Default product recommendations in PrestaShop treat loyal, returning shoppers just like first-time visitors, often showing them items they have already bought. By segmenting customers based on their actual purchase history, category affinity, and price bands, you can transform generic blocks into tailored suggestions that drive repeat sales.
Targeting returning buyers with relevant recommendations is one of the highest-ROI strategies available, as you have already paid to acquire their trust. Whether using a dedicated PrestaShop module or native customer groups, personalizing the storefront experience turns existing purchase data into predictable revenue.
Why Repeat-Buyer Relevance Matters More Than Traffic Growth
PrestaShop recommendations are usually the first thing a returning customer sees, and usually the least relevant thing they see.
To make recommendations more relevant to returning customers in PrestaShop, segment shoppers by order history and behaviour, then drive product blocks from that data rather than the default "viewed products" and "also bought" outputs. A recommendation module or a Cart Rule-backed custom block can personalise by past purchases, category affinity and price band.
Think about the shopper who bought a coffee grinder from you in March. She comes back in November, logs in, and your homepage offers her the same grinder under "Viewed products" plus a "Customers who bought this also bought" row of three more grinders. She has already solved that problem. You know she has, because her order is sitting in the Orders tab of her customer profile.
That gap is a data-asymmetry problem. A logged-in repeat buyer hands you purchase history, category affinity, average order value and a price band. An anonymous visitor hands you a cookie and a wish. Most PrestaShop stores run both audiences through the same default blocks built for the second group, then wonder why returning traffic converts no better than new traffic.
The commercial case is straightforward. This is the cheapest audience you will ever address: acquisition is already paid for, trust is established, and the only variable left is whether the next product you show is worth their attention.
Method 1: Personalise with a PrestaShop Recommendations Module
A module is the practical way to turn stored order history into live product recommendations inside PrestaShop. The steps below assume you have a working catalogue and at least some repeat buyers in the database. If your store is brand new, the segments will still build, but they will take time to fill.
Step 1: Install the Module from the Module Marketplace
The module has to be installed and registered before any recommendation block can read customer data. Without installation, PrestaShop has nothing to hook into beyond its default blocks.
Go to Modules > Module Manager and use the Upload a module option if you have a module archive, or search the marketplace directly from the Modules catalogue. Once uploaded, select Install, then Configure on the module card.
- Status, which switches the whole recommendation engine on or off across the storefront.
- Data source, which decides whether recommendations draw from order history, browsing behaviour, or both.
- Cache lifetime, which controls how long a generated recommendation set is stored before it is rebuilt.
You should now see the module's configuration screen with its settings tabs available in the left column.
Tip: Install on a staging copy first if your live store carries a large catalogue, so you can watch the recommendation set rebuild without slowing real shoppers.
Step 2: Connect The Module To Customer Order and Browsing Data
Recommendations only become personal when the module can read what each customer actually did. A module that only reads the current page will produce the same output for everyone, which is the problem you are trying to solve.
Open the module's Data Sources tab and review the following toggles:
- Order History, which lets the module read past Customer and Order records to see what each logged-in shopper bought and when.
- Browsing History, which tracks product page views so the module can recognise interest without a purchase.
- Cart History, which captures abandoned carts as a signal of intent.
- Guest Tracking, which decides whether anonymous visitors get generic recommendations or nothing at all.
You should now see each data source marked as active, with a summary of how many customer records the module can read.
Step 3: Build Customer Segments
Segments are where the data asymmetry becomes useful. A shopper on their fourth order should not see the same block as someone who arrived ten minutes ago.
Move to the Segments tab and create rules using Add new segment. Common starting points:
- Order Count: Splitting first-time buyers from repeat customers (for example, one order versus two or more).
- Last Order Date: Isolating shoppers who bought recently from those who have gone quiet.
- Category Affinity: Based on which categories dominate a customer's past orders.
- Price Band: grouping customers by the average order value range they habitually buy within.
You should now see your segments listed with their rule conditions and a live count of matching customers.
Warning: Keep segments mutually exclusive where possible. Overlapping rules can cause a customer to match two feeds, and the module will only apply one of them.
Step 4: Assign A Product Feed to Each Segment
Each segment needs its own product source; otherwise, the segmentation has no visible effect on the storefront.
In the Feeds tab, select a segment and define what it receives: Related products from the same category, Cross-sell products from a complementary category, Best sellers within the customer's price band, or a Manual selection you curate by product ID. Set the Product limit per feed so a block never overwhelms the page.
You should now see each segment paired with a feed and a visible preview of the products it will surface.
Step 5: Place The Blocks Via Hooks
Placement decides whether the recommendations get seen or get lost below the fold.
Open the Display settings tab and choose a hook for each feed, such as displayProductAdditionalInfo for the product page, displayShoppingCartFooter for the cart, or displayHome for the front page.
Set Position to control order against existing blocks, and leave Default blocks untouched so the native Viewed products and Also bought modules keep their original placement.
You should now see the recommendation block rendering on the chosen hook, in the position you set, without disturbing the default product page layout.
Method 2: Manual Segmentation and Code in PrestaShop
You can build a workable returning-customer layer without buying anything, using the native customer data PrestaShop already stores. The catch is that most of the adjustment has to happen in templates or a small custom module, because the default "Viewed products" and "Also bought" blocks do not read customer history on their own. This path suits a developer, or a merchant comfortable editing files on a staging copy first.
Step 1: Build Customer Groups That Represent Buying Behaviour
Groups are the cleanest native hook for targeting. A group can be attached to a customer, then used for pricing, cart rules and custom display logic, which makes it the anchor for everything below.
Go to Customers > Groups and use Add new. Set a Name that describes the behaviour rather than the person ("Repeat buyer, high AOV", "Second-purchase window"). Leave Price display method on Tax excluded unless your pricing model needs otherwise, and keep the Discount and Price display fields at zero and default.
Assign customers manually under Customers or automate assignment with a scheduled job that queries orders and customer tables.
You should now see your group listed under Customers > Groups with its member count.
Tip: Name groups by behaviour, not by value tier. "Repeat buyer" survives a replatform; "VIP" invites an argument about who qualifies.
Step 2: Query Order History to Define the Rule
The rule has to mean something specific, or your recommendations become noise. A returning customer with two orders in a repeated category needs a different block from someone who bought once eight months ago.
Write the condition as plain logic first, then translate it:
- Order count of two or more within 90 days, and the last order contains no repeat-purchase item
- Total spend band, grouped into two or three ranges, so you can match a higher-value customer with premium items
- Category affinity, scored by quantity bought rather than order count, so one bulk order does not skew the profile
Run the query against your database on a schedule and write the result to your group membership or a custom field. You should now see each customer tagged with a behaviour profile you can test.
Step 3: Override the Default Recommendation Templates
The default blocks live in your theme, so an override keeps your changes upgrade-safe. Edit copies, never the originals.
In your child theme, copy the relevant template files, such as product.tpl and any block templates rendered in the footer of the product page, into the same path under themes/your-child-theme/. Then wrap the product display loop in a condition on the logged-in customer's group, and swap the product source: for returning buyers, pull from their previously purchased categories rather than the global bestsellers list. Clear the cache under Advanced Parameters > Performance after each change.
You should now see a different recommendation block when signed in as a test customer in your repeating group.
Warning: Test on staging first. Template overrides survive core upgrades only while the file path and PrestaShop's template inheritance stay unchanged.
Step 4: Add Cart Rules for Genuine Returning Buyers
A returning customer who sees a relevant recommendation and then finds a reward for coming back is the cheapest conversion you will get. Cart rules keep that reward inside PrestaShop's native logic.
Go to Catalog > Cart Rules. Set a name, then under Conditions, restrict the rule to your returning group and add a Minimum amount if you want it to trigger only on a meaningful basket. Under Actions, choose the discount type and add a Product selection if you want the reward tied to a cross-sell rather than a flat reduction. Set Validity dates so the rule expires.
You should now see the cart rule badge applied when you log in as a returning test customer and add a qualifying product.
Step 5: Respect Consent Before You Personalise
Purchase history and browsing data are personal data. Showing a returning customer a block built from their own orders is generally within a legitimate interest, but combining it with tracking data or emailing them about it needs a lawful basis under GDPR.
Go to Customers > Customers, open the record and check the consent flags set at registration. Only pass a customer into a personalised recommendation path when the relevant flag is present. Keep the anonymous path as your fallback: bestsellers, category leaders and a simple "recently viewed" list. Customers who have not consented should see exactly what a first-time visitor sees.
You should now see that a customer record with the relevant consent flag missing is routed to the anonymous block.
Tip: Log every consent decision with a timestamp. If a data subject asks why they were shown a personalised block, the answer needs to be retrievable.
This approach works, but it is engineering work: a scheduled query, template overrides that need checking after every theme change, and manual group maintenance. If your team would rather not own that, a maintained module removes most of it, and the FME Modules PrestaShop modules from an official PrestaShop Partner are built for exactly this kind of store logic.
Which Signals Actually Make A Recommendation Relevant?
A returning customer is not one signal but several, and they do not carry equal weight. Purchase recency decays fast, while brand affinity can stay useful for months. Ranking each signal by how much it tells you, and how quickly it goes stale, is what separates a relevant block from a noisy one.
They fall into two groups that behave very differently: implicit signals the customer gave you by accident (what they browsed, how much they spent) and explicit signals they gave you on purpose (a wishlist add, a written review).
| Signal | Type | Use it for | Decay rate |
|---|---|---|---|
| Recency of Purchase | Implicit | Suppressing products already bought | Fast |
| Category Affinity | Implicit | Choosing the next category to surface | Slow |
| Brand Affinity | Implicit | Ranking within a category | Slow |
| Price Band | Implicit | Filtering out items above the customer's ceiling | Slow |
| Replenishment Cycle | Implicit | Timing a reminder or refill suggestion | Clock-based |
| Wishlist and Reviews | Explicit | Weighting what the customer genuinely wants | Very slow |
Price band deserves special attention because it is the signal most PrestaShop stores never use. A customer who has only ever bought from your lower-priced ranges will ignore a block full of expensive items, however well those items match their category. Filter first, then rank.
Replenishment cycle is the one signal that is timing-led rather than preference-led. If a customer buys the same consumable on a repeatable rhythm, the best recommendation is the same product again, shown slightly before they would otherwise think of it.
For a store with thin data, work in this order:
- Exclude what they already own, ranked by recency.
- Filter to their historical price band.
- Weight category affinity above brand affinity.
- Layer in explicit signals, wishlist first, reviews second.
- Leave replenishment for last, once you can identify a real cycle.
If you only implement one rule, make it the price-band filter, because it removes the single biggest source of irrelevant recommendations before ranking even begins.
If you would rather not build this logic yourself, the PrestaShop modules from FME Modules are built to sit on top of the customer data PrestaShop already holds, so the ranking happens without custom code.
Troubleshooting: Why Your PrestaShop Recommendations Still Look Generic
If the personalisation you configured is not showing up on the storefront, the cause is almost always one of four things: caching, an unapplied segment, a missing recency rule, or a theme that never fires the hook your module relies on. Work through the sections below in order, because the first two account for most "it looks the same for everyone" reports.
Logged-in and guest visitors see identical blocks
A full-page cache serving the same HTML to everyone is the usual culprit. PrestaShop's built-in Performance page under Advanced Parameters lets you clear the cache and toggle caching, but a third-party full-page cache layer sits in front of it.
- Cause: The block renders server-side and gets cached before the customer session is known.
- Fix: Exclude the recommendation block from full-page caching, or render it after the session loads. Then clear the cache and re-test in a private window while logged in.
- Also check that the customer is actually being matched to a segment. A guest with no order history falls into your default group, so identical output can simply mean the segment rule never matched.
Product Pages Slow Down After Enabling the Module
A recommendation query that scans the full ps_orders and ps_order_detail tables on every page load will drag response times down as the store grows. Anonymous visitors never trigger that cost, which is why the slowdown often appears only for returning customers.
- Fix: Ask your developer to confirm the order-history lookup is indexed on the customer ID column, and that results are cached rather than recalculated per request.
- Watch for: A module that queries on every render instead of storing a prepared list. A short-lived cache of the generated recommendations usually removes the delay.
The Block Recommends A Product The Customer Just Bought
This is the recency rule failing, not the recommendation logic. Without a filter that excludes recently purchased items, a repeat buyer sees the same consumable they ordered last week, which reads as a broken store rather than a helpful one.
Add an exclusion window that matches your repurchase cycle. A coffee subscription might justify a two-week gap; a laptop purchase should be excluded for months. Test with a real customer account that has order history, not a fresh one.
Customer Groups are ignored entirely
If wholesale buyers and retail buyers see the same block, the theme is probably not executing the display hook your module attaches to. Custom PrestaShop themes sometimes omit standard hooks in their templates.
- Check: Whether the block renders at all when you disable the module. If the slot is empty either way, the hook is missing.
- Fix: Add the hook to the theme template, or switch the module to a native hook the theme already fires, such as the product footer area.
Test every fix with two accounts side by side, one returning and one brand new. That single habit catches caching and segment failures faster than any log file.
Conclusion
You now have a personalised recommendation setup in PrestaShop that reads real customer history rather than guessing. Either method gives repeat buyers a block built from their own orders, category affinity and price band, and the signal table shows which rules to tune first.
Filter by price band before you rank anything, exclude recent purchases, then test with two accounts until returning and new visitors stop seeing the same products. Start with the segment you can measure this week, not the whole catalogue at once.
If you would rather skip the template overrides and scheduled queries, install the FME Modules PrestaShop recommendation modules on a staging copy, build one returning-buyer segment, and compare conversion for that group against your default block after a fortnight.