How to Do Keyword Research for SEO: A Guide for New Online Business Blogs

Most people who start an online business blog make the same mistake on day one. They write about things that feel relevant, publish a few articles, wait a few months, and then wonder why no one is reading. The problem is almost never the writing. It is the keywords. They either picked topics with no search volume, went after terms they had no chance of ranking for, or wrote content that attracted the wrong kind of visitor. All three of those problems have the same fix: learn how to do keyword research for SEO before you build a single piece of content.

Keyword research is not a creative exercise. It is closer to market analysis. Before a retailer stocks a product, they check whether people are actually buying it and whether the competition has already captured the shelf space. Before you write a single word of content, you need to do the same thing for each keyword on your list. The good news is that the data is all publicly available and most of the tools that surface it offer free tiers. The skill is in knowing which numbers to look at and how to interpret them together.

This guide walks through the full process: what to measure, what the numbers actually mean for a new site, how to match keyword intent to the right type of content, and how to build a working list from scratch. The examples throughout are drawn from a real niche research session using live Ubersuggest data.

Why Learning How to Do Keyword Research for SEO Matters Most at the Start

A new website has no authority. Google does not know who you are, you have no backlinks, and your content has no track record. In that position, choosing the right keywords is not a nice-to-have. It is the only lever you control. You cannot outspend established sites on links or outrank them on brand recognition. But you can choose keywords they are ignoring, or serve search intent better than they do on specific long-tail queries. That is where new sites grow.

The mistake that kills most new blogs is starting with high-volume, high-difficulty terms. It feels logical: more searches means more traffic. But a keyword with 300,000 monthly searches and a difficulty score of 73 is not an opportunity for a site with zero domain authority. It is a dead end. Your content will rank on page 8 and receive no traffic at all. Meanwhile, a keyword with 480 monthly searches and a difficulty score of 20 can put you on page one within months, because the competition is genuinely beatable.

The goal of keyword research for a new site is not to find the biggest keywords. It is to find the most winnable ones that still attract the right audience.

The Four Numbers That Tell You Whether to Pursue a Keyword

Every keyword evaluation comes down to four core metrics. Once you understand what each one actually measures, the decision on any keyword becomes straightforward.

Search Volume: Is Anyone Looking for This?

Search volume tells you how many times per month a keyword is entered into Google. It seems like the most important number, but it is frequently the most misleading one for beginners. Volume is context-dependent. A keyword with 1,000 monthly searches in a niche where buyers are actively looking to purchase is more valuable than one with 50,000 searches driven by casual curiosity. Volume tells you the size of the audience. It does not tell you whether that audience will ever buy anything.

For a brand-new blog, you are generally looking for keywords in the 100 to 5,000 monthly search range. Below 100, the audience may be too small to justify the content investment. Above 5,000, the competition is usually severe enough that a new site will not crack the first page for years. The sweet spot shifts slightly depending on your niche, but this range gives you enough traffic potential to matter while keeping the competition realistic.

There is also a distinction worth understanding between search volume and traffic potential. Search volume counts how many times a keyword is searched per month. Traffic potential estimates how much actual traffic the top-ranking page for that keyword receives. The two numbers are often very different. For many informational queries, Google now serves an AI Overview or a featured snippet at the top of the results page, which means a significant portion of searchers never click through to any website at all. A keyword with 3,000 monthly searches might realistically deliver only 400 to 600 clicks to the top-ranking page once you account for this. When you use Ubersuggest or any other tool, treat the volume figure as a ceiling on potential traffic, not the actual number you will receive.

Live Data from Ubersuggest

Searching “keyword research” returns a volume of 301,000/month with a difficulty score of 73. That same search reveals that “keyword research seo” gets 22,200/month at difficulty 20, and “keyword research for small business” gets 260/month at difficulty 39. Three very different opportunities hiding inside the same topic.

Keyword Difficulty: Can You Actually Win This?

Keyword Difficulty (abbreviated as KD or SD depending on the tool) is a score from 0 to 100 that estimates how hard it would be to rank on page one for a given term. It is calculated by analyzing the authority and backlink profiles of the sites currently ranking in the top 10. A score of 80 means those top results are held by sites with thousands of backlinks and years of domain history. A score of 15 means the top results are held by sites you could realistically compete with in 6 to 12 months.

The practical threshold for a new site is a difficulty score below 30 to 35. Some content strategies push this to 40, but only when the site already has some traction. At launch, stay disciplined. Every article you publish on a high-difficulty keyword is content that may never see page one and will not build your organic traffic base. Every article on a winnable keyword compounds over time.

Search Intent: Will This Traffic Do Anything Useful?

Search intent is the purpose behind a query. Someone typing “blood glucose monitor” wants to buy one. Someone typing “how does a blood glucose monitor work” wants to understand something. Someone typing “best blood glucose monitor for type 2 diabetes” is comparing options before making a decision. These three people are at completely different stages of the buying process, and they need completely different content to be served well.

Intent is usually classified into four types: informational (learning something), commercial (comparing options before buying), transactional (ready to buy now), and navigational (looking for a specific brand or site). The content you create needs to match the intent of the keyword you are targeting. If you create a product page for an informational query, or a blog post for a transactional query, you will rank poorly and convert worse. Intent matching is one of the most commonly overlooked factors in keyword strategy.

One additional variable worth factoring in: Google increasingly places AI-generated summaries, called AI Overviews, at the very top of results for informational queries. When an AI Overview appears, fewer people click through to individual websites, because the answer is already visible on the results page. This does not mean informational keywords are not worth targeting. It does mean that if you are choosing between two informational keywords of similar difficulty, the one where the search results show mostly blog posts with no AI Overview is likely to drive more actual clicks. Check the live results page for any keyword you are seriously considering before committing to it.

A Real Keyword Research Example: The Diabetic Supplies Niche

Theory is useful, but the numbers tell the real story. Here is a live example pulled directly from Ubersuggest using the diabetic supplies space, one of Anton’s actual business categories. The goal is to show how dramatically different keywords within the same niche can be, and how to read those differences correctly.

The highlighted rows tell the real story. “Blood glucose monitor” has tremendous volume but a difficulty of 79. A new site has effectively zero chance of reaching page one. Meanwhile, “blood glucose monitor for dogs” has the same volume (4,400/month), a CPC of $1.27, and a difficulty of just 18. That is a winnable first-page position that you could realistically capture within 60 to 90 days of publishing quality content. The niche is narrow, but 4,400 monthly searches from people specifically looking for a glucose monitor for their dog is a highly qualified, commercially ready audience.

This is the pattern you are looking for in every niche: sub-keywords and long-tail variants where the competition has not yet invested, and where the audience intent is specific enough that conversion rates are high.

How to Do Keyword Research for SEO: A Step-by-Step Process

The process below is tool-agnostic, but it is most efficiently done with Ubersuggest, which gives you volume, intent, CPC, and difficulty in a single view with a free daily query allowance. The steps are the same regardless of what tool you use.

1. Start with your niche topic, not specific keywords. Open the Keyword Ideas section of your tool and enter a broad topic word related to your product or service category, not a specific keyword phrase. For example: “glucose monitor,” “leather wallet,” “virtual assistant services,” “online course platforms.” This gives the tool enough surface area to return hundreds of variations.
2. Check what is already driving traffic to competitor sites. In Ubersuggest, go to Competitive Research and open Top Pages by Traffic. Enter a competitor’s domain and you will see every page on their site ranked by estimated monthly visits, along with the keywords driving that traffic. This is one of the fastest ways to find proven keyword opportunities: topics where real demand already exists and where a competitor has validated the audience. Run this for two or three sites in your niche before building your keyword list.
3. Look at the Related and Questions tabs before the main list. The main Autocomplete list shows the most-searched variations, which are usually the most competitive. The Related tab and especially the Questions tab surface long-tail queries like “how do I,” “what is the best,” or “can I use X for Y” that often have far lower difficulty because they are more specific and get less aggressive targeting from established sites.
4. Filter by SEO Difficulty under 35 and Volume over 100. This immediately removes the unwinnable keywords and the near-zero-volume ones. What remains is your working list of candidates. Most tools let you apply these filters directly in the interface.
5. Identify the intent and business fit for each remaining keyword. Look at the Intent column if your tool provides it. Then ask a second question: if someone visits my site through this keyword, is there a natural path to my product, service, or program? A keyword with good volume and low difficulty is still a weak choice if the audience it attracts has no reason to ever buy what you sell. Prioritize keywords where the topic connects directly to what you offer.
6. Check the actual page one results manually for your top candidates. Search your target keyword in Google. Look at what type of content is ranking: are the top five results blog posts from small sites, or product pages from major retailers, or Wikipedia? If you see forums, smaller blogs, or thin content ranking, that is a green light. Also check whether Google is serving an AI Overview at the top of the page. If it is, factor in that click-through traffic will be lower than the search volume suggests.
7. Build a master list organized by topic cluster, with primary and secondary keywords assigned. Group your shortlisted keywords by theme. Each cluster should have one main article targeting the highest-value term, supported by three to five articles on long-tail variations. Within the main article, identify two to four secondary keywords closely related to your primary term and work them naturally into your headings and body copy. Pages that rank for a primary keyword typically also pick up traffic from dozens of related terms, and deliberately incorporating secondary keywords accelerates this.

Understanding Long-Tail Keywords and Why New Sites Should Prioritize Them

A long-tail keyword is any keyword phrase that is more specific than the broad head term. “Shoes” is a head term. “Waterproof hiking shoes for wide feet” is a long-tail keyword. The long-tail version gets far fewer monthly searches, but every one of those searchers is highly specific about what they want, which makes them much more likely to engage with content that actually answers their question, and much more likely to buy if your site sells it.

For new blogs, the long tail is not just a fallback strategy. It is the only viable strategy in the first year. The data consistently shows that long-tail keywords have lower difficulty scores because fewer content creators bother to target them specifically. Yet the conversion rates are often higher precisely because the searcher’s intent is so clear. A visitor who searched “waterproof hiking shoes for wide feet” is not browsing. They are looking for a solution to a specific problem.

Notice that “keyword research seo” has 22,200 monthly searches and a difficulty of only 20. That is a genuinely exceptional mid-tail opportunity: low difficulty, reasonable volume, clear informational intent. These exist in every niche. The job of keyword research is to find them before you write a single word of content.

The Role of CPC in Keyword Evaluation

Cost per click is the average amount advertisers pay to appear at the top of Google for a given keyword. You are not running ads, so why does it matter? Because CPC is a proxy for commercial value. When advertisers are willing to pay $9, $12, or $18 per click, it means those clicks convert into sales or leads. A high CPC keyword with manageable difficulty is extremely valuable organic traffic. You are getting clicks that advertisers pay heavily for, at zero cost per click.

A keyword with 500 monthly searches, a difficulty of 25, and a CPC of $11 is more valuable to build content around than one with 2,000 searches, difficulty of 30, and a CPC of $0.80. The first audience is commercially motivated. The second may be informational browsers with low purchase intent. CPC is not the deciding factor, but it is a useful signal when you are choosing between two keywords with similar volume and difficulty.

Business Fit: Will This Audience Ever Buy From You?

The fourth metric does not come from a keyword tool. It comes from thinking clearly about your own business model. Business fit is the question of whether the audience a keyword attracts is actually capable of becoming a customer. This matters more than most beginners realize, because it is entirely possible to rank well for a high-volume keyword and generate traffic that never converts into anything because the people arriving have no reason to buy what you sell.

The way to assess business fit is to ask: if someone finds my site through this keyword, is there a natural path from that article to my product, service, or program? For a site like this one, an article about “how to choose keywords for your online business blog” has strong business fit because the person reading it is already building an online business and is therefore a natural candidate for a course that teaches exactly that. An article about “what is Google Analytics” has weaker business fit because the audience is far broader and much harder to connect to a specific offer. Both might be valid for different reasons, but business fit should influence which keywords you prioritize when time and resources are limited.

Tools You Can Use to Do This Research

Several tools make this process manageable without requiring a technical background. Each has a different strength depending on what stage of research you are in.

For most new online business owners, Ubersuggest covers 90% of what you need in the first year. As your site grows and you need deeper competitor intelligence or content optimization at scale, Semrush and Ahrefs become the natural next step. The free Google tools, Search Console and Analytics 4, should be set up from day one regardless of what else you use, because they give you real data about your own site that no third-party tool can replicate.

Keyword Scoring Calculator: Should You Target This Keyword?

Enter the metrics from your keyword tool below and get an instant recommendation on whether to target it.

Building Your First Keyword List: A Practical Framework

Once you understand the metrics, the practical question is how to organize a workable keyword list for your blog. A keyword list that is just a collection of terms is not a strategy. A keyword list organized into topic clusters, with a clear sense of which terms you are targeting first and why, is a content plan.

The structure that works best for new sites in the first year looks like this. You identify five to eight core topic areas related to your niche. Within each core area, you find one or two mid-volume keywords with difficulty under 35 that become your primary content targets: the main pillar articles. Then you find three to five long-tail variations and question-based keywords around each pillar, with difficulty under 25. These become the supporting articles that internally link back to the main piece, building topical authority over time.

This is not just an organizational preference. Search engines evaluate topical depth. A site that covers one topic thoroughly across many interlinked articles ranks better than a site that publishes one article on a topic and moves on. The cluster structure is how you signal to Google that your site is a genuine authority on a subject, even as a new domain.

Example Topic Cluster

Niche: diabetic health products. Core pillar: “blood glucose monitor for dogs” (4,400 vol, difficulty 18). Supporting articles: “how to check dog blood sugar at home” (informational, difficulty likely <20), “signs of diabetes in dogs” (informational, high intent for the niche), “best glucose meter for small dogs” (commercial, very specific), “continuous glucose monitoring for dogs” (informational/commercial). Each supporting article links back to the pillar. The pillar becomes the strongest page on this subtopic across your entire site.

The Most Common Mistakes in Keyword Research for New Sites

Most of the errors that lead to poor organic traffic come down to a few predictable patterns. Understanding them upfront saves months of wasted content production.

The first mistake is targeting keywords by volume alone. Volume is only meaningful in the context of difficulty. A keyword with 50,000 monthly searches and a difficulty of 85 is not an opportunity. It is a dead end that will cost you 10 hours of writing and return nothing. A keyword with 400 monthly searches and a difficulty of 15 is a genuine win waiting to happen. Train yourself to look at difficulty before getting excited about volume.

The second mistake is ignoring intent. Writing a product page for an informational query, or a blog post for a transactional query, results in content that ranks poorly and converts even worse. Google’s algorithm is sophisticated enough to identify the dominant intent behind a query and preference content that matches it. If you search your target keyword and find page one is full of blog posts, do not build a product page for it. Match the format the search engine has already decided the audience wants.

The third mistake is treating keyword research as a one-time event. Your first keyword list will not survive contact with reality intact. Some keywords you target will prove harder than the difficulty score suggested. Others will perform faster than expected. Google Search Console will start showing you queries you are already ranking for on pages two and three: keywords that may not have been on your original list but represent real opportunities to improve. Keyword research is an ongoing process, not a launch task.

The fourth mistake is keyword cannibalization, which is when two or more pages on your own site compete for the same keyword. This weakens both pages in Google’s eyes because the site is sending conflicting signals about which one should rank. It happens most often when a blog grows without a clear topic cluster structure and multiple articles end up covering the same ground. The fix is to plan your keyword assignments before you publish: each keyword should belong to exactly one page, and every supporting article in a cluster should target a distinct variation, not overlap with the pillar.

The new sites that build organic traffic fastest are not the ones that wrote the most content. They are the ones that chose winnable keywords, matched intent precisely, kept their topic structure clean, and treated the list as a living document rather than a launch checklist.

Frequently Asked Questions

How many keywords should I target in the first year?

Focus on quality over quantity. Thirty to fifty well-chosen keywords organized into topic clusters will outperform 200 randomly selected terms. Start with five or six topic areas and build out five to eight articles per cluster before adding new topic areas. A tightly structured site with deep topical coverage builds authority faster than a sprawling site that touches on everything superficially.

What SEO difficulty score should a new site target?

Aim for difficulty scores below 35 for the first six to twelve months. Once your site has a handful of articles ranking on page one and has started accumulating organic backlinks, you can start testing keywords in the 35 to 50 range. Difficulty scores above 50 are best avoided until your site has genuine domain authority built from earned links and consistent traffic history.

Is Ubersuggest accurate for keyword data?

Like all third-party SEO tools, Ubersuggest pulls from its own data sources and the figures will differ somewhat from Google’s internal data. Treat the numbers as directional, not precise. The relative comparisons between keywords are reliable and useful: this one has lower difficulty than that one, this one has higher commercial value. The absolute volume figures are best understood as estimates within a margin of roughly 20 to 30 percent.

Can I rank for a keyword without backlinks?

Yes, especially for low-difficulty keywords. Backlinks are a major ranking factor, but on terms with difficulty scores below 20, the sites currently ranking often have very few backlinks themselves. Well-structured, thorough, intent-matched content on a technically sound site can rank for low-difficulty keywords within weeks of publication. As difficulty increases, the backlink requirement increases with it.

How do I know if a keyword is worth writing about if the volume shows zero?

Zero volume in a keyword tool does not always mean zero searches. It can mean the tool’s data sample is too small to detect it. Before discarding a zero-volume keyword, check whether Google Autocomplete suggests it, look at whether forums or Reddit threads about the topic exist, and check whether any pages are already ranking for it. If you find content that targets the term and is getting engagement, there is real demand even if the tool cannot quantify it. Treat these as optional additions to a cluster, not primary targets.

Where Keyword Research Fits in the Bigger Picture

Keyword research is where every successful online business blog begins, but it is not where the work ends. Knowing which keywords to target gets you in front of the right audience. What you do with that audience once they arrive, how you structure your site, build your offer, and create systems that turn traffic into revenue, is the larger challenge. If you are building an online business from scratch and want a structured path through every stage of that process, from niche selection and content strategy to unit economics and scaling, the Business to Passive Income program was built for exactly that.