you already have actually sending?

Open your store’s abandoned‑checkout emails or app and send yourself a real test: add something to cart, enter an email, and leave without paying. Then wait. If nothing arrives within the window the flow is set to, you do not have a tactic to add — the ranking guides are right that recovering abandoned checkouts matters, but several of them write about it as something to build, when for most stores past their first year it is a feature Shopify’s own guide to recovering abandoned checkouts describes as sending automatically once configured, and that is exactly the kind of setting that quietly stops firing after an app or theme change without ever telling you.

Common, boring causes: a sending domain that lost its verification after an email platform migration, a discount code baked into the flow that expired and now errors out, or a flow that was paused during a sale and never resumed. Every one of these fails silently. The store owner finds out from a shopper complaining, or never finds out at all.

Can you trust the numbers before you act on them?

Before you decide a channel is underperforming and move budget away from it, confirm the conversion number driving that decision is real. Google Analytics’ own guidance on measuring ecommerce is specific about firing each purchase event exactly once per order — and duplicate firing, a common side effect of adding a second tracking snippet during a theme or app change, inflates the conversions a channel gets credited for without a single extra sale happening.

The reverse failure is just as common and more expensive: a tracking snippet that stops firing after a checkout redesign makes a channel that is still converting look dead, and the obvious response — cut its budget — cuts spend on the thing that was working. Pull your order count directly from your store admin for the same period your ad platform reports conversions for, and check the two numbers are in the same range before trusting either one to make a budget decision.

Where did your repeat customers go?

A store can have healthy traffic and a healthy conversion rate and still have flat revenue if every customer buys once and never comes back. Pull your returning‑customer rate for the last 90 days and compare it with the 90 days before. A falling number here usually is not a sales problem at all — it is the absence of a reason to come back: no email after the first order, no second offer, nothing that reminds someone the store exists before they forget about it.

On the Shurhold account, a relationship now in its seventh year, spend grew 35% one year and revenue grew 44% over the same period — revenue outpacing spend is what it looks like when what improves is retention and the existing funnel rather than only the volume of new traffic bought. That gap between the two numbers does not come from adding a new channel. It comes from more of the people already arriving buying again.

The ten-minute check for each leak

None of these four require new spend, a new channel, or a new tactic. They require looking at something you already have and confirming it still works, which is the step every tactic list skips because it assumes the answer is yes.

Should you add spend before fixing these?

No — not until these four come back clean. Adding spend, a channel, or a promotion on top of an out‑of‑stock bestseller, a dead checkout‑recovery flow, untrustworthy conversion numbers, or a store nobody returns to does not fix any of those problems. It pays for more traffic to arrive at the same leak, which is a more expensive way of discovering the same problem you could have found in half an hour for free.

Once the four checks come back clean, the traffic‑versus‑conversion‑versus‑order‑value framework in how to increase online sales is the right next step for deciding where new effort actually goes. That post assumes the foundation is sound. This one is how you confirm it actually is before you spend anything building on top of it.

Questions people ask