> ## Documentation Index
> Fetch the complete documentation index at: https://docs.synscribe.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Reading the SERP Cliff Before You Pick a Keyword

> Before writing anything, we gate keywords on winnability (a visible SERP cliff) and drop the ones we can't win, even high-volume ones. We pick on winnability, not volume.

Before you write a single page, you run a **SERP cliff** analysis to decide whether the fight is
winnable. We select on one thing — a visible "cliff" in the SERP — not volume or
keyword-difficulty scores, so a high-volume keyword gets dropped the moment it fails the cliff.
Here, *"pick fights you can win"* is an actual step in the pipeline: the check a keyword has to
clear before anyone writes a page.

## What a "cliff" is, and why it's the criterion

A SERP has a *cliff* at the first of the top results that either has no exact keyword phrase in its
title tag or sits on a low-authority domain. Above the cliff are high-authority pages built on
purpose to rank for the query, exact phrase in the title. Below it are pages that are there by
accident: low DA, no exact-match title, thin content, ranking only because nobody better has shown
up yet. The cliff is your insertion point, and the earlier it appears, the easier the fight. The
keyword eval also surfaces *"how other sites are getting asymmetric wins,"* the places where a
competitor ranked without really deserving it, which gives you a template to copy. This is exactly
why the small, long-tail fights from [1.1](/theory/thesis) are so attractive: the cliff appears
fastest there.

This inversion is deliberate. In the [Organise-keywords SOP](/sops/organise-keywords#reference-—-the-signal-hierarchy),
volume sits *last* in the signal hierarchy, not first:

```text theme={null}
1. SERP cliff           — can we actually win this?
2. Strategic fit        — does the client want this?
3. Intent match         — right audience seeing our page?
4. No self-competition  — not competing with our own pages?
5. Volume               — how big is the prize?
```

## How we score a cliff

Pull the top seven Google results for the keyword and, for each, note three things: its position,
whether the title tag contains the exact keyword phrase word for word (partial doesn't count, so
"QuickBooks Automation Tools" is not a match for "automate quickbooks"), and whether the domain is
a real brand or a weak site. The cliff is the first position that fails on either title match or
authority. If all of the top five hold, check six and seven; if those hold too, call it 8 or
deeper.

The cliff position tells you what to do with the keyword:

| Cliff | Read      | What we do                                               |
| ----- | --------- | -------------------------------------------------------- |
| 1–3   | wide open | publish now, we can win on day one                       |
| 4–5   | winnable  | build the page and target it                             |
| 6–7   | crowded   | defer to next month, revisit once we have more authority |
| 8+    | locked    | park it, come back after adjacent wins                   |

> **Good to know — rules we don't break:**
>
> * Only the title tag counts for the exact match, never the H1, body, or URL slug.
> * We only scan positions 1–7; a cliff at 8 or deeper isn't worth chasing now.
> * A major aggregator (G2, Capterra, Gartner) with an exact-match title is locked — pick a different keyword rather than fight it.
> * Cliff scores decay: a wide-open SERP today can fill in within a few months once competitors notice, so we publish fast on the ones open now.

## Winnability is a first-class survival gate

Winnability isn't advisory. A keyword doesn't survive without passing it. The **Keyword Quality
Gate** asks *"Does this look like they want to buy? · Does my company offer this (or solve the
same problem)? · Is the source of keyword credible? · **Can I win on this keyword?**"* The
**Discovery Quality Gate** closes with *"Have I looked up those terms and verified I can win?"*
and points you at the SERP gap document to do it. And the live operator prompt says it plainly:
*"look at ALL the keywords right now and run it through the keyword evals,"* which returns
winnability plus asymmetric-win intel, per keyword.

## Operationalize it

* Run the analysis on a batch: [2.2.2 — Evaluate + SERP-cliff a batch](/platform/evaluate-serp-cliff).
* The full cycle discipline (with survival gates + the signal hierarchy): [3.4 — Organise keywords for a cycle](/sops/organise-keywords).