Skip to main content

Keyword research for SaaS

Keyword research is the most important work you do in Google Ads. The keywords you pick determine which queries enter your ads' auctions. Good keywords send your budget toward people likely to buy. Bad keywords send your budget toward people who will click but never convert.

This article covers how to think about keywords for B2B SaaS specifically.

Start with intent, not volume​

The instinct is to find keywords with high search volume. That instinct is wrong. Volume matters, but it is not enough on its own. What matters is intent: is this searcher likely to fit your product right now?

Five buckets of intent, in roughly the order they convert:

1. Branded competitor terms​

People who search for a specific competitor by name. "Salesforce alternative", "alternatives to Asana", "Mailchimp competitors".

These convert exceptionally well. The searcher is dissatisfied with a known competitor and looks for alternatives. In effect, they ask to switch.

The catch: not all competitor terms are biddable. Some categories restrict bidding on trademarked names. Check Google's policies and the competitor's trademark stance. Do not assume you can bid on the name.

2. Switching intent terms​

Variants of the terms above that do not name a specific competitor. "Switching CRMs", "best Asana alternative", "leaving HubSpot".

Slightly less specific, but still high intent. The searcher is actively looking to buy.

3. Specific category + qualifier terms​

Terms that narrow your category to a specific use case. "CRM for sales engineering teams", "project management for consulting", "email marketing for ecommerce".

These match buyers who know what category they want and have a specific use case. They convert well if your ad and landing page address that use case.

4. Broad category terms​

The category name without a qualifier. "CRM software", "project management tool", "email marketing platform".

Highest volume, highest competition, mixed intent. Some searchers are buyers. Many are researchers, students, or people who are just browsing. These terms can convert, but each conversion costs more.

5. Top of funnel research terms​

"What is a CRM", "how to manage a sales pipeline", "introduction to email marketing".

Almost never convert directly. The searcher wants to learn, not buy. Target these terms if you have content that captures email addresses to nurture leads. Skip them if your only goal is a direct conversion.

Avoid these patterns​

Your own brand name​

If you are already known, people who search for your brand will find you organically. When you bid on your brand, you pay for clicks you would get for free anyway. The exception is when a competitor bids on your brand name. Then bid on it too, to defend it.

Generic adjectives​

"Best", "top", "free". These add nothing. "Best CRM" is just "CRM" with a meaningless qualifier.

The exception: "free X" can be a useful intent signal if you have a free tier. Otherwise, skip it.

Job titles​

"Account executive", "marketing manager". Job titles are not search queries. People who search their own job title probably look for a job, not your product.

Industry names​

"SaaS", "B2B", "fintech". Too broad, too top of funnel. Useful as filters or categories, not as keywords.

How to find keywords​

A few approaches, in order of effectiveness:

Mine your own data​

If you have any organic traffic, your search console (Google Search Console) shows what queries bring people to your site. Those are real intent signals. Some will be good keywords for paid ads. Many will not.

Look at competitor ads​

Type your competitor brand name into Google. Look at the ads that show up, especially ads from other companies that bid on your competitor's brand. Their keywords are public. Their ad copy gives you ideas.

Use Google's keyword planner​

Google Ads has a built-in keyword planner. Type a seed phrase. Get keyword suggestions. Volume estimates are noisy, but they give a rough guide.

Use AI-driven keyword tools​

Hero Marketer's campaign wizard generates keyword clusters from a single seed plus your product context. It pulls from Google's keyword data and clusters by intent. It is faster than manual research, and it stays grounded in what your product is. See Choose keyword clusters.

Talk to customers​

Especially newer customers. Ask: "When you were searching for a tool like ours, what did you type into Google?" The answers are often surprising and rarely match what marketing thinks customers search.

Match types matter​

Google Ads has three keyword match types:

  • Exact match. Only triggers on queries very close to your keyword. Tightest control.
  • Phrase match. Triggers on queries that contain your phrase plus other words. Looser.
  • Broad match. Triggers on Google's interpretation of related queries. Loosest, most volume, most risk.

For most B2B SaaS, start with phrase match. It captures more queries than exact match, while it still respects intent. Move to broad match once you have a strong negative keyword list and conversion data to optimize against.

Hero Marketer creates campaigns with sensible default match types. Ask Hero AI to add or remove keywords with a different match type once a campaign is live.

Negative keywords are half the work​

A keyword tells Google what to bid on. A negative keyword tells Google what never to bid on, regardless of other matches.

Negatives are how you cut wasted spend. Common B2B SaaS negatives:

  • "Free" (if you do not have a free tier, or if you do not want people who only look for free options).
  • "Job", "career", "salary" (people who research jobs, not tools to buy).
  • "Definition", "what is", "examples", "tutorial" (research, not a purchase).
  • "Reddit", "YouTube" (people who look for opinions or content, not products).
  • Brand names of products with completely different use cases that share keywords.

Run your campaigns for a week. Pull the search terms report from Google Ads. Add anything irrelevant as a negative keyword. Repeat weekly for the first month.

How many keywords per campaign​

Tight clusters of 10 to 30 keywords work well. Hundreds of keywords in one campaign usually mean the cluster is too broad.

If you are tempted to throw 200 keywords into one campaign, the right move is usually 4 campaigns of 50 keywords each, organized by intent. See Building keyword clusters.

A starter list for a B2B SaaS​

If you start from zero, build campaigns around these intent buckets:

  1. Alternative to [biggest competitor]. One campaign, often the highest converter.
  2. [Category] for [your specific customer type]. "CRM for sales engineering", "project management for consulting".
  3. [Specific job your product does]. "Track engineering work", "automate email outreach".
  4. [Pain point your product solves]. "Reduce manual data entry", "speed up sales handoff".

Four campaigns, each with one tight cluster, get you quickly to the data that tells you what works.

Next​