In short: A GEO strategy is three decisions, not a document. Which questions you want to be the answer to, which of the five filters you are currently failing, and what you will stop doing to pay for the work.
Anything longer than that is usually a plan to look busy. The test is whether somebody could disagree with it: a strategy that nobody in the room could argue against has not chosen anything.
And a strategy written before a baseline is a plan to be unable to tell whether it worked, which is the most expensive mistake available here because it costs a quarter before anyone notices.
Three Decisions
Most documents called a GEO strategy are a description of the field with a budget attached. The parts that make it a strategy are the parts that exclude something.
The three: which questions you want to be the answer to, which step of the sequence you are currently failing, and what you will stop doing to fund the work. Each one rules something out. A document that rules nothing out is a briefing.
The rest of this module is the first two. The third is not a GEO problem and this course cannot help with it, beyond one observation: work funded from new budget survives a bad quarter and work funded by cutting proven search does not.
Why It Has To Follow A Baseline
The lesson on association ended with an hour of measurement: eight questions, three fresh runs each, sources recorded. Everything in a strategy depends on having done it.
Without it you cannot name the failing step, so the plan defaults to doing everything at once. You cannot tell a result from drift, because you never measured how much your own runs disagree. And you cannot say what to stop, because nothing has been shown to be working or not.
The order that wastes a quarter: write the strategy, do the work, then measure. At the end you have a number with nothing to compare it against, and the honest answer to "did it work" is that nobody can say. That is not a reporting problem to fix later; it is decided at the start.
Naming The Failing Step
The sequence from module three is what a diagnosis attaches to. A system rewrites the question, retrieves candidates, discards most of them, reads the survivors, and names a few. You are eliminated at the first step you fail.
- Not resolvedDescribed wrongly or not at all when asked by name. Fix the naming, the category words and the stated relationships
- Not classifiedAccurate when named, absent from category questions. Off-site work, and the slowest of the four
- Not retrievedIn the category, never in the answer. Check crawler access and conventional indexing first, in Bing as well as Google
- Not selectedRetrieved and dropped. The page is diffuse, or its answer only makes sense in the context of the page around it
A strategy names one of these as the constraint and sequences everything else behind it. Naming two is usually a sign that the baseline was not run properly.
What Does Not Belong In It
Three things turn up in most GEO strategy documents and none of them survive contact with the mechanism.
Platform-specific tactics. The platforms differ in how they cite and which sources they favour, and they change without notice. A strategy built on one platform's current behaviour has a shelf life measured in months.
A content volume target. Publishing more pages does not address any of the four failures above. If you are not classified into the category, forty more pages are forty more pages nobody retrieves.
A tool purchase. Tools measure. None of them can tell you which step you are failing, because that judgement needs somebody to read the answers and notice what is missing from them.
What a good one looks like on paper: a page. The ten questions, the named failing step with the evidence for it, the work in order, the measurement that will be repeated, and the thing being stopped. If it runs longer than that, the extra pages are usually there to survive a meeting rather than to direct any work.
Key Takeaways
- A strategy is three decisions: which questions, which failing step, and what you stop doing to pay for it. Each one excludes something.
- A document nobody in the room could disagree with has not chosen anything.
- Write it after the baseline, never before. Measuring at the end leaves you with a number and nothing to compare it against.
- Name one failing step and sequence the rest behind it. Naming two usually means the baseline was run badly.
- Platform tactics, content volume targets and tool purchases are not strategy: none of them addresses a named failure.
Check yourself
Before you move on
Not scored, not recorded, and not part of the certificate. Both answers are settled by a sentence in this lesson, and the reasoning appears whichever option you pick.
- 01
What makes a document a strategy rather than a briefing?
- 02
Why does a strategy have to follow the baseline measurement rather than precede it?