Tutorial: A Frequently Bought Together Offer
Build the classic bought together block: first the free version and what it can honestly do, then the Pro version backed by your order data. About 15 minutes.
"Frequently bought together" is one of the best-earning recommendation formats in ecommerce, and one of the easiest to build badly. Done well, it says people who bought this also needed that. Done badly, it's a related-products grid with a misleading heading.
This tutorial builds it honestly. That starts with knowing where the "bought together" data comes from.
Where the co-purchase data comes from
The Frequently Bought Together template and sorting rule are free and never locked. But the rule ranks products using co-purchase pairs that the Pro add-on mines from your completed orders. On the free plugin alone, nothing collects those pairs yet, so the rule has no data and has no effect on the order: the other rules decide. Part 1 builds a block that's still useful on the free plugin. Part 2 shows what changes once Pro is mining your orders.
Part 1: The free version
Step 1: Start from the FBT template
Go to Recommend → New Recommend, and under Ready-Made Recommendations, choose Frequently Bought Together.
On the free plugin, the builder opens on Setup with:
- Show On: Product
- Position On Page: Product page · After tabs
- How Products Are Chosen: Rules (Default)
- Filter Type: Same Category & Tags
- Sorting: Frequently Bought Together, Based on orders from: Last 90 Days
- On Design: Grid, heading "Frequently bought together"
This adds a new block. It doesn't touch WooCommerce's own Related products section.
Step 2: Add a fallback sort
Under Sorting, click Add Sorting Rule and add Popularity (Popularity window: Last 30 Days).
| Rule | Setting |
|---|---|
| Frequently Bought Together | Based on orders from: Last 90 Days |
| Popularity | Popularity window: Last 30 Days |
Rules are blended, each counting the same. Where the bought-together rule has data, it pulls the real companions up. Where it has none (on the free plugin, or on a product nobody has bought with anything yet), it stands aside and Popularity decides. Without a second rule, a rule with no data leaves you with random order, reshuffled on every page view.
That second row is the difference between a block that only works on your best-sellers and one that works across your whole catalogue.
Step 3: Make the heading honest
Click Continue to go to Design, and decide whether this block has earned its heading.
On the free plugin alone, this block is really "popular products from the same category". So change Heading Shown to Shoppers to something true, like "Popular in this category" or "You may also like". A heading that overclaims is worse than a modest one.
When Pro is mining your orders and you have real order history, change it back to "Frequently bought together".
Step 4: Pick a position and a count
Go back to Setup. The template chose After tabs, the browsing position further down the page.
Consider Product page · After meta instead. It's higher on the page, close to the price and the add-to-cart button. Bought-together offers work better close to the buying decision than below the reviews.
Then on Design, set Products to Show to 3. A bought-together block works best as a short, specific list; four general products reads like a related-products grid.
Step 5: Save and check
Click Save Changes. It goes live straight away.
Open a product page in a private window. Then open something obscure from a small category and check the block still shows something sensible rather than nothing.
Keep filters light
The automatic Same Category & Tags pool keeps this free version on-topic. Don't add more category filters on top. Once you're on Pro with real data, the most valuable pairs often cross categories (the camera and the memory card), and narrow filters would remove exactly those.
Part 2: The Pro version Pro
With Pro active, the same template card starts somewhere different:
| Free | Pro | |
|---|---|---|
| How Products Are Chosen | Rules (Default) | Store Statistics |
| Position On Page | Product page · After tabs | Product page · After add-to-cart |
| Display Style | Grid | Bundle box |
What Store Statistics changes
Instead of filters picking the pool and sorting rules ordering it, the products come from co-purchase pairs mined from your orders, ordered by how strongly they're bought together. The Sorting section is replaced by a note: "Store Statistics orders by what shoppers actually buy together, so there are no sorting rules to set." Filters still narrow the result.
Two settings appear on Setup:
- Order History Window ("Which past orders to learn from"). The default is Last 90 Days, which is a good start.
- Minimum Co-Purchases ("How often products must be bought together"). The default is 10+ Orders. On a smaller store, start at 5+. A pair has to pass statistical checks as well as the count, so 5+ isn't as loose as it sounds. Store Statistics explains the checks.
The builder always shows this warning: "Store Statistics needs order history, so until enough data builds up it falls back to category & popularity." Shoppers see a normal block in the meantime, never an error.
What the bundle box changes
Instead of three separate cards each with its own button, the shopper sees the products side by side with a Bundle total and one Add bundle to cart button. It's placed right under the add-to-cart form, so it's one decision at the moment the shopper is already buying.
Two things to know:
- Let Shoppers Uncheck Items is on by default. Each item gets a checkbox and the total updates. Leave it on: a shopper who doesn't want one item still buys the other two, where an all-or-nothing bundle would lose all three.
- It needs at least 2 products. With fewer, the block isn't shown rather than appearing as a "bundle" of one. The builder warns you first: "This display needs 2 products and the current configuration yields fewer, so it would be suppressed."
Bundles include simple products only
One button can't add a product that needs a size or colour chosen first, so the bundle box only includes simple products and leaves variable products out. If your catalogue is mostly variable products, the bundle box won't have enough to show. Use Grid or Inline list instead.
If it shows nothing, or only popular products
You're probably below the data threshold, and the block has fallen back to category and popularity. Check the state of the mined data:
wp sprcm stats statusIt reports each window's state and tells you plainly when there isn't enough data. Lower Minimum Co-Purchases before you widen the Order History Window: a lower minimum responds faster and stays more current. Troubleshooting has more.
What you learned
- Know where the data comes from. Co-purchase pairs are mined by Pro. On the free plugin alone, the bought-together sort stands aside.
- Always add a fallback sort. Popularity next to a bought-together rule covers products with no history, and avoids a random order.
- Headings should match the data. Say "bought together" only when it's true.
- Keep filters light for real FBT data. The valuable pairs cross categories.
- Bundles need two or more simple products. Below that, the block is held back rather than shown broken.
Next
- A cart upsell: the next tutorial.
- Store Statistics: the full reference.
- Displays: the bundle box and the other display styles.