Keyword and search term research
A documented keyword map with intent tiers, estimated costs and the negative lists that protect the budget from day one.
Google is the only channel where people tell you exactly what they want, in their own words, at the moment they want it. Most accounts waste that gift on broad match and default settings. I rebuild them around intent — so budget follows the searches that actually end in enquiries.
Typical movement after a structural rebuild across managed accounts. Your numbers will depend on competition, offer strength and landing page quality.
Three things decide whether a Google Ads account is profitable: what you bid on, what you say, and what happens after the click.
Keyword volume is the most over-rated number in the account. A term with 200 searches a month from people ready to buy is worth more than 20,000 searches from people writing a school project. Research starts with the language your existing customers actually used.
Campaigns are then split so that budget can be controlled by intent tier — brand, high-intent commercial, competitor and research — instead of everything competing inside one bucket where the cheapest clicks win by default.
Google rewards relevance with cheaper clicks and better positions. That reward is earned by tight alignment between the search term, the headline, and the page that loads next — which is why copy and landing pages are never treated as separate jobs here.
Responsive search ads are built with deliberate variety rather than fifteen rephrasings of the same sentence, pinned where a message must not move, and supported by every relevant asset so you take up more of the results page.
Automated bidding works well — provided it is optimising toward the right thing. Feed it every form fill including the junk, and it will cheerfully buy you more junk. So conversions are defined narrowly, values are attached where possible, and low-quality actions are excluded from bidding.
Once real sales data exists, closed deals are imported back into Google as offline conversions. The account then optimises toward revenue rather than toward form submissions, which is where most of the compounding gains come from.
Everything below is built in your account, under your login.
A documented keyword map with intent tiers, estimated costs and the negative lists that protect the budget from day one.
Search, Performance Max, Shopping or YouTube — structured, named consistently and built so reporting is readable six months from now.
Responsive search ads with real angle variety, plus sitelinks, callouts, snippets, images and lead forms filled in properly.
Ongoing work on the three components Google actually scores, because every point of Quality Score is a discount on future clicks.
Weekly search term review, negative keyword expansion, placement and audience exclusions, and PMax brand-traffic containment.
GA4 and GTM configured, conversions validated with test submissions, offline imports where you have a CRM, and a weekly written summary.
I go through structure, search terms, wasted spend, Quality Score, conversion setup and landing page experience, then send you the findings in writing — including anything you can fix yourself for free.
Research grounded in the language your customers actually use, sorted into intent tiers with the negative keyword lists built before a rupee is spent, not after.
Campaigns, ad groups, responsive search ads, all assets and landing pages are built together. Conversion tracking is validated with live test submissions before launch.
Manual or tCPA bidding with capped budgets while conversion volume builds. Search terms are reviewed every few days in this phase, because early waste is where most accounts lose the month.
Budget shifts to the winning intent tiers, ad copy is tested systematically, offline conversions start feeding the bid strategy, and spend increases only while cost per qualified lead holds.
The gap between an account built by Google's setup wizard and one built around your margins.
| Google's default setup | Typical 'set and forget' account | Structured build | |
|---|---|---|---|
| Keyword targeting | Broad match on everything | Broad match with a short negative list | Intent-tiered exact and phrase, broad only where guarded |
| Negative keywords | None until money is already gone | Reviewed occasionally | Built pre-launch, reviewed weekly, shared across campaigns |
| Ad copy | Auto-generated from your website | One RSA, fifteen similar headlines | Distinct angles, strategic pinning, full asset coverage |
| Landing page | Your homepage | Homepage or a generic contact page | Campaign-matched page with message match and fast load |
| Conversion tracking | Page-view based, counts everything | Pixel fires, never verified | Defined, deduplicated, validated, plus offline imports |
| Bidding | Maximise clicks | Maximise conversions from day one | Staged — manual, then tCPA, then tROAS once data supports it |
| Performance Max | Eats brand traffic and takes credit | Left running unexamined | Brand excluded, assets controlled, reported separately |
Quoted after the free audit, once the account size, campaign count and market competitiveness are known. Never a percentage of your ad spend.
Single-location businesses starting with search
Businesses that need search, PMax and remarketing working together
E-commerce catalogues or several branches
Your ad budget is paid directly to Google. The Google Ads account, Merchant Center, GA4 property and every conversion action are created under your ownership and stay with you.
Coaching institute, Mumbai — plenty of clicks, very few enrolments.
The account was not broken in an exotic way — it was broad match with almost no negatives, everything pointed at the homepage, and the conversion action counted every page view of the contact page. We split campaigns by intent, wrote a page for the specific course being advertised, and redefined the conversion to a real form submission. The click cost fell, but the more important change was that we could finally tell which searches produced students.
The things worth understanding before you approve a budget.
Google Ads is an auction for attention that has already declared itself. Somebody typing “best CA coaching in Thane fees” is closer to a decision than somebody typing “what is CA”. The first search may have a fraction of the volume and cost three times as much per click — and still be dramatically cheaper per enrolment. Cost per click is a vanity metric; cost per qualified lead is the real one.
Google scores every keyword from 1 to 10 on three things: how often people click your ad, how relevant the ad is to the search, and what the landing page experience is like. A higher score means you pay less for the same position. It is one of the few levers where doing the work properly gives you a permanent discount rather than a temporary bump.
In most underperforming accounts I audit, somewhere between a fifth and a third of spend is going to searches that could never have converted — free, jobs, courses, PDF, meaning, salary, competitor names you cannot serve. Reviewing the search terms report weekly and adding negatives is the single highest-return hour in Google Ads management, and it is the task most commonly skipped.
Automated bidding is genuinely good, but it optimises toward whatever you call a conversion. If your conversion is “visited the contact page”, Google will find you people who love visiting contact pages. Define conversions narrowly, attach values where you can, and once you have real sales data, import closed deals back so the algorithm learns what a good customer looks like rather than what a good form-filler looks like.
Month 1 — build, instrument, launch, aggressive search term cleanup; cost per lead is at its highest. Month 2 — structure settles, negatives compound, first copy tests conclude. Month 3 onward — bidding has enough data to work with, offline conversions start feeding back, and budget can rise while cost per qualified lead stays flat. That last sentence is the actual goal.
There is no official minimum, but a useful floor is enough spend to generate 30 or more conversions a month, because smart bidding and your own reporting both need that volume to be reliable. In low-competition local markets that can be modest; in competitive education or finance keywords it is considerably more. I will model the number for your specific keywords during the free audit.
Often yes, but rarely on its own and never without guardrails. Performance Max will absorb your brand searches and report them as its own success unless brand terms are excluded and reporting is separated. Run search as your foundation, then add PMax with controlled assets and a clear read on incremental performance.
Usually one of four things: low Quality Score from weak ad-to-keyword-to-page relevance, bidding on broad research terms instead of commercial ones, competing in an auction where your offer is not differentiated, or a bid strategy chasing volume rather than efficiency. An audit will identify which of these applies to your account.
Yes. Shopping performance is decided mostly in the product feed — titles, attributes, image quality, GTINs and disapproval fixes — so feed work is treated as part of the campaign management rather than a separate project.
Boosting has no keyword targeting or intent at all. Against an agency, the difference is that the person auditing your account is the person working in it every week, the fee is flat rather than a share of your spend, and landing pages and tracking are included instead of quoted separately.
Yes. Everything is built in your own Google Ads, Merchant Center and GA4 accounts. You keep all campaigns, keyword history, audiences, conversion data and landing pages with no export or handover fee.
Ready to scale your growth? Drop me a message or reach out directly.