# What's the Best Way to Do Keyword Research for a SaaS Company?

> What's the best way to do keyword research for a SaaS company? Start with buyer-intent queries, not volume. Here's the full 2026 process, step by step.

Published: 2026-09-08

What's the best way to do keyword research for a SaaS company? Start with the queries that signal a readiness to buy: comparisons, alternatives, pricing questions, integrations, and reviews. That's the short answer. If you do keyword research the way most generic tutorials describe it, meaning you type a broad term into a tool, sort by search volume, and start writing about whatever shows 5,000 monthly searches, you'll build a blog that gets traffic and never signs up a single customer.

Run properly, SaaS keyword research in 2026 follows a specific sequence:

1. Start with the bottom-funnel page types closest to a purchase decision
2. Mine the language customers actually use (sales calls, support tickets, Reddit, review sites)
3. Score keywords by intent and business value before volume
4. Map one keyword to exactly one page
5. Run a competitor gap analysis
6. Check the live SERP before committing to anything
7. Add an AI-search layer, because ChatGPT and Perplexity now send buyers directly

Aim for 20 to 50 well-researched core keywords to start, not 500 mediocre ones. That's the range experienced SaaS SEO practitioners keep converging on, and it matches what I've seen work in practice. The details are where teams win or lose, so let's walk through each step properly.

## What's the best way to do keyword research for a SaaS company, step by step

### Why most SaaS keyword research fails before it starts

Most SaaS keyword research fails for one unglamorous reason: teams treat it as a traffic exercise instead of a revenue exercise. They chase terms like "project management tips" because the volume looks juicy, then wonder why 40,000 monthly visitors produce 12 trials.

The problem is structural. Third-party keyword tools (Ahrefs, Semrush, and friends) estimate volume from clickstream data and databases that are thinnest exactly where SaaS money lives. Long-tail, high-intent queries like "[competitor] for [specific use case]" often show zero or tiny volume in the tools. Buyers still type them. Google still serves results. And when someone searches "Asana vs Monday for small agencies," that person is weeks away from swiping a credit card.

So rule one: stop letting search volume be the deciding factor. Volume is a tiebreaker, not a strategy.

### Step 1: Start at the bottom of the funnel with the SaaS page-type map

Most advice on SaaS keyword research jumps straight to tools. Start one layer earlier: reverse-engineer the page types that sit closest to a purchase. Before you touch any keyword tool, list the queries your product should own. Six types cover nearly everything.

Comparison pages come first. "[Your product] vs [Competitor]," and the reverse. Five real competitors means ten pages right there, and every one of them targets a searcher who has already narrowed the field and is choosing.

Alternative pages convert even harder. Someone typing "[Competitor] alternatives" or "[Competitor] alternatives for [use case]" is dissatisfied with something else and actively shopping. In my experience these are the cheapest trials in SaaS SEO.

Pricing and cost queries deserve their own pages too: "[category] pricing," "how much does [category] cost," "[category] for [budget]." Even with a freemium model, own the pricing query. The people asking it are doing budget math right now.

Round the map out with best-of and category pages ("best [category] software," "best [category] tools for [industry]"), integration pages ("[Your product] [other tool] integration," which are underrated and often zero-competition), and review or trust pages ("[Your product] reviews," "is [your product] legit," "[your product] demo").

Work through your top five to ten competitors and every adjacent category, and you'll have a concrete keyword list before you've opened a single tool. This page-type map is the foundation of the whole process, and it's the single highest-leverage change most teams can make. Most start at the top of the funnel and spend eighteen months earning traffic that never converts. Flip it around.

One caveat, and I'll be blunt about it: if your product genuinely has no close competitors or your category has no search demand yet, bottom-funnel research alone won't carry you. You'll need problem-aware content. But for the majority of SaaS companies, competitors exist and buyers search for them by name. Mine that.

### Step 2: Mine buyer language, not feature names

Founders describe their product the way they built it. Customers describe it by the job they're trying to do.

A scheduling tool gets marketed as "AI-powered resource orchestration." Nobody has ever typed that into Google. The customer types "how to schedule shifts for a small team" or "rota software for restaurants." The gap between your feature vocabulary and your customer's job-to-be-done vocabulary is where keyword opportunities hide.

Where to find the real language:

- **Sales and support transcripts.** Tools like Gong or even plain Zoom recordings. Search transcripts for the verbs customers use. What did they call the problem before they knew your product existed?
- **Your existing Google Search Console data.** If you have any history, export every query where you appeared. You'll find terms you're ranking 8 to 15 for with no dedicated page. Cheap wins.
- **Reddit, Quora, and niche communities.** Search "site:reddit.com [category keyword]" and read what people actually ask. Threads titled "looking for a tool that..." are essentially pre-written keyword lists, and the phrasing is verbatim customer language.
- **Review sites.** Capterra and G2 reviews for your category and competitors. People describe what they replaced and why. Those replacement stories map directly to alternative keywords.
- **Your onboarding flow.** The fields you ask people to fill in during signup tell you exactly how they self-describe their use case. Cluster those answers and you have your use-case keyword map.

Ask any keyword tool for its opinion and it will hand you a volume list. This mining step is what separates real SaaS research from the generic e-commerce or blogging variety. A keyword tool can't see inside your support tickets. You can.

### Step 3: Score intent and business value before volume

Once you have a raw list, you need a scoring system. Score intent first. It's the heart of the whole exercise, because most of the list eliminates itself when you get it right. Here's one I'd actually recommend, kept simple enough that a team will use it:

Give every keyword a business value score from 1 to 3:

- **3 (buyer-ready):** comparison, alternative, pricing, integration, review queries. The searcher knows the category, probably knows competitors, and is choosing.
- **2 (solution-aware):** "best [category] tools," "how to [job] with software," category educational content where software is the obvious solution.
- **1 (problem-aware):** "why is my team always behind on deadlines" type queries. Top of funnel. Valuable in bulk, useless in small quantities.

Then apply a rule: a keyword only earns a spot if its business value is 3, or if it's a 2 with meaningful volume and clear winnability. Problem-aware 1s only make the cut when they cluster into a genuine topic authority play or the SERP is trivially weak.

Notice volume never appears in the scoring. That's deliberate. A term with 70 monthly searches and an intent score of 3 will often outperform a term with 5,000 searches and an intent score of 1, both in conversion rate and in total pipeline. I've seen a single "[competitor] alternative" page outperform an entire 50-article top-of-funnel blog in trial signups. It's not even close.

This is also how the sharper SaaS SEO agencies answer what's the best way to do keyword research for a SaaS company in 2026: traffic potential and business value over raw volume, with intent and expected conversion as the primary filters. If you remember nothing else from this article, remember the funnel-first scoring.

### Step 4: Check the actual SERP before you commit

Here's a step most teams skip, and it costs them months.

Before you assign any keyword to a page, search it yourself. In an incognito window, from a relevant location. Look at what's actually there.

You're checking for three things:

1. **Is the SERP winnable?** If every result is a DR 90 giant (Salesforce, Gartner, Forbes) and you're a DR 20 startup, this keyword is a two-year fight, not a three-month win. Thin forum results, dated content, or niche blogs ranking? That's your green light. The filter you're applying is winnability: queries where your site can actually rank, not just queries that happen to be popular.
2. **Does the SERP still have clicks?** Since 2024, AI Overviews have absorbed clicks on a lot of informational queries. If the top ten results are buried under a sprawling AI answer, the informational keyword may deliver a fraction of its stated volume. Commercial queries hold up better, because Google and the AI engines still funnel buyers to vendor pages for comparison and pricing.
3. **What format is Google rewarding?** If every ranking page is a listicle, don't write an essay. If they're all landing pages, don't write a blog post. Format mismatch is one of the quietest reasons pages fail.

This live-SERP check matters more in 2026, not less. Enriched results and AI Overviews change whether a term is worth targeting at all, which is why every process worth following includes this manual check instead of trusting tool difficulty scores alone.

### Step 5: Map one keyword to one page

Keyword cannibalization quietly kills SaaS sites. You write a comparison post, a blog post about alternatives, and a "best of" roundup, and all three target "best CRM for startups." Google can't figure out which one to rank, so it half-ranks all of them, usually below where any single one would have landed.

The fix is a strict one-keyword-per-page rule. Build a simple spreadsheet: every keyword gets exactly one URL, one primary target, and one assigned funnel stage. If two keywords are so similar you can't justify separate pages, merge them. If they're genuinely different intents ("best email marketing software" vs "Mailchimp alternatives"), they get separate pages with different angles.

Then map the whole list to your funnel:

| Funnel stage | Page type | Example keywords |
|---|---|---|
| Decision (bottom) | Comparison, alternative, pricing, integration pages | "[competitor] vs [your product]", "[competitor] alternatives for [use case]" |
| Consideration (middle) | Best-of lists, category explainers, "software for X" pages | "best [category] for [industry]", "how to [job] with [category]" |
| Awareness (top) | Problem-solving guides, frameworks, templates | "why [problem happens]", "[problem] checklist" |

A sensible 2026 allocation for an early or growth-stage SaaS: roughly half your effort on bottom-funnel, a third on middle, and the rest on top. If your allocation is inverted, that's a red flag. This mapping is also the moment your research turns into an actual content plan rather than a list.

### Step 6: Run a competitor gap analysis

By this point you have a solid list built from intent. Now expand it systematically.

Export your competitors' organic keywords from Ahrefs or Semrush (the lightweight versions are fine; Ahrefs starts around $29/month and Semrush around $139/month as of 2026, with higher tiers reaching roughly $1,499 and $549 respectively, so start small). Filter for:

- Keywords where a competitor ranks in the top 10 and you don't rank at all
- Keywords with buyer-intent modifiers you haven't covered
- Pages of theirs that rank for several related terms (these reveal their clusters, which reveal their strategy)

Anything that survives your intent and SERP checks gets appended to your spreadsheet. Gap analysis is the least glamorous stretch of the process and one of the highest-yield: exporting competitor keywords, finding the pages you don't rank for, and mapping the gaps to new or updated content is still standard practice at the best SaaS SEO agencies, and for good reason. Your competitors have already done years of SERP testing for you. Their rankings tell you what Google rewards in your exact niche.

### Step 7: Add the AI-search layer

This is the part older keyword research guides completely miss. No 2026 research process is complete without this layer.

A growing share of SaaS buyers ask ChatGPT, Perplexity, or Gemini things like "what's the best tool for [job]" or "what's a good [category] alternative to [competitor]." If your product doesn't appear in those answers, you're invisible in a buying conversation that happens without a click on your site at all.

Practical ways to win AI visibility:

- **Get into the sources the AI engines cite.** For most category queries, that means comparison articles, listicles, and review sites. Pursue inclusion in "best [category] tools" roundups and on the major review platforms. Third-party mentions now function like backlinks did in 2015.
- **Publish content that directly answers buyer questions in plain language.** AI models favor content that states a clear answer up front, with concrete criteria and numbers, over vague marketing fluff. Ironically, that's also what ranks in Google's AI Overviews and gets cited by ChatGPT. Write so a model can quote you.
- **Check the AI answers manually.** Once a month, ask ChatGPT and Perplexity the queries that matter to your category. Note who gets mentioned and who doesn't. That's your competitive intelligence now.

If you want the bigger picture, our guide to [the best SEO strategies for SaaS companies in 2026](https://www.tryspook.com/blog/what-are-the-best-seo-strategies-for-saas-companies-in-2026) covers where AI citations fit alongside technical SEO and link building.

## How many keywords should a SaaS company target?

Direct answer: 20 to 50 well-researched core keywords is the right starting set for most SaaS companies, and 20 to 30 is plenty for an early-stage one. Ask ten practitioners how they do it and they'll disagree on plenty; on this point they won't. Smaller beats bigger, every time.

Why? A list of 500 keywords produces decision paralysis and shallow coverage. A list of 30 lets you build complete pages, cluster related terms under each, and actually finish things. Every core keyword can also carry 3 to 10 supporting long-tail variations, so your real coverage is bigger than the core number suggests.

Cluster around topics, not just individual terms. If you're a scheduling SaaS, "shift scheduling software," "employee rota software," and "shift planning app" probably belong to one cluster served by one strong page, plus supporting content. This clustering is how you build topical authority, which Google's systems reward with rankings across the whole topic, not just one query.

## The tool stack, honestly

You don't need much tooling for any of this, and you definitely don't need the $1,499/month tier of anything.

![The tool stack, honestly](https://wd9kbpy1zu.ufs.sh/f/zfTUaHBShJ7Cg0zpjZ6tGPMQRY4JKExOX0o5VDUIlbeq2j3B)

- **Ahrefs or Semrush** (or SE Ranking at about $129/month as a budget pick): competitor keyword exports, SERP data, gap analysis. One is enough.
- **Google Search Console:** free, and the only source of data about queries you already touch. Underrated for keyword research because it shows real impressions and click-through rates, not estimates.
- **Google autocomplete and People Also Ask:** free lists of real queries people type. Start typing "[category] for" and watch what appears. Crude, but genuine.
- **Reddit, Quora, review sites:** free buyer-language mining as described above.
- **A scoring spreadsheet:** free, and the most important tool on this list.

And if all of this sounds like a lot of manual work, that's because it is. It's the same workflow [Spook](https://tryspook.com/) automates: identifying winnable queries for your site, generating SEO-optimized articles in your brand voice, and publishing them, with a built-in backlink network behind it. For a founder or a two-person marketing team, trading several hours of weekly research for automated output is a legitimate choice, and our comparison of [the best alternative to manual content marketing](https://www.tryspook.com/blog/what-is-the-best-alternative-to-manual-content-marketing) lays out the trade-offs on both sides.

## Prioritize pipeline, not rankings

One more mindset shift, and then I'll give you the weekly routine. The whole point of keyword research is pipeline, not positions.

A keyword that takes you from position 15 to position 3 on a buyer-intent term might be worth more than 200 informational rankings. Track which pages generate demo requests and trial signups, in whatever analytics setup you have, and let that feedback loop reshape your keyword priorities every quarter.

Some SaaS teams even tie keyword value to revenue math. If you know your average deal size and close rate, you can estimate what a page-one ranking on a given commercial term is worth per month, and that number makes prioritization arguments in your team disappear. Resources like [Targetmrr](https://www.targetmrr.com) exist specifically to help SaaS companies connect growth metrics like these to MRR, which is the only number your CFO cares about anyway.

Rankings feel good. MRR pays salaries. Optimize accordingly.

## How often should you redo keyword research?

Direct answer: do a quarterly refresh with monthly light check-ins. Competitor rankings shift, Google updates SERP features, and AI-search behavior changes fast enough that a keyword list written in January can be stale by July. That's why you should treat this work as a maintenance ritual rather than a one-off project.

The quarterly refresh looks like: re-run your competitor gap export, re-check SERPs for your 20 most valuable keywords (are new competitors ranking? did an AI Overview appear?), add and retire keywords, and re-score anything whose intent changed. The monthly check-in is lighter: skim Search Console for new queries you're earning impressions for, and spot-check the AI answers for your category.

There's no finish line here. It's maintenance, like flossing, only with spreadsheets.

## A concrete 4-week plan you can run starting Monday

If you want this as a schedule rather than a philosophy:

![A concrete 4-week plan you can run starting Monday](https://wd9kbpy1zu.ufs.sh/f/zfTUaHBShJ7CJggFhJPcwhK1gi6EpOJDArW0mHl5z8UCsn2e)

**Week 1: Build the raw list.** Write out your competitor set (aim for 5 to 10 names). List every comparison, alternative, pricing, integration, and review keyword from the page-type map. Mine sales calls, support tickets, Reddit, and your Search Console exports for buyer-language keywords. Target: 100 to 150 raw terms.

**Week 2: Score and prune.** Apply the 1-to-3 intent scoring, drop everything that scores 1 unless the SERP looks trivially weak. Run live SERP checks on the survivors. Target: 30 to 60 keywords left.

**Week 3: Map and cluster.** Assign one keyword per page, cluster supporting terms, and mark funnel stage for each. Export competitor keywords and fold in gaps that survive your checks. Target: a final roadmap of 20 to 50 core keywords with a publishing order that starts bottom-funnel.

**Week 4: Start executing.** Write the two or three highest-value pages first. Almost always, that's a competitor comparison and a top alternatives page. Publish, measure, and iterate quarterly from there.

If you're a brand-new site with no authority, temper expectations: even winnable keywords take weeks to months to rank. The honest answer for a new domain in 2026 is that long-tail, low-competition buyer terms are where it earns its first traction. Which is convenient, because that's also where your keyword research should start anyway.

## FAQ: SaaS keyword research questions people actually ask

**What is the best way to do keyword research for a SaaS company?** Start with bottom-funnel buyer queries like comparisons, alternatives, pricing, and integrations, mine customer language from sales calls and Reddit, score keywords by intent and business value before volume, map one keyword to one page, and check the live SERP before committing. Aim for 20 to 50 core keywords.

**What is the best way to do keyword research generally?** The core method is: list seed topics from your business, expand them using tools and autocomplete, check the SERP for each candidate, score by intent and difficulty, and map keywords to pages. For SaaS specifically, weight buyer-intent page types far more heavily than generic tutorials suggest.

**What is the 80/20 rule in SEO?** Roughly 80% of your results come from 20% of your pages and keywords. In SaaS terms: a handful of bottom-funnel pages will generate most of your pipeline, so protect and improve those before writing anything new. Check which 20% of your pages convert and double down there first.

**Is SEO still worth it for SaaS in 2026?** Yes, but the shape changed. AI Overviews and chat assistants have eaten clicks on generic informational queries, while commercial and comparison queries still send buyers to vendor pages. SEO for SaaS is more valuable and more selective than it used to be: fewer queries are worth targeting, and the ones worth targeting are worth more.

**Is SEO dead now with AI?** No. Search behavior moved, it didn't vanish. Buyers still search, and AI assistants cite sources, which creates a new visibility surface (often called GEO, generative engine optimization). The companies losing out are the ones optimizing for 2019 search only.

**Can I do SEO and keyword research myself?** Yes, with the process in this article and one affordable tool. The honest trade-off: doing it well takes several hours a week on research, writing, and maintenance. That's a fine use of a founder's time at the very start, and a poor use of it once you have paying customers. Tools that automate winnable-keyword identification and content production, Spook among them, exist for exactly the moment when your time gets expensive.

## The bottom line

Plenty of teams assume keyword research is a tooling problem. It isn't. Any tool will hand you the same 5,000-row export; the advantage lives in the judgment layers, deciding what counts as buyer intent in your specific category, spotting a SERP you can actually win, knowing which customer words beat your own feature vocabulary. That judgment is the real answer to what's the best way to do keyword research for a SaaS company, and it's why two competitors running identical processes can end up with completely different pipelines.

So skip the giant list. This week, write out your top five competitors and draft the comparison and alternatives pages for each. That's ten keywords, all buyer-intent, winnable at almost any domain strength, and they'll produce more pipeline than your next fifty blog posts. Spook can find the winnable queries and write those pages for you if you'd rather skip the spreadsheet work. Either way, start at the bottom of the funnel and work up.

**Watch:** [How to do Keyword Research for SaaS](https://www.youtube.com/watch?v=sCy7jsuBRAs)

## FAQ

### What is the best way to do keyword research for a SaaS company?

Start with bottom-funnel buyer queries like comparisons, alternatives, pricing, and integrations, mine customer language from sales calls and Reddit, score keywords by intent and business value before volume, map one keyword to one page, and check the live SERP before committing. Aim for 20-50 core keywords.

### How many keywords should a SaaS company target?

20-50 well-researched core keywords is the right starting set, and 20-30 is plenty for an early-stage SaaS. Each core keyword can carry 3-10 supporting long-tail variations clustered under one page.

### What is the 80/20 rule in SEO?

Roughly 80% of results come from 20% of pages and keywords. For SaaS, a handful of bottom-funnel pages generate most pipeline, so identify and improve those before publishing new content.

### Is SEO still worth it for SaaS in 2026?

Yes. AI Overviews have absorbed clicks on generic informational queries, but commercial, comparison, and pricing queries still drive buyers to vendor pages, making selective, intent-focused SEO more valuable than ever for SaaS.

Read on the site: https://www.tryspook.com/blog/what-s-the-best-way-to-do-keyword-research-for-a-saas-company