How to Pick Three Topics You Can Own as a Small Team

Competition here means "few strong pages answering this exact intent", not "low search volume". A keyword tool's difficulty score is a useful first filter: tools such as Semrush score difficulty from 0 to 100 and show the top ten ranking pages, which lets you see who you are up against before you commit the Semrush keyword difficulty checker. Surfer's guidance points out that lower-volume, lower-difficulty keywords are often the practical entry point for new sites and small businesses, because larger sites with more backlinks simply outrank them on the hard terms the Surfer guide to low competition keywords.

Use the top ten as a test of intent, not just authority. Open the first three results and ask one question: does each one fully answer the query, or do they partially answer it and pad? A page from a huge domain that half-answers the question is beatable. A dedicated guide with original data, a comparison table and a maintained update history is not, whatever the difficulty score says.

A practical scoring pass, run in one sitting by one person:

If a topic fails two checks, drop it that afternoon. Do not keep it on a "maybe" list; maybe lists are how quarters disappear.

This is the criterion small teams most often skip, and it is the one that decides whether you get cited. Salesforce's small-business guidance makes the point plainly: if your business has a niche, your specific expertise and perspective is something larger companies cannot replicate the Salesforce content marketing guide. That expertise has to live somewhere you can point to.

Before scoring, inventory what you hold. Walk through these six drawers and write down what you find:

1. Support tickets and sales call notes that repeat the same question in the customer's own words.
2. Internal numbers: conversion rates, delivery times, error rates, cost breakdowns, survey results.
3. Screenshots, recordings or logs that show a process actually running.
4. Named examples (with permission) of customers, projects or failures.
5. Opinions from your own practitioners that contradict the consensus.
6. Anything you have measured twice and can compare over time.

Two items from this list is the minimum for a topic to stay alive. Zero or one means you are about to write the same overview as everyone else, and you will lose that race to whoever has the bigger domain. First-party data matters for a second reason: the CMI's B2B research notes that first-party data collection is common while governance and strategy around it lag behind the CMI B2B trends report, which means clean internal numbers are also a differentiating asset, not just raw material.

Be honest about the difference between evidence and enthusiasm. "We know this space well" is not evidence. "We have eleven months of ticket tags showing this error spikes after upgrades, plus three logs" is.

An anchor piece is the one page a stranger would link to, quote and send to a colleague. You must be able to name it before you build it: its working title, the promise it makes, who would cite it and why. If the best you can say is "a guide to X", you do not have an anchor piece yet.

Anchor pieces share three properties:

• They answer the whole intent on one page. Not an intro that funnels to a download.
• They contain something that can only come from you. A table of your own measured numbers, a worked example with real inputs, a before-and-after.
• They age well enough to be linked six months later. Evergreen structure, with dated figures clearly labeled by source.

This is where repurposing thinking helps early. If the anchor piece is strong, it can be shrunk and shifted into a dozen smaller assets across channels without new research, a pattern Multiview describes as taking a high-value asset and breaking it into channel-ready pieces the Multiview repurposing guide. Small teams that plan the anchor first get a quarter of content from one research effort; teams that plan posts first get twelve disconnected posts. The CMI's own guidance on writing a one-page strategy makes the same structural point: decide the topics and formats before the calendar the CMI strategy guide.

Three questions to settle the anchor test:

• Can you write its headline right now, with a specific promise?
• Can you name one person or publication likely to cite it, and why they would?
• Can you sketch its main table or diagram from memory?

Two yeses out of three keeps the topic. One or zero, and the topic is not ready, regardless of how much you like it.

Score each candidate 0, 1 or 2 on each criterion. Total out of six.

The cutoff rule, in plain terms:

• Drop any topic scoring 0 or 1 on internal evidence, or 0 on anchor potential. These are the two criteria you cannot buy your way out of quickly.
• Drop any topic whose total is 3 or below.
• Test the rest with one small piece, roughly a day of work, published on your own site with a clear internal link target.

Write the rule down and apply it to the whole list in one meeting. The point of a cutoff is that it works when you are tired and the topic is interesting, which is exactly when judgment fails.

A word on timing: give a tested piece enough time to show a signal before judging it. Search visibility and citations both lag. Use leading indicators instead (impressions, scroll depth, replies from people you did not ask, inbound links), the same logic product teams use when rejecting vanity signup metrics in favor of actions that show real value the market demand validation guide.

The test piece is not a mini version of the anchor. It is a probe: the smallest piece that proves three things at once.

1. The intent is real. People search, ask or share it. The Surfer guide notes that Google's autocomplete and "People also ask" panels reveal queries people are actively making and that competitors overlook the Surfer guide to low competition keywords.
2. Your evidence lands. A specific number, screenshot or example gets mentioned in replies, shares or support threads.
3. The anchor is buildable. Writing the probe tells you whether the data is clean, whether the table works and how much research the full piece really needs.

A worked example. A three-person team tests four topics in two weeks, one probe each, about a day of writing per probe. Results:

• Topic A (onboarding errors): 40 impressions, two inbound emails from customers, one partner asked to quote the error table. Signal: strong. Keep.
• Topic B (legacy integration): 15 impressions, no shares, but the probe surfaced three additional internal projects worth writing up. Signal: mixed but evidence grew. Keep, with a rewritten anchor plan.
• Topic C (pricing model): 60 impressions, no replies, nothing quotable, because the team had no numbers. Signal: interest without differentiation. Drop.
• Topic D (hiring vs outsourcing): 25 impressions, one comment challenging the comparison. Signal: weak but the challenge points to a better table. Hold one quarter.

One quarter, three topics, one probe already written and reusable. That is the whole yield of the exercise: you spent four days to avoid spending three months on the wrong three topics.

If your bottleneck is producing the probe pieces and the anchor at a steady pace rather than choosing topics, a specialist can carry the build with you once the list is settled: Quanta takes a described piece of work and builds it alongside you.

• Scoring competition by volume. High volume with three thin results is a good topic. Low volume with one definitive owner is a bad one.
• Counting enthusiasm as evidence. See the drawers in criterion two; if nothing came out of them, the score is zero.
• Choosing the anchor after writing. You will end up with three blog posts and no page anyone links to.
• Testing three topics at once with one person. Stagger the probes; a half-finished probe proves nothing.
• Keeping a "later" list. Later lists are where dropped topics go to look undropped. Delete them or schedule a re-score with a date.
• Measuring the test with traffic only. First results are usually impressions and unsolicited replies. Track those explicitly, as the distribution guidance argues for measuring against defined goals per channel rather than raw reach the Mimeo distribution guide.