Amazon's search demand is not static. New phrases appear every week — a trend, a competitor's campaign, a seasonal shift, a product nobody was searching for a month ago.

TrackIQ counts them for you, on a tile most people read as a total and never look at again.

[Watch the walkthrough](https://kb.trackiq.com/kb-assets/kbvideo-new-queries-explainer-0649b86b.mp4)

*Prefer to watch? Eighty-seven seconds on a tile most people read as a total, and the ten-minute weekly routine behind it.*

## **The number is already on the screen**

![Keywords tracked, with the new count beside it](/kb-assets/shot-sqp-kpis-49353d.png)

**Search Query Performance → Keywords Tracked.**

The tile shows the total, and beneath it a green figure with the definition attached:

> **new = first seen in the latest week**

That is the whole feature. A count of queries that were not in your data last week and are now, refreshed every week, sitting on the dashboard you already open.

It appears on both **Account** and **Products**, so you can ask the question for the brand or for one ASIN. See [**Account vs Products: Two Ways to Read SQP**](https://kb.trackiq.com/hc/knowledge-base/articles/account-vs-products-sqp).

## **Why new queries matter more than they look**

A new query is demand that did not exist for you last week. Three things follow:

**You are almost certainly not bidding on it.** New queries arrive through broad, phrase and auto matching, at broad-match prices, if they arrive at all.

**Nobody has optimised for it yet.** Competition on a genuinely new phrase is thinner than on an established one, which is the cheapest position you will ever get.

**It tells you something about the market.** A cluster of new phrases around one attribute is a shift in how people are shopping your category.

## **The weekly routine**

Ten minutes, once a week:

**1. Read the new count.** Both tabs. A number well above its usual level is worth investigating on its own — something changed.

**2. Find the new queries.** Sort the SQP Analysis table by week and look at the most recent cohort. Queries with data in the latest week and nothing before it are your new arrivals.

**3. Sort them into three piles.**

| Pile | What it is | Do |
|---|---|---|
| **Relevant, converting** | real new demand you are already winning | **harvest it** |
| **Relevant, not converting yet** | too new to judge | watch a week |
| **Irrelevant** | broad match drifting | **negative-match it** |

**4. Act on the first and third piles.** The middle one keeps.

## **Harvesting a new query**

A new query that converted is the best harvest candidate there is — proven, and probably cheap.

Move it to an exact-match keyword, in a campaign where you can see it, and **negative-match it in the campaign it came from** so you are not bidding against yourself. That last step is the one people skip. See [**Harvest Winning Search Terms into Exact-Match Campaigns**](https://kb.trackiq.com/hc/knowledge-base/articles/harvest-winning-search-terms).

**Do not harvest on one order.** A new query with a single sale is a coincidence until it repeats.

## **Reading the count itself**

The trend in the number tells you something the individual queries do not:

| Pattern | Reading |
|---|---|
| Steady low number each week | normal churn |
| **Sudden spike** | a trend, a season, or your match types widened |
| Persistent high number | your broad/auto campaigns are casting very wide |
| Near zero, consistently | you have tight targeting — efficient, but you will miss shifts |

**A spike is not automatically good.** If a bid change or a match-type change widened your net, the new queries are your own doing rather than the market's, and many will be irrelevant. Check what you changed before concluding the category moved.

**Consistently near-zero is a strategic choice**, and it has a cost. Exact-match-only accounts are efficient and blind: they cannot discover the phrase that becomes next year's main term.

## **Where else new demand shows up**

**Recommended Actions → Add as keyword** flags queries you have impressions and clicks on but no keyword bidding, which overlaps with this population and comes pre-sorted by money. See [**Recommended Actions: The Queries That Need a Decision**](https://kb.trackiq.com/hc/knowledge-base/articles/sqp-recommended-actions).

**Brand Keywords** in AMC shows new *branded* searches your advertising caused — a different and complementary signal. See [**Brand Keywords: Proving Upper-Funnel Spend Works**](https://kb.trackiq.com/hc/knowledge-base/articles/brand-keywords).

## **What not to conclude**

**New to you, not new to Amazon.** "First seen in the latest week" means first seen **in your data**. The phrase may be long-established in the category and simply new to your ads.

**A new query is not free traffic.** It arrived through a match type you are paying for.

**Weekly cohorts.** SQP reports whole weeks, so "the latest week" is a complete week, not the last seven days. See [**SQP Is Weekly: Reading Week Cohorts Correctly**](https://kb.trackiq.com/hc/knowledge-base/articles/sqp-is-weekly-reading-week-cohorts).

**Volume takes time.** A new query with a handful of impressions cannot be judged for a fortnight.

## **Related**

- [**Understanding Search Query Performance**](https://kb.trackiq.com/hc/knowledge-base/articles/understanding-search-query-performance) — the dashboard overall.
- [**Harvest Winning Search Terms into Exact-Match Campaigns**](https://kb.trackiq.com/hc/knowledge-base/articles/harvest-winning-search-terms) — what to do with the good ones.
- [**Recommended Actions: The Queries That Need a Decision**](https://kb.trackiq.com/hc/knowledge-base/articles/sqp-recommended-actions) — the same population, money-sorted.
- [**Search Terms and Aggregated Search Terms**](https://kb.trackiq.com/hc/knowledge-base/articles/search-terms-and-aggregated-search-terms) — the advertising-side view of the same searches.