What is a recommendation engine?
A recommendation engine is the part of the online store that automatically selects which products to display to each individual visitor. Instead of you manually pairing products together, it uses the store’s own data to determine which products go well together and displays them where they’ll be most effective.
In practice, these are the boxes your customers recognize from all major online stores: “Others also viewed” below the product, “Others also bought” in the cart, and “Popular right now” on the homepage. Behind these features is an analysis of which products appear together in customer visits and orders.
Shoporama has the recommendation engine built-in
At Shoporama, recommendations aren’t an add-on, an app, or an external service you need to connect to. It’s part of the platform, included in the price, and there’s nothing to install. You simply drag a module into the Page Designer, choose what it should display, and save.
Because the engine runs within your own store, your customer data never leaves the platform. No product feeds or visitor data are sent to a third party, and you don’t have to keep track of an additional data processing agreement.
Five types of recommendations
- Others also viewed —items that other visitors viewed during the same visit. The most “intelligent” type, and the most effective on product pages.
- Others also bought —items that were actually purchased together in real orders. Most effective in the shopping cart, where the customer has already made up their mind.
- Related products —the items you’ve selected yourself. Here, you’re in complete control.
- Popular Right Now —the store’s best-selling items over the past 90 days. Works well on the homepage, in an empty cart, on an empty search results page, and on the 404 page.
- Customer’s Recently Viewed – the individual visitor’s own browsing history.
If you’re unsure, there’s an automatic setting that selects the type based on the type of page the box is on.
It also works when you’re just getting started
The classic weakness of recommendation engines is the first few weeks, when there isn’t enough data. A new product has no history, and a new store has no behavioral data at all. Many solutions therefore fall back on the same bestsellers on every single page, and so the recommendations feel generic right when you’re trying to decide if they’re worth the money.
Shoporama solves this with multiple layers. If there isn’t enough behavioral data on an item, it’s first supplemented with what’s been purchased alongside it, then with items similar to it based on category, brand, features, and price range, followed by popular items in the same category, and finally with the store’s bestsellers. The result is that a box for a brand-new product displays information that’s actually relevant to the item, rather than the store’s four bestsellers.
Built for Speed
Recommendations must not come at the expense of speed. That’s why the page itself is loaded without them, and the box is fetched only after the page has been displayed. Space is reserved in advance so the page doesn’t jump when the products load, and no external files are fetched. All the code is inline and takes up about 5 KB. The results are cached, so two visitors on the same product page within the same hour only require a single database lookup.
This also means that pages with recommendations can still be cached statically, just as before.
You can see the results
Under Statistics → Recommendations, you’ll find views, clicks, click-through rate, orders, and attributed revenue—aggregated and broken down by the five types—with a daily graph and CSV export.
The attribution is intentionally conservative. An item is only counted if the customer has clicked on it in a recommendation box and purchases that exact item within seven days. Then, the item’s amount is counted once per product per order. Impressions without clicks never generate revenue, and the other items in the order are not counted. The figure is therefore a floor, not a ceiling, and it’s designed that way on purpose: a number you can trust is more useful than a higher number you have to take with a grain of salt.
Views are only counted when the box is actually at least half visible on the screen, and your own visits are excluded from the numbers while you’re logged into the admin panel.
Your own theme? It works, too
If you’re using one of Shoporama’s built-in themes, Recommendations is a module in the Page Builder. If you’ve created your own theme, you can still add them without editing code: specify a CSS selector in the admin panel, choose whether the box should appear before, after, or inside that element, and save. If the element isn’t present on a page, nothing happens, so it’s safe to experiment.
If you’re a developer, you can instead position the box precisely using a Smarty tag in the template and give the theme its own template for the product cards.
How to Get Started
- Enable visitor statistics under Account → Webstore if you want to use “Others also viewed.”
- Go to Design → Page Designer and select the page layout where you want the recommendations to appear.
- Add the Recommendations module, set the heading and type, and save.
- Track your results under Statistics → Recommendations.
A good starting point is three boxes: “Others Also Viewed” on the product page, “Others Also Bought” in the cart, and “Popular Right Now” on the homepage.
Ready to give it a try?
Sign up for a free trial and test the recommendations on your own products. There’s no commitment, and you don’t need to provide credit card information to get started.
Frequently Asked Questions
What is a recommendation engine?
A recommendation engine is the part of the online store that automatically selects which products to show each visitor. It uses the store’s own data to determine what people are viewing during the same visit, what items are bought together, which products are similar, and what’s selling best right now. The result is the boxes the customer sees labeled “Others also viewed” and “Others also bought.”
Does the recommendation engine cost extra with Shoporama?
No. Recommendations are part of the platform and are included in the price regardless of how many views, clicks, and orders you have. There’s no pay-as-you-go billing, no extra subscription, and nothing to install.
How long does it take for the recommendations to become effective?
The boxes work from day one because they populate with similar items, popular items in the same category, and the store’s bestsellers until there’s enough behavioral data. The “Others Also Viewed” algorithm runs every night based on the last 90 days of behavior, and the proportion of true behavior-based recommendations increases as traffic grows. A store with a few hundred visits a day typically has a usable foundation within a few weeks.
Can I use this if I have my own theme?
Yes. In Shoporama’s own themes, recommendations are a module you drag into the Page Designer. If you have your own theme, you can instead specify a CSS selector in the admin panel and point to where the box should go, without editing the theme’s code. If you’re a developer, you can place it precisely using a Smarty tag in the template.
Do the recommendations slow down my online store?
No. The page itself is served without recommendations, and the box is loaded afterward, so it never delays the initial display. Space is reserved in advance so the page doesn’t jump. No external files are loaded, the code is about 5 KB, and the response is cached for one hour.
How do I measure the impact of the recommendations?
Under Statistics → Recommendations, you’ll see impressions, clicks, click-through rate, orders, and attributed revenue—both in total and by recommendation type—with a daily graph and CSV export. Attribution is conservative: the recommendation must be clicked, and the item must be purchased within seven days for the amount to be counted.
Can I decide for myself which products are recommended?
Partially. If you select the “Related Products” type, the system displays exactly the items you’ve manually added to the product. The other types are calculated automatically. Regardless of the type, products you’ve marked as “Offline” will never be displayed, and sold-out items are automatically moved to the back of the list.
What do I need to get started?
A Shoporama online store, and visitor statistics must be enabled under Account → Online Store if you want to use “Others Also Viewed.” Then, drag the module into the Page Designer on the pages where you want it to appear, select the type, and save. It typically takes less than five minutes.