Ask a happy CRM customer what life was like before, and you often hear about the mess before the product. One CRM customer on Capterra put it this way: "Before using pipedrive I had Post-It Notes and Spreadsheets scattered all over my office and it was complete chaos." A HubSpot user said customer information "was spread across emails and spreadsheets, making it easy to miss follow-ups." A Calendly user was glad to stop "spending three days playing email tag."
Those people didn't start their search by typing a software category. They started by trying to fix a problem they could describe but not yet name. Some of them searched for the category later. Others searched in their own words.
This chapter covers how a software account reaches both kinds of buyer: the ones who type the category, and the long tail of people who describe their problem in their own words. The data comes from Keyword Planner's US numbers for four software categories and from buyer reviews on Capterra.
Exact Match Covers the Software Searches You Can Name. Broad Match Finds the Thousand You Can't.
A software account grows in two stages, each its own campaign, under labels we invented (you won't find them inside Google Ads):
- Isolation goes live first. Every keyword in it is exact match, every bid is a number you typed in, and every search is one you could name in advance. "Crm for small business." "Shift scheduling app." "Payroll software for restaurants." These are predictable, high intent and usually expensive.
- Exploration comes second: broad match keywords under Google's automated bidding, guided by a cost target, so Google can go after the phrasings nobody on your team would have written down.
Isolation is the foundation. It proves which searches turn into qualified leads, and it gives Google clean data to learn from. Exploration is where the scale comes from, because the searches you can't name add up: in our pulls, 13% to 36% of non-brand volume came from searches too small to list one by one.
Every Software Category We Pulled Had More Than 1,000 Searches Too Small to Bid On One by One.
We took the Keyword Planner pulls from Chapter 5, collapsed close variants, removed brand names, logins, job and stock searches, free, template, Excel and spreadsheet searches and pure definitions, and counted what was left: the non-brand searches from people looking for software or describing the problem it solves.
In every category, a handful of big searches carried a lot of the volume. Project management had nine searches above 10,000 a month; CRM had four. Underneath them sat a long tail:
- Project management: 1,069 different searches under 1,000 a month each, together about 15% of the non-brand volume.
- CRM: 1,191 such searches, about 24%.
- HR and payroll: 1,162, about 13%.
- Scheduling: 1,064, about 36%.
That's over a thousand searches per category, and those are only the ones Keyword Planner can see. You could write each of them into an exact match ad group by hand. You'd spend weeks doing it, and most of the groups would get a few clicks a month, too few for anyone to judge. Broad match exists for exactly this ground.
Buyers Describe Spreadsheets and Email Tag Before They Name a Product. Most of Those Searches Are Too Small for Keyword Planner to Count.
The reviews show where the long tail comes from. In the Capterra reviews we read, phrases like "all in one place" came up in 66 of them. Spreadsheets or Excel came up in 16 reviews and 18 Reddit buyer posts and comments. "Back-and-forth," "email tag" or "phone tag" came up in 20 reviews. Buyers remember the problem in vivid, specific language.
That language does reach the search box, a little at a time, and these searches aren't even in the counts above, which left out anything mentioning Excel or spreadsheets. US numbers: "spreadsheet crm," roughly 260 a month; "payroll spreadsheet," roughly 260; "all in one crm," roughly 210; "employee schedule excel," roughly 140. Other phrasings that sound just as natural, like "crm instead of spreadsheet" or "scheduling without back and forth," don't register at all; Google's planning tool shows no volume for them. Our reading: that usually means a search is too rare to measure, not that nobody ever types it.
That's the territory broad match covers. It reads the meaning of a search, not just its words, so, in our reading, a search like "tired of tracking leads in excel" could match a keyword like "crm for small business" if Google judges them close enough. Sometimes it's brilliant. Sometimes it matches a student's homework question. Which is why the next two sections matter.
A Broad Match Campaign Built From Scratch Learns From Nothing. Copy the Exact Match Campaign and Its Best Searches Instead.
Exploration doesn't start from a blank page. Make a copy of your Isolation campaign, then change three things in the copy:
- Seed it with three to five proven keywords, the exact match searches from Isolation that already bring in qualified leads.
- Set them to broad match. Only broad match runs in this campaign.
- Let Google bid, but give it a target. Pick the Maximize conversions strategy and set a target cost per acquisition as its guardrail. Set it from what qualified leads actually cost you in Isolation, not from a number you'd like them to cost.
Timing matters as much as setup. A broad match campaign gets smarter only as fast as conversions arrive. We launch Exploration once the account brings in about 30 qualified leads a month. For most software accounts that's around months two to three. Bigger accounts get there sooner, and accounts in a narrow niche can take longer. Expect another month or two of learning before Exploration settles.
This is where Chapter 1 pays off. If the account's primary conversion is a raw free-trial sign-up, broad match will go looking for more people who sign up for free things, and it will find them. If the primary conversion is a qualified lead from the CRM, broad match has something worth chasing.
The Search Terms Report Shows Every Student and Job Hunter Broad Match Paid For. Read It Every Day for a Month.
The first thirty days of Exploration are the expensive part of learning. Open the search terms report every day and look at what triggered your ads. In a software account, you'll see the same kinds of junk Chapter 5 described, now arriving in new phrasings:
- Students and people studying for certifications.
- Job seekers searching for roles at software companies.
- Customers of other products trying to log in.
- People checking a software company's stock price.
- People looking for free tools and templates.
- Rival products' names, which go on the negative list here; Chapter 14 covers targeting them in a campaign of their own.
Sort the report by cost, not by clicks, so the expensive mistakes surface first. For each search, ask one question: would a person who typed this ever buy our software? If the answer is no, it becomes a negative. If the answer is maybe, leave it and check again next week, when there's more data. If it already brought in a qualified lead, it may be a keeper. Be careful with negatives that are too broad: blocking the word "free" can also block "free trial," which is often exactly what a buyer types. Negate two-word phrases like "free template" rather than the single word.
Junk goes on the negative list the same morning you find it. After the first month, in our experience, weekly is usually enough. The negative list is never finished; it just grows more slowly.
Watch for the good surprises, too. When a phrase you'd never have listed keeps producing qualified leads, promote it: give it its own exact match keyword and bid inside Isolation. That's how the account's list of known buying searches grows.
Exact Match Keeps Running After Broad Match Starts. It Guards the Searches You Already Know Convert.
Turning on Exploration doesn't mean shutting off Isolation. The two run side by side, and the balance shifts over time.
Early on, Isolation carries most of the qualified leads, and Exploration is learning. As Exploration gets better, it may start to cover searches Isolation used to handle, and some accounts eventually pause Isolation altogether. If one search you care about slides down the page, Isolation is how you take it back: switch that keyword on again and set the price yourself.
A simple way to watch the balance is to compare the two campaigns' cost per qualified lead every month. If Exploration's cost drifts well above Isolation's for the same searches, the broad match campaign is paying extra for leads the exact match one could have bought for less.
Think of it as a balance, not a handoff. Exact match protects what you know works. Broad match finds what you didn't know to look for. A software account that scales needs both.
Next: Google Wants 30 Conversions Before You Judge Target CPA. A Niche Software Account Should Bid by Hand Until It Has Them.
Isolation's bids are set by hand, and Exploration's target comes from what Isolation proves. Chapter 7 covers the bidding side: setting prices by hand at launch, and the conversion count that tells you Google's automated bidding has enough to work with.
Every chapter, from tracking to brand campaigns, competitor campaigns and the rules for free trials, is on the guide's home page. Chapter 5 builds the exact match list this chapter starts from.
On the software accounts we manage through Google Ads management, Exploration waits until the CRM has confirmed enough qualified leads, not just sign-ups. If your account runs broad match today, open the search terms report and read the last week of it before you change anything else.
