**ASIN Purchase Upsell** finds the products your customers buy together, and pairs them up: a **lead ASIN**, an **overlap ASIN**, and the share of the lead's buyers who also bought the other one.

It is the raw material for two things — a bundle, and a remarketing campaign aimed at people who have already bought from you.

[Watch the walkthrough](https://kb.trackiq.com/kb-assets/kbvideo-upsell-explainer-df4316a6.mp4)

*Prefer to watch? Eighty-three seconds on lead and overlap pairs, and why the denominator matters.*

## **The headline numbers**

![Lead ASINs, overlap ASINs, overlap users and average overlap ratio](/kb-assets/shot-b28-upsell-top-7fa12d.png)

> *Identify products that customers tend to buy together. Each row pairs a lead ASIN with an overlap ASIN — the overlap ratio shows what share of the lead's purchasers also bought the overlap product.*

| Tile | Reads |
|---|---|
| **Lead ASINs** | primary products in upsell pairs |
| **Overlap ASINs** | companion products surfaced |
| **Overlap Users** | customers buying multiple ASINs |
| **Avg Overlap Ratio** | share of lead buyers who also bought the overlap |

## **Read the base before the ratio**

This is the dashboard where enthusiasm most needs a check, and the check is in the tiles above.

**Cross-purchase is rare.** An average overlap ratio under one percent, and overlap users in the double or low triple digits, is entirely normal — and it means every pair below is built on a handful of people. A ratio computed from three customers is not a finding. It is a hypothesis.

So read the two numbers together:

| Overlap users | Ratio | Reading |
|---|---|---|
| high | high | a genuine pair — act on it |
| high | low | a popular lead, weak companion pull |
| **low** | **high** | **noise** — a big percentage of very few people |
| low | low | nothing here |

The bottom-left row is the trap, because a high ratio is exactly what catches the eye.

## **The two charts**

**Overlap User Count by Lead ASIN** ranks lead products by how many customers bought them **and** at least one other ASIN. *Higher bars surface stronger upsell candidates to remarket.*

**Overlap User Share by Lead ASIN** shows what share of all cross-purchase behaviour each lead product accounts for — *which products dominate cross-purchase behaviour*.

Both are counts, not rates, which is why they are the right place to start. Find the products with real cross-purchase volume first, then look at their ratios.

## **The pair table**

![Lead ASIN, overlap ASIN, overlap users, lead distinct users and the ratio](/kb-assets/shot-b28-upsell-mid-efad9d.png)

Every pair is listed with both products — image, title, ASIN, SKU and a **Dashboard** chip each — plus:

**Overlap Users · Lead Distinct Users · Overlap Ratio · Lead Purchases**, each with a change against the previous month.

The arithmetic is straightforward and worth confirming as you read: **overlap ratio = overlap users ÷ lead distinct users**. Three overlap users against 165 lead buyers is 1.82%.

Seeing the denominator beside the ratio is the point of the table. The charts give you candidates; this tells you how much to trust each one.

## **What to do with a good pair**

**Build a remarketing audience.** The clearest use, and the one the dashboard suggests: customers who bought the lead product but *not* the overlap are the audience. That is a DSP remarketing campaign with an unusually specific creative brief. See **DSP Campaigns**.

**Test a bundle.** If the pair is genuinely popular, a listed bundle removes a step for the customer. Test it before committing inventory — a high overlap ratio says people buy both, not that they want them in one box.

**Check the sequence.** A pair tells you *that* both were bought, not in which order. If one is consistently the entry product, it belongs in your gateway list — see **NTB ASINs Analysis**.

**Sanity-check against the catalog.** Some pairs are obvious (a product and its refill) and some are surprising. The surprising ones are where the value is, provided the base is large enough to believe.

## **What not to conclude**

**Overlap is not causation.** Both purchases may follow the same need rather than one leading to the other.

**Pairs are not ordered.** Lead and overlap are labels for the row, not a claim about sequence.

**This is ad-attributed behaviour**, so purely organic cross-purchase is not represented.

**Small numbers move violently.** A pair going from 2 to 4 customers is a 100% increase and means almost nothing. Watch the change column with that in mind.

**Monthly cohorts.** One calendar month at a time.

## **Related**

- [**NTB ASINs Analysis: Finding Your Gateway Products**](https://kb.trackiq.com/hc/knowledge-base/articles/ntb-asins-analysis) — which product starts the relationship.
- [**DSP Campaigns**](https://kb.trackiq.com/hc/knowledge-base/articles/dsp-campaigns) — where a remarketing audience gets built.
- [**Find your next bundle from what customers already buy together**](https://kb.trackiq.com/hc/knowledge-base/articles/find-your-next-bundle) — the playbook.
- [**Why AMC Data Is Monthly and Sometimes Empty**](https://kb.trackiq.com/hc/knowledge-base/articles/why-amc-data-is-monthly) — the reporting cadence.
- [**Six Ways to Find Product Candidates**](https://kb.trackiq.com/hc/knowledge-base/articles/six-ways-to-find-product-candidates) — using cross-purchase data as a research method.