Chapter05A long row of cars parked at dusk outside an office building, each behind a small reserved sign, ending in two empty visitor spaces lit by a lamp post
Google Ads

Google Ads for SaaS · Chapter 05 of 16 · All chapters

Most Software Searches We Pulled Name a Brand. Build Your Keyword List From the Ones That Don't.

Brand names made up 58% to 84% of the software searches we pulled. Here is how to build a keyword list from the rest, and what to cut before launch.

David SmaniaFounder, BrandRocket9 min read · October 8, 2026

Pull every search in your software category out of Keyword Planner and sort it by volume, and the top of the list won't be the category at all. In project management, the biggest searches are "smartsheet," "asana" and "trello." In CRM, they're "salesforce" and "hubspot." In payroll, "adp" and "gusto." The category's own name shows up further down, behind a wall of company names.

That's the first thing a software keyword list teaches you: most of the volume belongs to somebody else. The second is that a surprising share of what's left isn't about buying software at all. This chapter walks through building a keyword list for a software account the way we build them, using Keyword Planner's US data for four categories, CRM, project management, payroll and scheduling, and the twelve software companies from earlier chapters as seed sites. Your category's numbers will differ. The steps don't.

Chapter 5 of 16

Brand Names Made Up 58% to 84% of the Software Searches We Pulled. They Belong in Other Campaigns.

We pulled Keyword Planner ideas for each category from seed keywords, from two leading vendors' sites and from a list of the leading brands' own names, collapsed the close variants that Keyword Planner counts more than once, and classified every search by what it was asking for. Brand and navigational searches, people typing a company's name, made up this share of the in-category searches we pulled:

All four categories, same story. The share runs high partly because we pulled the leading brands' names on purpose, and partly because some brand names are also ordinary words ("monday," "gusto," "basecamp"), and it describes the searches we pulled, not every search on Google. Still, no reading of those numbers puts the category's buying searches anywhere near the majority.

Brand searches don't disappear from your account. They just don't belong in this list:

In software, most of the parking lot is reserved. Build your keyword list from the spaces nobody owns.

Google's Ad Preview Tool Shows Who Bids on Your Category Without Paying a Competitor for the Click.

First, learn which software companies already bid on your category. Inside your account, under Tools, the Ad preview and diagnosis tool shows how your ad would appear on a results page for a search you type. In our experience, that preview also shows the other ads competing for the search. Run your category's main buying searches through it and note which software companies appear.

Don't do this research by searching on Google and clicking the ads you find. Every click costs that company money, and it tells you nothing the preview doesn't. Google's Ads Transparency Center fills in the rest: type a competitor's domain and it lists the ads they've run, with dates. When we checked 66 software companies there, 58 had Search ads for their brand running in the previous 90 days. In most software categories, assume your competitors are already in the auction.

Keyword Planner Seeded With a Software Site Hands Back What the Site Is About, Not What Buyers Search.

Next, give Keyword Planner each competitor's website address, set the location to the US (or the countries you sell to), and let it suggest searches. Then run your own address through it, which shows you how Google reads your business. That second pull is often the more useful one. If your own site comes back full of searches you'd never pay for, such as blog topics, careers or a free tool's audience, that is what Google thinks you're about, and it will shape everything from your ad relevance to where automated campaigns send traffic later. Better to find out from a free report than from a month of wasted clicks.

Read the results carefully, because Keyword Planner reflects the site back at you, content and all. When we seeded it with our twelve sample sites:

None of those lists is wrong, exactly. Each one shows what the site's own pages signal. But none of them is the list of searches a buyer types when shopping for that kind of software. Seed from several competitors, keep the buying searches, and treat each site's quirks as noise. Three or four seed sites usually give you a list long enough to work with and varied enough that no single site's habits dominate it.

One more quirk: seeding from a site leaves out the site's own main brand name. A pull from asana.com doesn't return "asana." For your keyword list, that's a feature.

Stock, Job and Certification Searches Were 14% of the CRM Searches We Pulled. Salesforce's Ticker Is "CRM."

Here's a trap that only software has. Salesforce trades on the New York Stock Exchange under the ticker CRM. So a CRM keyword list fills up with investors: "crmstock" alone draws about 201,000 searches a month, "salesforcestock" about 135,000, "nyse crm" about 40,500. Together with job and certification searches, that noise was 14% of the in-category CRM searches we pulled.

Every software category has its own version. Before anything goes live, build a negative keyword list for:

The list isn't finished at launch. Once the Exploration campaign switches on broad match, new junk turns up every day, and someone has to read the search terms report daily for the first month after it launches and keep adding negatives.

Every software keyword list has a few investors, a few job hunters and a few people who just want the free template. Pay for none of them.

Five or Six Software Searches That Ask the Same Question Make an Ad Group. Fifty Make a Junk Drawer.

A keyword list isn't finished when it's long. It's finished when every search on it could become a customer.

What survives the cuts is your working list: category buying searches, industry and use-case searches, and the problem searches that sound like a buyer. Now group it.

Each ad group holds five or six searches that ask the same question. "Staff scheduling software," "employee scheduling software" and "shift scheduling software" belong together. "Construction crm" and "crm for contractors" belong together. "Crm for small business" and "small business crm" belong together. Some of these groups are bigger than they look from the outside: "staff scheduling software" alone draws about 12,100 searches a month, and "field service software" about 22,200, each a category in its own right inside the broader one. Each group gets one ad that answers its question and one page that matches it, which Chapter 4 explained.

Launch every keyword in these groups on exact match. Exact match holds the ad back until a search carries the same meaning as the keyword. In the first weeks, that means every dollar lands on a search somebody on your team chose. We call this the Isolation campaign: exact match with bids set by hand, used to prove which searches turn into qualified leads. Broad match comes later, in a separate campaign we call Exploration, once the account has about 30 qualified leads a month to learn from. Chapter 6 covers that step.

A Project Management Buying Search Bids Near $44 for the Top of the Page. Payroll, Near $100. Budget by Category.

Most searches in Keyword Planner come with a two-number bid range for the top of the page. Both numbers are percentiles of what advertisers have paid in the past for that top spot, the 20th and the 80th, so most past winners sat somewhere between them. Across the buying searches in our pulls, the typical high end was:

Those are estimates from past auctions, not quotes. In our experience, what a new account actually pays usually lands within about 30% of the range either way. That's close enough to plan with. Put the high end into the payback math from Chapter 3 before you commit a budget, and you'll know whether your category's buying searches pay for themselves or whether you need to start narrower.

Next: Software Buyers Search for the Mess, Not the Product. Exact Match Covers the Category, and Broad Match Finds the Rest Later.

The keyword list covers the searches you can predict. The bigger opportunity is the ones you can't: the many ways buyers describe a problem before they know what to call the software. Chapter 6 covers how to reach them.

Every chapter, from tracking to brand campaigns, competitor campaigns and the rules for free trials, is on the guide's home page. Chapter 4 explains why every ad group in this list needs its own page.

On the software accounts we run through Google Ads management, the negative list is usually longer than the keyword list on launch day. Pull last month's search terms this week and tally the clicks that came from a rival's name.

Google Ads for SaaS

Software Buyers Compare Two Products, Not Ten. Make Sure Yours Is One of Them.

We've run paid ads for 25+ years and seen just about every way a budget goes sideways. Get on the phone with someone who does this every day. Bring your questions, your numbers and your skepticism. You'll hang up knowing what we'd do, whether you hire us or not.

David Smania · Founder, BrandRocket

25+ years running paid media for small businesses, and a low tolerance for agency theater.

Follow on LinkedIn