The “best selling” product or product category on an ecommerce site is not normally the one with the highest number of sales or largest dollar amount in sales. It could also change depending on your goals like moving inventory, increasing AOV, and improving profit per customer.
If a product shows up in Google via SEO for a high volume query and with an intent to shop, that product or category going to sell better than everything else. When that ranking eventually goes away, the sales do too and then you find your actual best selling product. The same goes for a product getting a ton of Pinterest or PPC referrals from Merchant Center and datafeeds.
A lower converting or priced product that gets reorders, has regular cross-sells added to cart, and builds customer loyalty is where you want to look as that is likely your best seller. It does not depend on a specific traffic source, but works across your audience as a whole.
To find these in an ecommerce store, try looking at:
- The number of sales and conversion rates from branded traffic coming in
- Conversion rates from non-product specific traffic
- Conversion rates for non-branded traffic
- Repeat purchases and LTV for the product
Here’s a recent example of how we optimized the upsell tree in a client’s checkout, then three ways you can use the above to find your best sellers and hopefully drive more conversions and revenue.
A Real Recent Example With Conversion Rates and Expiration
The client sells a specific type of food and uses WooCommerce. They have FunnelKit installed in their checkout process which has a very easy-to-use logic tree and data display feature.
- The logic section lets you define that if X is in the cart do not show the product, or if A and B are in it, then definitely prioritize ABC as the next product to feature.
- The data section shows you the order of the upsells and the total conversions across all site traffic that reach them at checkout in a list.
This client needs to move desserts faster than other upsells, so desserts were always the primary upsell. It turns out that desserts were not the highest converter compared to complementary proteins and some of the side dishes. But that didn’t mean I should shift these top to the top as that could work against us.
Unlike retail products that can be liquidated, perishables are a loss once they pass the expiration or sell by date.
The food space is a bit unique as you have:
- Product expiration
- Perishables going out with unpredictable weather
- Tariffs reducing margins substantially (this has always been the case historically)
I shifted one of the side dishes to the second position as it was something we stock regularly, has a high conversion rate, and good margins. This put the dessert to the second upsell. After two weeks the side dish kept its strong conversion rate and there was no shift in the dessert’s rate, so the client was now more profitable as there was more product in the cart and more money being made.
Since dessert sales were not impacted, I looked site wide at the biggest protein upsell (think bacon, wagyu burgers, salmon filets, shrimp, jumbo crab meat, etc…). This was already something people would add on based on the conversion rates, but not many had it already in cart because it isn’t a heavily pushed feature on the website. So I moved a protein to the second position making dessert now the third.
The protein in the second upsell position did get added but the conversion rate dropped a bit. The dessert was moved to position four and actually kept its conversion rate. The exposure rate declined a little bit as it was a few clicks in, but not enough to cause concern. This is natural with upsells that aren’t forced click throughs.
At this point total cart value was high and helped offset the expensive shipping costs, but we still need to move desserts. Since upsells weren’t working as fast as we needed them to, we’re running promotions via email and ads, and we’re going to be trying “free dessert with XYZ” types of packages. These tend to move the needle. If the client gives permission in the future I’ll do a case study with numbers or examples in it.
The result was an increased AOV and more profitability by simply knowing what our actual best sellers were in an upsell tree. Now ignore the upsell example and needing FunnelKit. Here’s how to get your best sellers from your own shopping cart whether it is Shopify, WooCommerce, BigCommerce, or a custom built cart.
Sales and Conversion Rates from Branded Traffic
Branded traffic converts high because the visitors know about your business and came ready to shop, or they’re repeat buyers coming back. There is likely going to be a skew towards what they purchased before or know about you. For this part you want to isolate the data groups and then combine them later on.
If you ran ads for a specific product line or category, or had something go viral on social media, that skews as the best seller because you promoted it and got traffic specific to it. If you had no ad budget or the above didn’t happen, those sales will potentially tank.
The first thing to do is to segment those sales out. For the items below, pull the total count and dollar amount for the first two, and the conversion rate for the third:
- SKUs and product types added to cart from non-influenced traffic (traffic that does not have specific product intent).
- Use your tracking parameters like ad groups, keywords, etc… to find influenced traffic.
- How many returns you get from their products as this can make a top seller a lot less profitable than a third best seller.
- The conversion rate on new products from your existing loyal base, and check whether that same base converts on new products when they arrive through branded search versus other channels.
- This help show whether it’s the product is working across the audience or just brand trust carrying it to checkout.
Having this lets you know what your current audience is interested in, and shows the products that convert in general without brand loyalty and external influences. These are products to feature in your emails to existing customers, as well as remarketing ads if they recently shopped for the complementary products but didn’t buy them.
Conversion Rates from Generic Traffic
Generic traffic could be someone asking ChatGPT for the best stores for apples, or someone looking on Google for bras, but not specifying the type, material, or style of bra. This is where they find the brands that sell the products and go to visit them. In both cases the person knows they have a specific purpose and use, but it isn’t defined to a specific variety.
The person shopping for apples could be looking for apples to bake with or to eat as a snack. If it is snacking apples and you have 100 visitors coming in from the word “apples”, you can now find your best selling snacking apple as there’s no modifier other than snacking, it is just apples in general.
If you ranked in Google and got LLM traffic for “fuji apples” and that was mixed in, this skews the data and fujis would win, when in reality it was the honeycrisp or cosmic apple that are actually the best seller. If you focus everything on fuji, you could be less profitable because it is not your best seller, you just included product specific traffic in the mix.
The person searching for bras may want lift and support, so as they reach your page they’ll find shelf/balconette, push up, and side support. In the survey you do after purchase you’ll learn who came in specifically for this style, and then you can cross reference which of these three styles won out over the others. The non-influenced traffic (i.e. didn’t type balconette bra in specifically) is the one that shows your best selling bras for lift and support.
This information lets you know what to place as a higher priority in your product mix, in your shopping ads across online platforms, and to feature in email reminders to replace your undergarments. If most of the traffic was specific to push up, that is why push ups are your best seller, but that doesn’t mean they’re right for the general audience.
Repeat Purchases and LTV
The last way to know what your best sellers are may not be reliant on the number of orders or AOV (average order value) at all. It could be the lifetime value and total number of purchases from the customer. Especially if the return and replace rates are lower than top converting products.
You may have a product category that converts lower than others and has a lower AOV. But these customers stay for five years or purchase 8 times over their lifetime as your customer. The top converting option may have subscribers for 3 years, or that only place 5 orders in the lifetime. While we’d be paying attention to the higher converting and larger order numbers, they may not be as profitable and this is why they’re not necessarily the best sellers.
If the first product has an AOV of $100 and 5 years, that is $500 per customer in gross. The second product has an AOV of $120 and the customer stays for 2 and 1/2 years. This results in $300 gross. The AOV is lower on product 1 but you make an extra $200 in gross profit. And if your company gets refer a friend traffic, you have an additional 2 and 1/2 years of high-value referrals coming in as they’re customers for a longer period of time.
Best selling products are not always the ones that have the highest number of orders or the largest AOVs, and top selling products don’t always mean they’re the most profitable. Top sellers are the ones people gravitate to when they know they need something from a category of products with a specific idea in mind, and then the majority choose one or two options. From there it is a matter of how long they stay a customer and how many repeat orders they make.
By having clean data sets you can discover your best sellers and most profitable products so you know where to invest marketing budget to grow your company for long term success. These are things we do with our clients, especially in merchandising, upsell and cross sell paths, and I hope this helps you do the same.
