Personalized product recommendations aren’t just for flashy B2C shops. If you’re running a B2B site, your buyers expect the same “just for me” experience—and you probably want to drive bigger carts and repeat orders. But here’s the thing: B2B buyers are different. They don’t want to “discover” a random stapler—they want to reorder 200 boxes of the same gloves every quarter, and they want it fast.
This guide is for B2B marketers, merchandisers, or developers who want to set up product recommendations in Optimizely without getting lost in a sea of features built for someone else’s business. We’ll walk step-by-step through what to do, what to skip, and how to avoid the usual pitfalls.
1. Get Real About What Personalization Means for B2B
Before you start clicking around, take a minute to ask: what do your customers actually want? In B2B, “personalization” often means:
- Showing buyers their contract pricing and only the products they’re allowed to buy
- Recommending frequently ordered items, not just “trending” ones
- Making reorder and bulk purchase easy
- Nudging buyers to complete their usual set (think: “You typically order gloves and masks together”)
What usually doesn’t work? Stuff like “You might also like” widgets pushing unrelated products, or generic “Others also bought…” suggestions. That’s B2C thinking; it rarely moves the needle in B2B.
Pro tip: If you’re not sure what your buyers want, ask your sales or customer service team. They’ll tell you what actually gets ordered together—and what just annoys people.
2. Set Up the Basics in Optimizely
Assuming you’ve got an Optimizely account and your product catalog loaded, let’s get started. (If you’re still evaluating, skip this article for now—get the basics working first.)
a. Confirm Your Data Is Ready
Personalization is only as good as your product and customer data. Double-check:
- Product Catalog: Make sure every product has up-to-date data (name, SKU, pricing, category, images, etc.). In B2B, contract or customer-specific pricing is key.
- Customer Data: Ideally, you have accounts or logins for your buyers. If you don’t, personalization will be limited.
- Order History: The system needs to know what people have bought before. Import order data if it’s not already there.
If your data is messy, fix that first. No recommendation engine can paper over garbage data.
b. Enable Recommendations
In Optimizely, product recommendations are usually part of the “Personalization” or “Product Recommendations” module. Depending on your setup (Commerce Cloud, B2B Commerce, CMS, etc.), the steps might look a little different, but the basics are:
- Activate Recommendations: Go to the admin panel and turn on the Product Recommendations feature.
- Integrate Tracking: Make sure the site is tracking page views, add-to-carts, purchases, and logins. This is what teaches the system what to recommend.
- Connect Your Catalog: Double-check that your product feed is connected and updating regularly.
Don’t overcomplicate it early on. Out-of-the-box placement options (homepage, product detail page, cart) are fine for a first pass.
3. Pick the Right Recommendation Strategies
Now comes the fun part. Don’t just turn on every widget Optimizely offers. Choose what actually helps your buyers:
a. “Frequently Purchased Together” and “Reorder” Modules
- Best for: Buyers with regular replenishment needs
- Why it works: B2B buyers hate hunting for stuff they already know they need. Show them their usual products first.
b. “Recommended for You” Based on Past Purchases
- Best for: Logged-in users with order history
- Why it works: This helps buyers spot products they’ve forgotten to reorder—and increases basket size.
c. “Customers Like You Bought”
- Best for: When you have enough data to segment by industry, company size, or buying group
- Why it works: If you’re selling to a variety of verticals, this can surface relevant products for specific sectors.
d. “Contract or Permission-Based Recommendations”
- Best for: Companies with different product entitlements or contract catalogs
- Why it works: Don’t recommend products your buyer can’t actually purchase. Optimizely lets you filter recs by user group, contract, or permissions.
What to skip (at least for now):
- “Trending” or “Popular” products (unless you sell commodities)
- “You might also like” on unrelated items
- Too many recommendation widgets on one page (it just looks desperate)
4. Build and Place Your Recommendation Blocks
Let’s talk logistics—how do you actually show recommendations on your site?
a. Use Optimizely’s Built-In Widgets
Optimizely has prebuilt blocks for recommendations. These usually require minimal setup:
- Drag and drop the “Product Recommendations” block onto your page templates (homepage, category page, product page, cart, etc.).
- Configure each block: choose which recommendation logic to use (e.g., “Reorder,” “Frequently Bought Together”).
- Set rules for when to show each block (e.g., only show “Reorder” to logged-in users).
b. Customize the Display
Out of the box, these widgets are… fine. But B2B buyers don’t need shiny carousels. Keep it simple:
- Show key info: SKU, contract price, inventory status, “Add to Cart” button
- Don’t crowd the page: Limit to 3–5 recommendations per panel
- Think mobile: Many B2B buyers are on the shop floor or in the field
c. Respect Permissions and Contracts
Make sure every recommendation respects user permissions, pricing, and contract terms. Optimizely supports this, but you have to configure it—don’t assume the defaults won’t leak cross-catalog products.
Pro tip: Test with multiple user accounts to make sure nobody sees products or prices they shouldn’t.
5. Test, Tune, and Ignore the Hype
Personalization is never “set it and forget it.” Here’s how to keep it honest:
a. A/B Test Placement and Content
- Try different recommendation types on different pages (homepage vs. cart vs. PDP)
- Test the number of items shown (3 vs. 5 vs. carousel)
- Measure actual results: Are people clicking? Reordering more? Or just scrolling past?
b. Watch Out for False Positives
Just because a widget gets clicked doesn’t mean it’s helping sales. Look for:
- Increased order size or frequency
- Faster checkout for repeat buyers
- Fewer abandoned carts
If you’re just getting more random clicks and no sales lift, dial it back.
c. Don’t Get Sucked Into Over-Engineering
Ignore the siren song of “AI-powered hyper-personalization” unless you have the data and resources to do it right. Most B2B companies get 80% of the value from just showing the right products to the right company, with contract terms in place.
6. Common Pitfalls (and How to Dodge Them)
Even smart teams can get tripped up. Here’s what to watch for:
- Ignoring Data Quality: If your catalog or pricing is messy, recommendations will embarrass you.
- Showing Products Users Can’t Buy: Double-check permissions and contract filtering.
- Overwhelming the Page: One or two well-placed recommendations beat a cluttered sidebar.
- Not Measuring Impact: If you can’t prove recommendations are helping, they’re just decoration.
7. Take It Live—Then Keep It Simple
Once your blocks are in place, go live and watch what happens. Don’t expect miracles on day one. Instead:
- Start with your top customer segments or biggest accounts
- Tune placements and recommendation types based on real feedback (talk to buyers, not just your analytics dashboard)
- Iterate—don’t overhaul everything at once
Remember: In B2B, the best personalization helps buyers get their job done faster and with fewer headaches. Skip the flashy features and focus on what actually helps your customers.
That’s it. Personalized recommendations in Optimizely aren’t magic, but when you stick to the basics—clean data, the right logic, and simple presentation—they can quietly drive more revenue and happier buyers. Keep it straightforward, keep testing, and don’t let the promise of “AI” distract you from helping your customers buy what they actually need.