# How to auto-warm your hottest pages

> Let xSpeed Cache Pro rank your URLs by real traffic, recency and content, then warm the top ones on a schedule — so your most popular pages are never cold.

- Category: Cache
- Canonical: https://xspeedcache.com/docs/predictive-warming/

---

Warming every page is wasteful — most sites have a handful of pages doing most of the work, and a long tail nobody reads. **Predictive Warming** spends your crawl budget where it counts: it ranks your URLs by real traffic, recency, and content, then keeps the top ones warm on a schedule. Your popular pages stay fast; the forgotten ones don't burn resources.

> **Where to find it**
>
> 1. In your WordPress admin, click **xSpeed Cache** in the left menu.
> 2. In the xSpeed Cache sidebar, open the **Cache** group.
> 3. Click the **Preloader** card, then the **Predictive Warming** tab.
>
> *Shortcut: open `wp-admin/admin.php?page=xspeed#/cache/preloader/predictive` directly.*
>
> ⭐ This tab is part of **xSpeed Pro**.

![The Predictive Warming settings](https://xspeedcache.com/_astro/predictive-warming-settings.D2cphO06.png)

1. **Enable predictive warming** — the master switch for this tab.
2. **Crawl budget** and **How often** — how many URLs per run, and how frequently.
3. **Skip already-warm pages** — spend the budget on cold pages only.

---

## Settings at a glance

| Setting | Default | What it does |
| --- | --- | --- |
| **Enable predictive warming** | Off | Ranks your URLs and warms the top ones on a schedule. |
| **Crawl budget (URLs per run)** | 25 | How many top-ranked URLs each run warms. Range 1–500. |
| **How often to warm** | Hourly | How frequently the warmer runs — **Hourly**, **Daily** or **Weekly**. |
| **Skip already-warm pages** | On | Don't re-fetch URLs that are already cached and fresh. |

---

## How it works

Each URL gets a score, and the highest scores get warmed first. Three signals feed it:

| Signal | What it favours |
| --- | --- |
| **Traffic** | Pages people actually request. Popularity is the strongest predictor of future popularity. |
| **Recency** | Recently published or updated content, which tends to spike then settle. |
| **Content** | Page type and substance, so a real article outranks a thin utility page. |

The scoring is why this differs from [Crawl Now](https://xspeedcache.com/docs/crawl-now/), which treats every sitemap URL equally and works through them; Predictive Warming decides what *deserves* the budget and spends it there. On a large site that's the difference between warming 25 pages that matter and 25 pages nobody will visit today.

---

## Next to warm

![The ranked queue of URLs to warm](https://xspeedcache.com/_astro/predictive-warming-queue.CiQe6i-S.png)

1. **Next to warm** — the ranked list, refreshable on demand.

This is the panel's most useful feature and worth checking before you enable anything: it shows exactly which URLs the warmer will fetch on its next run, in rank order, with each one's score and why it ranked (a **Recent** badge, for instance).

Treat it as a preview. If the list is full of pages you'd expect to be popular, your budget is well spent. If it's dominated by things like `/404-2/` or `/privacy-policy/`, your budget is probably too high for the amount of content you have — those low-value pages are only in range because the genuinely popular ones ran out.

**Warm now** runs a job immediately rather than waiting for the schedule, which is the fastest way to see whether this does anything for you.

---

## Tuning the budget

**Crawl budget** (default **25**, range **1–500**) is how many URLs each run fetches. The right number depends on how much content you have and how much of it is genuinely popular:

- **Small site (under 50 pages)** — a low budget is plenty. Warming 25 of 40 pages means you're warming almost everything, which defeats the purpose of ranking.
- **Large site** — a higher budget covers more of the popular tail, but each run costs more origin requests.

**How often** offers **Hourly**, **Daily** and **Weekly**, and defaults to hourly. Pair it sensibly with the budget: a large budget running hourly is a lot of sustained background traffic. Frequent-and-small beats occasional-and-huge, because it spreads the load.

**Skip already-warm pages** is on by default and should stay on. Without it you'd spend budget re-fetching pages that are already cached and fresh, which is pure waste — the whole point is to spend it on cold pages.

> ⏱ **Runs fire on WP-Cron**, so on a low-traffic site they can drift or stall, since cron only fires when someone visits. A real server cron hitting `wp-cron.php` keeps timing reliable.

---

## Related

- [How to warm your cache ahead of visitors](https://xspeedcache.com/docs/preloader/)
- [How to crawl your sitemap](https://xspeedcache.com/docs/crawl-now/)
- [How to schedule cache prewarming](https://xspeedcache.com/docs/scheduled-prewarming/)
- [How to activate your licence](https://xspeedcache.com/docs/license/)