Chapter 14: Workflow, Career and Judgment

First 90 days in a new SEO or GEO role

Days 1–30 — inventory and quick wins

• Week 1: get access and get numbers. Analytics, search console, rank tracking, the site’s crawl tooling, tag manager, and the CRM. Run a full crawl and export titles, meta, headings, status codes, canonicals and internal link counts. Read sixteen months of search performance, and the list of pages that are not indexed. Then book thirty minutes each with product, sales and support — what customers ask for, what they object to, what they cannot find.
• Week 2: find the cheap wins. Content-gap the three closest competitors. Pull the queries where you rank 4–15 with a below-average click rate and rewrite those titles. Check the LCP, INP and CLS on the homepage and the five pages that carry the most traffic, and file the top three performance tickets. Fix 404s and redirect chains.
• Weeks 3–4: start the pipeline. Build the query map and a content calendar. Write the first five briefs. Refresh five older pages that sit at 11–20. Turn on rank tracking with alerts, and write down the baseline you will be judged against.

Days 31–60 — build the system

• Ship the first batch of new pages against the brief standard, not against a deadline.
• Start the authority work that does not depend on budget: profiles, directories, the first outreach round.
• Run the first GEO baseline: a fixed prompt panel, crawler access verified, llms.txt published.
• Set the content audit rule: flag everything with no traffic, no links and no update in twelve months, and decide update, merge or remove.
• Automate one report so that the monthly review stops consuming a day.

Days 61–90 — prove and adjust

• Take the thirty-day read on the first batch: what moved, what did not, and why. Write the why down; it is the only part that compounds.
• Reallocate toward the clusters that responded and kill the direction that did not, out loud.
• Deliver the first quarterly review: what was spent, what changed, what is next. One page of numbers and one page of decisions.
• Leave behind five documented processes, so the work survives your holiday.

Interview questions and answer frameworks

Q1: You join a company that has never done SEO. The CEO wants 100,000 organic sessions a month within six months. How do you respond?

Framework: Agree with the direction, then replace the number with an evidenced one. Establish the baseline — domain authority, indexed pages, current traffic, competitor scale. On a young domain, going from near zero to six figures a month is typically a multi-year project, not a two-quarter one, and saying so now is cheaper than explaining it in month seven. Decompose the target — work backwards from sessions to impressions to rankings to pages: how many keywords need to reach the top three, how many pages that implies, how many links those pages need. Now the conversation is about inputs, which you can control, instead of an output, which you cannot. Offer scenarios — conservative, base and aggressive, each with the resourcing it requires and the assumptions it depends on. And commit to early proof — a list of quick wins deliverable in thirty days, so the first month produces evidence rather than a plan.

Q2: How do you explain to a non-technical CEO why SEO takes months, without sounding like you are making excuses?

Framework: One analogy, then leading indicators, then a date. The analogy: paid search is buying fruit at the shop — you pay and eat today, and stop paying and stop eating. SEO is planting an orchard — the first months are soil and irrigation, and nothing appears above ground. The analogy only works if you immediately show that the trees are alive, which is what leading indicators are for: month one, indexed pages and technical health; month two, impressions climbing and more queries entering the top twenty; month three, first clicks and first earned links. Then commit to a review date and the specific question you will answer at it. What destroys credibility is not slowness; it is a promise with no interim evidence and no review point.

Q3: A stakeholder wants to ship a tactic you believe is risky. How do you handle it?

Framework: Write down the disagreement in the terms the business uses — expected upside, downside, reversibility, and the cost of being wrong. If the tactic is against platform policy, say plainly that the downside is not a dip but removal, and that you will not put your name on it; offer the compliant version that captures most of the value. If it is merely aggressive rather than prohibited, ask for a bounded test with a stop condition agreed in advance, and log the decision so the outcome becomes organisational knowledge rather than a private argument. The professional failure here is not losing an argument; it is implementing something you know is wrong and then being surprised when it breaks.

Q4: What separates a senior practitioner from a mid-level one?

Framework: Three things. They pick the right problem — a mid-level practitioner executes a backlog; a senior one asks which ten of the four hundred items will actually move the business, and is willing to say the rest can wait. They design the evidence — they decide in advance what would prove the work succeeded and what they will compare against, so the conversation afterwards is about results rather than impressions. They build capability, not dependency — documentation, templates and handover, so the team can run without them. Add the unglamorous fourth: they tell people bad news early and precisely.