How to increase online sales, that is the question. We all invest money in marketing, campaigns for acquisition or re-targeting to assure the end visitor will finally convert. But if the visitor already knows that he will buy from you, but he just needs some time, what’s the point? In that case we want to know if there’s a clear reason on the webpage that’s possibly hindering sales and/or in marketing.

Here are a couple of check points you should focus first to know if there is something you should do differently to improve your sales.

Funnel conversion and conversion rate

There is no real rule that applies to all business regarding the conversion rates. Except maybe this one: If your conversion rate is more than 61%: and especially over 80%, you can be sure that you’re missing something either in your funnel, or in your tracking implementation. Something is definitively missing. For example if you have a funnel from a customer that is already sitting at the table then I’m pretty sure that the that the likelihood of him drinking something will be over 95%, as your customer has already picked his place and dropped off his coat. You should definitively consider having a funnel that has more steps, and especially making sure that the potential customer is looking at your menu before entering your place.

As stupid as it can be, always consider the largest scope of your audience when thinking about funnels. The aim is not to have the best number of your network, it’s to have relevant data to work with.

Up-sell and cross-sell

Those two strategies are not the best way improve the number of sales online. Instead, they are good for increasing the average purchase value, even though you need to be careful and adjust them to your business and audience.

If a visitor is willing to spend 2000€ in a cruise, then he should be quite OK to spend a couple more hundred euros for an extra such as a restaurant or a spectacle. But if your selling less expensive products, such as pasta (yes people are buying pasta online), then you need to adjust the up-selling strategy. If your visitor will be happy to spend 100 euros for pastas, he will not that much be interested for 40€ vintage boxes decoration, but surely happy to add an extra 10 euros for some Italian olive oil. You always need to adjust the up-selling product regarding the current amount of the cart and its content.

Same for the cross-selling strategy: always suggest different range of price for cross-sell product (1 very cheap, 1 very expensive, 1 in the average for example) so the visitor won’t get the impression that your goal is only about selling. You have to show him/her some interest. Also keep in mind that the main goal is not cross-selling or upselling but the initial product purchase.

AB Testing: take care of your potential product catalog changes

You should know very clearly and without any doubt which product is your best-seller online, which one is your best up-seller, and which one is the best known one (and therefore best for acquisition). If your best-seller is responsible for 80% or more of your online sales revenue, then any impact/changes on the product itself or on its product page will directly affect your sales.

Then it becomes obvious that you should be very careful about doing any bold A/B testing on those pages. Those are your references, and you have plenty of other products to use for those experiments where you can learn from.

Coupon code or not coupon code?

This again depends on your business, and especially on one question: how much you want your customer to convert? Does it make sense time-wise to give your potential customer a little push to convert, even if that reduces your margin? That’s mainly a business decision but in any case, you’ll have to track the usage of that coupon code to know if someone is using it or not, and to understand what the best strategy for your audience: % off on the next purchase, free delivery fees…


Your sales are business data, you’re the best person to analyse them.  If you’re struggling with your analytics tools, wondering if you can trust the data or not, or on how to interpret specific data, then for sure, you can contact us!