Before adding a keyword to Google Ads, ask a simple question: if someone searches for this and clicks, can your business give them what they came for?
That question is easy to lose sight of when a tool returns hundreds of ideas. A large search volume looks promising. A low bid looks affordable. Neither tells you whether the person wants your service, can buy from you, or will find a useful answer on your landing page.
This guide covers keyword research for Google Search campaigns. You will build a focused list, decide which searches deserve a test, and work out what to exclude before the campaign starts spending.
Start with the offer you can actually deliver
Write down the service or product you want to advertise, the area you serve, and the action you want a visitor to take. Be specific about what counts as a useful enquiry. A completed form is not much help if the person needs a service you do not provide.
Consider an air-conditioning repair business in Mohali. It might handle home repairs and installations, cover selected neighbourhoods, and want customers to book a technician. It may not sell spare parts, repair car air conditioning, or offer training.
Those details belong in the research brief. They will help you reject irrelevant ideas, choose negative keywords, and spot gaps in the website before you pay for traffic.
- Offer: what can a customer buy or book?
- Coverage: where can you serve them, and during which hours?
- Customer: who is the service intended for?
- Outcome: what makes a lead or sale worth paying for?
Build your seed list from customer language
Start with a few clear descriptions of the offer. For the repair business, useful seeds might include “AC repair”, “air conditioner servicing”, “split AC repair” and “AC installation”. Add a location where it changes the meaning, such as “AC repair Mohali”.
Look at sales enquiries, customer emails, product categories and the wording on your service pages. Customers may describe the problem before they know the service name. “AC not cooling” can reveal a different need from “annual AC maintenance”.
Keep separate notes for services, problems, buying signals and locations. You do not need to generate every possible combination. The first job is to cover the ways a real customer might describe the purchase.
Use the Highbav keyword research tool to expand one topic at a time. Copy the ideas that fit your offer, then research the next topic. A smaller, reviewed list is easier to organise than one large export containing several unrelated services.
Build a starting list with our guide to finding seed keywords if the first ideas still feel too broad.
Research the market you intend to target
Choose the country first. For a local service, type the city name and select the matching location from the suggestions. Check the full location label: two cities can share a name, and a familiar local name may appear under a different official name.
Research each important service area separately when demand or delivery costs differ. A business serving Mohali and Delhi should not assume that the same bids, customer needs or landing-page claims will suit both markets.
The location selected in the research tool does not change your Google Ads campaign settings. Review the campaign’s locations, language and location options separately before launch. For campaigns in Hindi, Punjabi or another language, also research that language in Google Keyword Planner; Highbav’s current tool requests English keyword ideas.
Google explains how to adjust research locations, languages and networks in its Keyword Planner targeting guidance.
Separate buying intent from general interest
Read each idea as a customer’s request. “AC repair near me” suggests someone looking for a provider. “How to clean an AC filter” suggests someone who may want instructions. Both are relevant to the business, but they call for different decisions about advertising.
| Example search | Likely need | Research decision |
|---|---|---|
| AC repair Mohali | Find a local repair service | A strong candidate if the area and service are covered |
| AC repair cost | Compare prices before booking | Consider testing if the page explains charges and the next step |
| AC not cooling | Diagnose a problem or arrange a repair | Review the results and decide whether the page can satisfy that need |
| How to clean an AC filter | Find instructions | Usually a lower priority for a limited lead-generation budget |
| AC repair jobs | Find employment | Exclude from a customer-acquisition campaign |
These are illustrative decisions, not performance findings. Review the actual search results, your offer and your budget before deciding. A word such as “cost” does not make a search unqualified; many serious buyers check prices before contacting anyone.
Our search intent guide explains how to validate what a searcher needs before choosing the page and offer.
Use the metrics to judge the opportunity
Average monthly searches estimate historical demand. They are not a forecast of the clicks your campaign will receive. A phrase with fewer searches can still be useful if it closely matches a profitable service. Seasonal averages also need context: an annual figure can hide a short, busy period.
Advertiser competition describes competition in Google Ads. It is not an organic ranking difficulty score, and a low value does not prove that a keyword will be profitable.
Low and high top-of-page bid ranges give you a view of historical bidding conditions. They are not your guaranteed cost per click or a promise of a particular ad position. Use them to identify ideas that deserve a closer budget check.
Google’s historical metrics documentation describes these measures. Keep missing figures marked as unavailable rather than treating them as zero.
In your shortlist, record the keyword, location, intended customer, relevant landing page, bid range and the reason you want to test it. That last column is often more useful than another score.
Group the list and choose match types deliberately
Group keywords around a shared need that one ad and landing page can address. Repair, installation and maintenance may deserve separate ad groups because the offer and message differ. Splitting every small wording variation into its own group usually creates more work without improving the customer experience.
Choose the match type as part of the test plan:
- Exact match: written as
[ac repair mohali]. It can match searches with the same meaning or intent; it is not restricted to an identical string of words. - Phrase match: written as
"ac repair mohali". It can reach searches that include the keyword’s meaning. - Broad match: written without brackets or quotation marks. It can reach a wider range of related searches.
Google recommends using Smart Bidding with broad match. Before expanding reach, make sure the conversion actions used for bidding reflect outcomes you value. An account optimising towards low-quality form submissions can appear successful while producing poor sales enquiries.
There is no need to add every misspelling or rearrangement of the same phrase. Close variants are part of keyword matching. Read Google’s match type guidance when deciding how much reach and control the campaign needs.
For a fuller page-planning workflow, use our keyword clustering and landing-page mapping guide to connect each theme with a suitable destination.
Prepare negative keywords before launch
Build a separate list of searches you do not want to pay for. For the repair business, that might include employment, courses, do-it-yourself instructions or spare-parts shopping, depending on what it sells.
Avoid downloading a generic negative list and applying it without review. Excluding “free”, for example, could block a useful search for “free AC repair quote”. Excluding a brand could remove customers whose equipment you are happy to service.
Choose the negative match type carefully. Negative keywords work differently from positive keywords and do not expand to close variants in the same way. Where relevant, review synonyms and singular or plural forms instead of assuming one exclusion covers them all. Google’s negative keyword documentation explains the rules.
After launch, use the search terms report to identify irrelevant queries and new opportunities. It is a useful review tool, but it does not expose every individual search because reporting is subject to privacy thresholds.
Check whether the numbers make sense
A competitive keyword can still be worth testing if the sale supports the acquisition cost. A cheap keyword can waste money if the clicks rarely turn into suitable customers.
Here is a hypothetical service-business example. Each completed job contributes ₹4,000 after delivery costs, and one in four qualified leads becomes a customer. That gives an expected contribution of ₹1,000 per qualified lead before advertising and other overheads.
If the business chooses a target cost of ₹600 per qualified lead and expects 8% of ad clicks to produce such a lead, the implied planning ceiling is:
₹600 × 8% = ₹48 per click
These numbers are assumptions, not a forecast. The business would need to measure its actual conversion rate, lead quality and close rate. If a keyword’s likely click costs leave little room for error, start with a more focused market or improve the offer and landing page before increasing spend.
Use your own margins rather than copying another advertiser’s target. Revenue per sale alone is not enough to set a sensible acquisition cost.
Treat the first list as a test plan
Before launch, check that each keyword group has a relevant page, a clear offer, working forms or call handling, and verified conversion tracking. Make sure the team answering enquiries knows which services and locations the campaign promises.
Review early traffic for obvious mismatches, but allow for conversion delays when comparing performance. Look beyond the number of leads: which groups produced suitable enquiries, booked appointments and profitable work?
Keep a record of what you changed and why. Useful new search terms may deserve their own test. Repeated irrelevant searches may need exclusions. A relevant keyword with poor results may point to a landing-page problem rather than a problem with the keyword itself.
Your finished research should give you a short, defensible answer to four questions: what are we targeting, why are we targeting it, where will the visitor land, and what would make the test worth continuing?
Return to the keyword research tool to build your shortlist. If you are also planning organic content, read our keyword research for SEO guide. For help structuring and managing the campaign, explore Highbav’s Google Ads services.
