A lot of small businesses are staring at the same Google Ads problem right now. Clicks are coming in, the spend looks active, and the search terms look busy, but the phone isn't ringing enough to justify the budget. The issue usually isn't that the account needs more keywords. It's that the account is buying the wrong intent.

That's why PPC keyword research in 2026 has to be treated less like a brainstorming exercise and more like a filtering system. Google's keyword auction model has been evolving since GoTo.com opened a keyword-based paid-search auction in February 1998, and Google shifted AdWords into a keyword auction structure with AdWords Select in February 2002 (history of PPC keyword auctions). What changed over time is the amount of data, automation, and semantic matching sitting between your keyword and the actual search query.

Small businesses feel that shift first. Budgets are tighter, broad informational queries get soaked up by richer search results, and AI has become mainstream enough that even small teams are using it operationally. One 2026 roundup reports 58% of small businesses use generative AI, up from 40% a year earlier, while another says 68% use at least one AI tool (small business AI adoption statistics). Used well, AI helps small teams sort search terms faster, cluster intent, and clean up negatives without turning keyword work into a weekly time sink.

The Real Problem With Most PPC Keyword Lists

The account usually looks fine at first glance. A local HVAC company has campaigns running, keywords loaded, ads approved, and a monthly budget of $2,400. Then you open the search terms and see the waste. A big chunk of clicks came from searches like “hvac salary” and “is hvac a good career.” The business owner didn't ask for job seekers. They asked for service calls.

Most keyword lists grow by accumulation. Someone adds “HVAC repair,” then “heating service,” then “air conditioner,” then close variations, then more city names, then broad terms because traffic feels slow. Very few accounts are built around a hard question: which searches deserve paid budget, and which ones should never trigger an ad?

That distinction matters more now because search behavior isn't as simple as matching a string to a term. Search ads have become more measurement-heavy as the market has scaled, with one industry summary projecting the global search advertising market could reach about $227 billion by 2026, alongside an average cost per click around $4.66 (search advertising and keyword research statistics). When clicks cost real money, sloppy intent mapping gets expensive fast.

Why keyword discovery isn't the main bottleneck

A decade ago, the edge often came from finding more terms than your competitor. That's not where most small accounts win now. Major platforms rely more on semantic matching, automation, and audience signals, which means the control point is often your exclusions and your landing-page alignment, not the size of your list. Recent PPC coverage also notes that exact match behaves more semantically, broad match tends to work best with Smart Bidding, and mature accounts are moving toward fewer keywords plus heavier negative-keyword management (2026 PPC trends and semantic matching).

Practical rule: If your keyword list keeps growing but lead quality keeps slipping, the problem usually isn't reach. It's weak exclusions.

What a better workflow looks like

The accounts that hold together usually do seven things well:

• Define conversion value first. They know what a call or form is worth.
• Mine real queries first. They start with existing search data, not guesses.
• Use match types and negatives together. They don't manage them as separate chores.
• Read the SERP before bidding. They check if the click is even available.
• Group by intent. They send different searches to different pages.
• Track keyword-level business outcomes. They don't optimize for click volume alone.
• Review weekly. Keyword research is ongoing, not a one-time setup task.

Set Conversion Goals Before You Touch a Keyword Tool

Before opening Keyword Planner, autocomplete, or any third-party platform, define what a conversion means in dollars and in operations. If you don't, PPC keyword research turns into educated guessing.

A plumber is a simple example. A booked standard service call might be worth $180. An emergency call might be worth $400. A basic quote request might be worth $90. Those values don't come from Google Ads. They come from the business. Once you have them, you can decide what you can afford to pay for a lead and, by extension, what you can afford to pay for a click.

Start with one primary conversion

Most small service businesses should choose one primary action per campaign. Usually that's one of these:

1. Phone calls for urgent services like plumbing, HVAC, towing, and dental emergencies.
2. Booked appointments for clinics, salons, and professional services.
3. Lead forms for estimates, consultations, and non-urgent service requests.

If every action is treated as equally valuable, the bidding logic gets distorted. Emergency intent and research intent shouldn't share the same tolerance for spend.

Build the worksheet before the keyword list

Use a simple worksheet like this.

Conversion Value Worksheet for a Plumbing PPC Account

Conversion Type	Average Value	Target CPA	Max CPC
Booked service call	$180	$60	$15
Emergency call	$400	$120	$30
Quote request	$90	$30	$7.50

This table isn't just for finance. It drives keyword decisions later.

A high-intent keyword like “emergency plumber near me” can tolerate more aggressive bids than a softer term like “plumber estimate.” It also affects your negative thresholds. If a query keeps spending against a low-value conversion path without producing qualified leads, you know where to tighten.

Why this matters operationally

Teams often try to solve poor PPC results by expanding keyword coverage. That's backward. The first fix is usually tracking and economics. If conversion tracking is shaky, clean that up first with a proper Google Ads conversion tracking setup. Otherwise you'll optimize toward form submissions that never became jobs, or phone calls that lasted ten seconds and came from the wrong city.

A keyword isn't valuable because it gets searched. It's valuable because the business can profit from the action that follows.

Once conversion values are clear, keyword research becomes less emotional. You stop asking, “Should we bid on this term?” and start asking, “Can this query produce a lead at a cost the business can sustain?”

Build Your Seed List From Searches You Already Get

The strongest seed list usually isn't hiding inside a fancy tool. It's already sitting in your business data. Most small advertisers skip that and jump straight into expansion, which is why they end up with jargon-heavy keyword sets that don't reflect how buyers search.

Google's own guidance points to the Search terms report as the native place to discover new negatives, and that same report is just as useful for identifying commercially relevant language in the first place (Google Ads Search terms report and negatives). That's where real query wording shows up.

Five places to pull seed terms this week

Start with sources that reflect actual customer language:

• Search terms report. Pull current paid queries and sort by conversions, spend, and obvious irrelevance.
• Site search logs. People often type cleaner intent on your own site than they do into a broad keyword tool.
• Google Search Console queries. Organic impressions reveal adjacent demand, especially for comparison and problem-based searches.
• Sales call transcripts and chat logs. Customers rarely speak in the polished terms marketers use.
• Auction preview and competitor ad copy. You're not copying competitors. You're identifying recurring commercial phrasing.

A dental clinic is a good example. Patients search “toothache relief,” “emergency dentist open now,” and “broken tooth repair.” They rarely search “endodontic pain management” unless they're in the field.

Expand carefully, then trim hard

After the raw seeds are collected, expand them with controlled inputs. Autocomplete and People Also Ask can help uncover modifiers, question variants, and local phrasing. But use them to widen understanding, not to dump every discovered phrase into the account.

If you need a quick way to compare themes and organize discovery ideas, a lightweight resource like browse keyword analysis can help you spot related term groupings before you decide what belongs in paid search.

A useful AI workflow for small businesses is to export raw queries, then use generative AI to cluster them by service, urgency, location, and intent. AI is excellent at sorting messy search language into usable buckets. It's much less reliable at deciding what deserves budget, so keep that judgment human.

Good seeds sound like buyers

Use phrases tied to needs, not internal labels:

• Better seed: “same day water heater repair”

• Weak seed: “residential hydrothermal system solutions”

• Better seed: “family dentist for anxious patients”

• Weak seed: “preventive oral care services”

• Better seed: “AC not cooling upstairs”

• Weak seed: “multi-zone residential airflow optimization”

Good PPC keyword research starts with live language. If your seeds don't sound like customers, the rest of the account won't either.

Match Types and Negatives Are Really One Decision

Small accounts often manage match types in one tab and negatives in another, as if they were separate systems. They aren't. They're one traffic-control system. The positive match type decides how far Google can stretch into adjacent meaning. The negative structure decides where that stretch stops.

Google Ads recognizes broad, phrase, and exact as the three positive match types, with broad being the widest and exact the most restrictive (overview of Google Ads match types). But “restrictive” doesn't mean literal anymore. Exact match can still trigger for close variants and same-meaning searches like plurals, misspellings, reorderings, and paraphrases (how exact match now behaves). That's why treating exact match as perfect precision is outdated.

What each match type is really for

• Broad match is for controlled discovery and scale. It works best when paired with Smart Bidding and a serious negative list.
• Phrase match is useful when word order or phrase integrity still matters commercially.
• Exact match is best when you need tighter control around a proven high-intent cluster, not when you're trying to build the whole account around single strings.

The old “exact match only” playbook gave advertisers psychological comfort. It also limited learning and often missed valuable variant searches. For many small accounts now, broad match plus strong conversion signals can outperform a rigid exact-only setup. But broad match without active exclusions is just expensive curiosity.

For a deeper explainer on how exclusions work inside campaign structure, this guide on negative keywords in Google Ads is a useful reference.

Build negatives in layers

Google documents negative match types clearly. Negative broad blocks queries containing all negative terms in any order, negative phrase blocks the exact phrase in order, and negative exact blocks only that exact phrase without extra words (Google Ads negative match behavior). That structure gives you a practical filtering hierarchy.

Use it like this:

• Broad negatives for obvious junk themes such as jobs, free, DIY, salary, tutorial, course.
• Phrase negatives for recurring low-value clusters that share wording.
• Exact negatives for one-off waste that you don't want to overblock.

Broad negatives are for themes. Phrase negatives are for patterns. Exact negatives are for isolated leaks.

Here's how that works in a water heater repair campaign. A broad keyword may still reach useful searches such as “urgent hot water heater fix” or “repair leaking water heater near me.” That can be productive. But if the account also attracts “how to repair water heater,” “water heater repair course,” or “water heater technician salary,” your negative list needs to absorb those paths quickly.

This video does a solid job showing how layered control works in practice.

What actually fails in practice

The most common failure isn't choosing the wrong match type. It's applying negatives too aggressively too early, or not aggressively enough after enough data comes in. Broad and phrase negatives can block valuable long-tail traffic if you panic after a few irrelevant searches. On the other hand, leaving the account open for weeks while “the algorithm learns” usually means you're paying for research the business didn't ask for.

Read the SERP Before You Bid on Anything

Some keywords shouldn't be bought, even if they sound relevant. The search results page tells you that before your budget does.

Recent reporting notes that AI Overviews can appear above the first text ad on informational queries, and one eCommerce analysis cited in 2026 coverage reported first-position click-through drops of 14% to 22% year over year on AI-Overview-active queries (AI Overviews and PPC click behavior). That doesn't mean informational terms are useless. It means you need to inspect whether the SERP still leaves room for a paid click that can convert.

A quick two-query comparison

Take these two HVAC searches:

• “best HVAC filter”
• “HVAC filter replacement near me”

The first query often attracts AI summaries, organic listicles, retailer results, and informational comparison content. Even if an ad shows, the searcher may not need a local service company yet. The second query has local commercial intent. The person likely needs a supplier, installer, or service provider.

That's the difference between relevance and commercial survivability. A term can be related to your business and still be a bad paid-search buy.

What to scan on the results page

Review each priority query manually and look for these signals:

• Ad density. Are there enough paid slots to make the auction worth entering?
• AI Overview presence. Does Google summarize the answer before users click?
• Local pack visibility. For local services, this can support intent or crowd out text ads depending on the layout.
• Shopping or rich features. Product-heavy results may shift clicks away from lead-gen ads.
• Competitor message quality. Weak competitor copy can create an opening. Strong aligned copy can raise the bar.

If you're seeing answer-first behavior across your informational terms, it's also worth understanding the broader trend around zero-click search behavior. It changes what counts as an efficient keyword, especially for service businesses.

If the SERP answers the question completely, don't assume your ad can still win the click profitably.

A fast SERP audit saves money because it stops you from funding low-intent curiosity. Paid search works better when you buy the queries that still require action.

Group Keywords and Match Them to the Right Landing Page

Keyword grouping shouldn't start with ad groups. It should start with intent and destination. If a searcher needs urgent service, they shouldn't land on a generic homepage. If they're comparing options, they may need more context before they convert.

That's where a lot of PPC keyword research breaks down. Advertisers collect terms well enough, then dump hundreds into one ad group and point all traffic to one page. The account becomes technically active but commercially fuzzy.

Group by service, location, and buying stage

A practical structure looks like this:

• Service cluster such as drain cleaning, emergency plumbing, or water heater repair
• Location cluster such as city, neighborhood, or service radius modifiers
• Intent stage such as urgent, evaluative, or educational

Those three dimensions usually produce clean groupings without overcomplicating the account.

Here's a simple mapping example.

Keyword Group to Landing Page Mapping Example

Keyword Group	Intent Stage	Landing Page Type	Primary CTA
Emergency plumber St. Petersburg	High intent	Fast-loading emergency service page	Call now
Water heater repair near me	High intent	Dedicated repair service page	Book service
Tankless water heater installation	Commercial investigation	Service comparison page	Request estimate
Water heater vs tankless	Research	Long-form comparison page	Get a quote

Two pages, two very different jobs

“Emergency plumber [city]” should lead to a page built for speed and action. Prominent phone number. Short form. Trust markers near the top. Tight headline that mirrors the search.

“Water heater vs tankless” needs a different experience. Longer copy, clearer explanation, comparison framing, and multiple CTA points. Forcing both keyword groups into the same page usually weakens conversion rates on both sides.

If you're improving page alignment, a focused guide to PPC landing page strategy is worth reviewing alongside the account structure.

Forecast conservatively

You don't need elaborate modeling to decide whether a keyword group deserves budget. Use historical conversion rate where you have it. Where you don't, stay conservative. Estimate the clicks required to hit one qualified lead, then compare that expected cost against your target CPA from earlier.

That exercise often exposes keyword groups that look attractive in a tool but have no realistic landing-page path. If you can't build a page that answers the specific search intent cleanly, the keyword probably hasn't earned its place yet.

One operational note here: this is also the point where a provider such as Netco Design LLC can fit into the workflow, since its PPC management and landing page work are built around keyword research, competitive analysis, and conversion-focused page alignment rather than traffic volume alone.

Track the Right KPIs and Run This on Monday

Keyword work drifts when teams watch the wrong numbers. Click volume feels productive because it updates fast. But clicks don't tell you whether the query mix is healthy, whether the landing page matches intent, or whether AI-heavy SERPs are reducing the available traffic on certain classes of searches.

The keyword-level view should stay narrow. Focus on Search Impression Share, Quality Score, Cost per Conversion, and Wasted Spend Rate. Those metrics tell you whether you're visible, relevant, efficient, and leaking budget.

The four numbers that matter most

• Search Impression Share. If strong keywords are barely entering auctions, visibility is the issue.
• Quality Score. If relevance is weak, check ad grouping and landing-page alignment before increasing bids.
• Cost per Conversion. Your earlier conversion worksheet becomes useful.
• Wasted Spend Rate. Track spend on queries that produce no qualified action or clearly irrelevant intent.

For teams that want a broader benchmark set around reporting discipline, these marketing KPI examples can help frame dashboards without slipping into vanity metrics.

A Monday operating checklist

Run this once a week. Don't overcomplicate it.

1. Pull the Search terms report. Review new queries by spend and conversions.
2. Add negatives. Start with obvious waste, then recurring low-value patterns.
3. Sort high-cost zero-conversion queries. Pause, exclude, or move them to a different intent bucket.
4. Check Search Impression Share lost to rank. Strong keywords may need bid or quality improvements.
5. Verify tracking. Make sure calls, forms, and booked actions are still recording correctly.

If AI Overviews are affecting a group of informational terms, you may not see the full picture in one metric. Look for a pattern instead: lower useful click volume, softer CTR, and weaker commercial outcomes on query classes that used to support top-of-funnel traffic. That doesn't always mean pause everything. It may mean demote those terms to research campaigns, lower bids, or route spend toward more transactional searches.

Weekly PPC keyword research isn't about finding new words. It's about deciding which searches still deserve access to your budget.

Small businesses can make this process lighter with AI. A practical use case is feeding weekly search terms into an AI tool to label them by intent, urgency, and negative theme before a human reviews final actions. That's the useful version of AI for small teams. It speeds up classification without outsourcing the money decisions.

The accounts that improve consistently aren't the ones with the biggest keyword lists. They're the ones with the sharpest filters, the cleanest intent grouping, and the discipline to review query behavior every week.

---

Netco Design LLC helps businesses turn PPC into a lead-generation system instead of a click-buying habit. If you need support with keyword research, landing pages, conversion tracking, or ongoing Google Ads management tied to real business outcomes, visit Netco Design LLC.