A working keyword research checklist that addresses discovery, qualification, intent mapping, and competitive feasibility—without relying on a single tool or arbitrary volume threshold. Built for practitioners who need a repeatable process that adapts to Canadian bilingual markets and local versus national contexts.
Begin with seed terms from three sources: actual customer language captured in sales calls or support tickets, your own service or product taxonomy, and competitor title tags or H1s scraped from their primary landing pages. Avoid starting in a keyword tool—you want unfiltered human language first. From these seeds, expand using synonym and question modifiers. For a law firm, a seed like estate planning becomes estate planning lawyer, will and estate lawyer, estate attorney near me, how to plan an estate in Ontario. If you operate bilingually, generate French parallels immediately—planification successorale avocat, testament et succession avocat—because Google treats these as separate keyword universes with different competition and volume distributions. Use a spreadsheet with columns for seed term, expanded variant, language, and estimated intent. Aim for 100 to 300 raw candidates before you touch a volume tool. The goal here is breadth and capturing the full vocabulary map, not filtering yet.
Now pull search volume data, but treat the numbers as directional ranges rather than gospel. Tools report blended averages and often merge close variants, so a term showing 880 monthly searches might include plurals, prepositions, and acronyms. For Canadian contexts, check if the volume is national aggregate or splits by province—a keyword strong in Ontario may be invisible in Alberta. Look at twelve-month trend lines to spot seasonal spikes or secular declines. A term spiking every December is useful for an ecommerce client but misleading for evergreen content planning. Export volume alongside the keyword list and flag any term under 50 monthly searches unless it is highly specific and high-intent transactional. Discard obvious branded competitor terms and misspellings unless you have a deliberate parasite strategy. At this stage you should cut your list by thirty to fifty percent, keeping terms with demonstrable demand and stable or growing interest over time.
Classify each surviving keyword by search intent: navigational if it targets a known brand or location, informational if it seeks an answer or explanation, commercial investigation if it compares options or seeks reviews, transactional if it signals immediate purchase or signup. This determines what content format can rank. Search the keyword yourself in an incognito window and note which SERP features appear—local pack, featured snippet, people also ask, video carousel, shopping ads. If the first page is dominated by a local three-pack and your client lacks Google Business Profile optimization, ranking organically is secondary. If featured snippets occupy position zero, your content needs structured answers in short paragraphs or lists. For Canadian queries, toggle between google.ca in English and google.ca with French language settings if applicable, because SERP layouts and featured content can differ. Record intent and SERP features in adjacent columns. A keyword that looks high-volume but returns only huge marketplace domains and shopping carousels may not be worth organic content investment.
Evaluate the top ten organic results for each keyword against your own domain metrics. Check the domain authority or rating of ranking URLs, their backlink counts, content length, and publish or update dates. If every ranking page has authority in the sixty-plus range, hundreds of backlinks, and your domain sits at twenty, you face a multi-month uphill battle. That does not mean skip the keyword, but it does mean deprioritize it or plan a long-tail cluster approach first. Look for gaps: keywords where the top results are thin, outdated, or off-topic. A query like keyword research steps might return generic marketing guides that never address sequencing—you can win with a process-focused article. For local Canadian terms, assess whether competitors have location pages, local backlinks from chambers of commerce or news sites, and whether they mention city or province names in title tags. Export a feasibility score—low, medium, high—based on this manual review. Avoid purely automated difficulty scores; they miss nuance like content relevance and SERP feature saturation.
Combine intent, volume, trend, and feasibility into a priority matrix. High-intent transactional keywords with medium competition and stable volume go to the top, even if absolute search numbers are modest. Informational keywords with strong volume but saturated SERPs get deferred or clustered under a hub page strategy. Group related keywords into topic clusters where one pillar page targets a broad head term and supporting pages address long-tail variations. For example, a pillar on keyword research process could support cluster pages on keyword research tools Canada, keyword research for local SEO, keyword research for ecommerce. Link these internally with descriptive anchor text. Canadian businesses operating in both languages should create parallel clusters—one English, one French—rather than machine-translating content, because user behavior and competition differ by language. Assign each cluster a target publish date and owner. This transforms a flat keyword list into an executable content roadmap.
Once initial content is live, cross-reference your keyword targets against actual queries triggering impressions in Google Search Console. You will discover semantically related queries you missed, question variants people actually type, and branded combinations if you rank for your own name. Filter by impressions over one hundred and average position worse than twenty—these are near-miss opportunities. If a keyword you never targeted is delivering consistent impressions but low clicks, it may warrant a dedicated page or a content refresh to capture that traffic. For Canadian clients, segment data by country if you serve international audiences, because query language and intent shift by region. Revisit your checklist quarterly, retiring keywords that decline in volume or become dominated by SERP features you cannot compete in, and adding newly discovered terms from Search Console or competitor content gap analysis. Keyword research is not a one-time audit—it is a feedback loop that improves as you publish and measure.
Intent classification is the most critical step because it determines whether a keyword can realistically drive the outcome you need. High search volume is irrelevant if the intent is informational and your goal is lead generation, or if the SERP is locked by local pack results and you lack a Google Business Profile. Correctly mapping intent ensures you build content that matches what the searcher expects and what Google rewards.
Treat English and French as separate keyword universes from the start. Generate seed terms in both languages, pull volume data for each, and analyze SERPs independently because competition, SERP features, and user behavior differ. Avoid machine translation—hire a native French speaker or use a qualified translator to ensure natural phrasing. Build parallel content clusters and interlink the language versions with hreflang tags if serving both on the same domain.
Yes, if the intent is highly specific and transactional, or if the keyword represents a gap in competitor coverage. A term with thirty monthly searches that converts at fifteen percent is more valuable than a term with three thousand searches that bounces at eighty percent. Low-volume keywords also tend to have lower competition, allowing faster wins. Aggregate multiple low-volume long-tail terms into a single comprehensive page rather than creating thin content for each.
Quarterly is a practical baseline for most businesses, aligned with Search Console data accumulation and competitor content cycles. After major algorithm updates or if your market experiences rapid change—new competitors, product launches, regulatory shifts—conduct an interim review. Keyword research is iterative; each content publish and ranking result feeds back into the next round of discovery and prioritization.
At minimum, Google Search Console for actual query data, a volume and difficulty tool that reports Canadian data specifically—some tools default to US volumes—and a spreadsheet for manual intent tagging and prioritization. If budget allows, add a backlink checker to assess competitor link profiles and a rank tracker set to Canadian geolocations. No single tool does everything well; the checklist is the process, tools are inputs.
Compare your domain authority and backlink count to the top ten ranking pages. If the median ranking domain is thirty points higher and has five times your backlink count, direct competition is inefficient. Instead, target long-tail variations of that keyword, build topical authority through a content cluster, earn links, then revisit the head term in six to twelve months. Competition is not binary—it is a resource allocation decision.