Building a maritime keyword research process from scratch

Maritime keyword research using only Ahrefs or Semrush will produce a list that misses most of the queries that actually drive pipeline. The tools are calibrated for B2C and high-volume B2B; they filter out the long tail where maritime intent lives. A research process built around them will give you a content programme aimed at queries no buyer ever types.

A useful maritime keyword research process has six steps. Each is unglamorous; together they produce a research output worth building content from.

Step 1: Map the buying decisions

Before you touch a keyword tool, sit down with your sales team and map the actual buying decisions your services support. For a ship management company, that might be:

• Owner switching managers
• Owner adding capacity (new vessels)
• Charterer evaluating outsourced operations
• Equipment manufacturer evaluating service partners

For each decision, what triggers it? What’s the buyer trying to learn? What’s the timeframe? Without this mapping, you’ll research keywords that nobody searches in the actual buying journey.

Step 2: Pull existing demand signals

Three sources, in this order:

1. Search Console queries from the last 12 months. Filter for impressions over 5, position 4-30 (where you have ranking potential), commercial intent in the query. This is what your site is already partially visible for.
2. Site search analytics. What are people searching once they land on your site? These are the questions your content isn’t answering.
3. Sales call transcripts. Run them through a tool like Otter.ai. The questions buyers ask in the first three calls are gold. They’re the queries that, written about properly, would have shortened the sales cycle.

Step 3: Run competitor SERP analysis

Pick five competitors, pull their top 200 ranking pages by traffic value (Ahrefs is fine for this) and look for:

• Pages ranking for queries you should be visible on but aren’t
• Pages where the competitor is in position 4-10 (potentially beatable with better content)
• Topical patterns where one competitor has a cluster you don’t

Don’t try to copy. Use the analysis to find gaps and angles, not exact queries to chase.

Step 4: Buyer interviews

Pick five recent buyers (won, lost or in-flight). Ask three questions:

1. “If you had to find a supplier like us tomorrow, with no brand recognition, what would you type into Google or ChatGPT?”
2. “What were the questions you needed answered before you made a decision?”
3. “What did you read or watch during your evaluation?”

The first question gives you queries. The second gives you content angles. The third gives you publishers to target for links and citations.

Step 5: Cluster the queries

By this point you’ll have a messy spreadsheet with 300-800 candidate queries from various sources. Cluster them by buyer intent and topic, not by exact-match keyword. A cluster might be “EEXI compliance services” with 25 queries underneath ranging from “what is EEXI” to “EEXI verification cost UK provider”. One page can serve the whole cluster.

The clustering is what stops you from writing 300 thin pages. It also surfaces where your existing content has overlap and could be consolidated.

Step 6: Score and prioritise

For each cluster, score on:

• Buyer intent (1-5)
• Existing content gap (1-5, where 5 means you have nothing close)
• SERP weakness (1-5, where 5 means the current top 10 is beatable)
• Strategic fit with your service offer (1-5)

Multiply through. The top 15 clusters become your next 12 months of content. The next 30 become roadmap. Everything else parks.

What good output looks like

A useful maritime keyword research deliverable is a clustered list of 15-20 priority topics, each with:

• The cluster name and target page brief
• 5-15 queries that the page would target
• Buyer intent and decision-stage notes
• SERP analysis showing what currently ranks and what would beat it
• Estimated effort to produce a genuinely better page

That’s a year of pipeline-relevant content right there. Not a 12,000-row keyword spreadsheet that nobody acts on.