How to Target Website Visitors Based on Order History

Most onsite campaigns target only current visitor behavior. However, two visitors viewing the same product can have completely different customer value, loyalty, and purchase intent.

One customer may regularly place high-value orders every month, while another purchased once during a discount campaign and never returned. Showing both visitors the same popup or offer often leads to wasted discounts, lower margins, and weaker conversion performance.

Modern onsite personalization in Tango.ad combines real-time behavior with historical order data to help marketers display the most relevant web layers to each visitor.

The most effective segmentation strategies usually combine four areas:

• Customer value

• Purchase timing

• Basket behavior

• Loyalty and product affinity

Customer value

Customer value segmentation helps identify how valuable a customer is based on historical purchasing behavior.

Customer value criteria:

Criterium	Description	values
avgItemsQtyPerOrder	average number of items per order	3
medianValue	typical order value	55
orderCount	total number of orders	4

Purchase intent prediction (purchaseIntent)

Purchase intent prediction estimates whether a customer is approaching their next expected purchase window.

Available criteria:

Criterium	Description	values
purchaseIntent.level	purchase intention	high, medium, low, unknown

For marketers, this becomes extremely practical. The key advantage is timing precision. Instead of interrupting visitors randomly, Tango.ad allows marketers to approach customers when their buying probability is naturally highest.

A customer who usually purchases every 30 days and is currently on day 27 is a perfect candidate for:

• reminder overlays

• replenishment campaigns

• urgency messaging

• personalized recommendations

Meanwhile, a customer with low purchase intention is more suitable for win-back offers, stronger incentives or “we miss you” campaigns.

Basket behavior (Customer)

Basket behavior segmentation identifies how customers typically shop.

Available criteria:

Criterium	Description	values
customer.basketStyle	shopping habbits based on previous orders.	single_item_bias, bundle_bias, mixed
customer.countryCodes	country based on order	CZ
customer.zipCodes	ZIP code	11000

This avoids the common mistake of showing “Buy 3 and save” campaigns to customers who historically only purchase one item at a time.

For example:

• Single-item buyers respond better to simple add-ons and cross-sells

• Bundle-oriented customers are better targets for “complete the set” offers, quantity discounts, and multi-product recommendations

• Based on ZIP code personalize location-wise content like nearest shop

Cusomer Lifetime Value(CLV)

Customer Lifetime Value segmentation helps distinguish between customers who generate occasional revenue and customers who create long-term business value.

Available criteria:

Criterium	Description	values
clv.aovTier	Helps identify whether the customer typically places low, medium, or high-value orders	unknown, low, mid, high
clv.ltvTier	Helps distinguish low-value customers from loyal high-value buyers.	unknown, low, mid, high

Two visitors may browse the same product page at the same moment while having completely different long-term business value. One customer may order every month with high basket value, while another purchased once during a heavy discount campaign.

Treating both visitors identically often reduces profitability.

Typical use cases:

• VIP campaigns for high LTV customers

• Early-access promotions for loyal buyers

• Premium product recommendations for high AOV segments

• Stronger incentives for low-value or inactive customers

Loyalty(Loyalty)

Loyalty segmentation helps marketers understand how active and engaged customers are over time.

Available criteria:

Criterium	Description	values
Loyalty.isChurn	is customer long-term inactive	true, false
Loyalty.frequency	how often customer place order	frequent, occasional, rare, unknown
Loyalty.daysSinceLastOrder	Number of days since last purchase	13
Loyalty., purchaseFrequencyDays	Expected days when next order will be placed	6

This allows marketers to build different experiences for different customer stages.

For example:

• Frequent + high LTV → VIP treatment, loyalty rewards, premium upsells

• Occasional + medium purchase intent → reminder campaigns and personalized recommendations

• Rare + churn risk → win-back campaigns and stronger discounts

• High AOV + bundle buyers → premium bundles and larger basket incentives

• Low AOV + single-item buyers → simpler conversion-focused campaigns

Product affinity (products)

Product segmentation helps marketers understand what customers repeatedly buy and which products they strongly prefer.

Available criteria:

Criterium	Description	values
products., distinctProductsCount	identifies how broad the customer’s shopping behavior is:, low values usually indicate narrow interest or replenishment behavior,, higher values often indicate strong engagement	11
products., favoriteProductIds	identifies products the customer repeatedly purchases or interacts with.	‘PID001’, ‘PID002’,, …

This enables highly personalized campaigns such as:

• replenishment reminders

• accessory recommendations,

• product compatibility campaigns,

• category-specific promotions,

• exclusion of already purchased products,

• personalized homepage banners.

Conclusion

The most successful onsite personalization strategies are rarely based on a single condition. The highest conversion uplift usually comes from combining:

• customer value,

• purchase timing,

• loyalty,

• basket behavior,

• and product affinity.

Instead of asking:“What popup should I show?” Modern personalization asks:

“What is the most relevant message for this exact customer at this exact moment?”