Somewhere between 30% and 50% of what most Google Ads accounts spend each month is genuinely wasted — not because the advertiser chose wrong products or services, but because of structural issues that compound quietly: match types that are too broad, Quality Scores stuck at 4, bid strategies running without enough conversion data, and audiences that include people who already converted three months ago. These are fixable problems. The 25 levers in this guide address them directly, with the highest-leverage items first.
The Audit Before the Optimisation
Before pulling any lever, spend 30 minutes in four specific reports. These four views reveal where roughly 80% of waste is hiding in almost every account.
Impression Share report: Go to Campaigns, add the columns for Search Impression Share, Search Lost IS (Budget), and Search Lost IS (Rank). If you're losing impression share to budget on your highest-converting campaigns, that's a reallocation problem — other campaigns are absorbing budget that should go elsewhere. If you're losing to rank, that's a Quality Score or bid problem to fix first.
Quality Score distribution: At the keyword level, add the Quality Score columns including Ad Relevance, Expected CTR, and Landing Page Experience. Filter for keywords with spend in the last 30 days. Any keyword with a score below 5 and meaningful spend is costing you 30–50% more per click than it should be.
Search Terms report: Set the date range to the last 30 days and sort by spend descending. Look at the top 20 search terms by spend. Count how many are genuinely relevant to your product or service. In a poorly managed account, 30–40% of the top-spend terms are off-target — those become immediate negatives.
Device performance split: Go to Campaigns and segment by Device. Compare CPA across Desktop, Mobile, and Tablet. A 2x difference in CPA between desktop and mobile is common, and a 2x bid adjustment is entirely appropriate when the data supports it. Most accounts apply identical bids across all three devices despite wildly different conversion rates.
Levers 1–5: Keyword Hygiene
Keyword hygiene is unglamorous but produces faster and more certain results than any bidding experiment.
Lever 1 — Match type strategy: If your account relies heavily on Broad Match keywords, you are funding Google's expansion into adjacent search territory you never intended to enter. For most Indian SMB accounts, a phrase match and exact match strategy gives far better control. Use Broad Match only in campaigns with strong audience signals — Customer Match lists, in-market audiences — where Google has something meaningful to anchor its expansion on.
Lever 2 — Negative keyword sculpting from search term reports: This is not a one-time task. Pull the search terms report weekly for the first three months of any campaign, and monthly thereafter. Add irrelevant terms as negatives at the campaign or account level. Focus especially on navigational queries (people searching for a competitor by name, or typing "login" or "careers"), informational queries that signal zero purchase intent, and terms from unrelated industries that broad match keywords are pulling in.
Lever 3 — Keyword-to-ad-group structure: Each ad group should contain keywords with the same intent and sensible shared ad copy. Mixing "web design company Kochi" with "website maintenance Kerala" in one ad group means no single ad can be fully relevant to both. Splitting them lets you write copy that directly matches the query pattern — which directly improves your Quality Score components.
Lever 4 — Remove duplicate keywords across campaigns: Duplicate keywords in multiple campaigns cause your campaigns to bid against themselves, inflating your own CPCs. Use Google Ads Editor to run a keyword diagnosis and filter for exact duplicates. Where duplicates exist, decide which ad group they belong in and remove them from the others, then add cross-campaign negatives to enforce that boundary going forward.
Lever 5 — Pause zero-impression keywords after 30 days: Keywords that have never triggered an impression in 30 days are either stuck in a Quality Score hole (below the first-page bid by a large margin) or simply not matching any real search traffic at that bid level. Pausing them cleans up the account. If you believe in a keyword strongly, investigate why it's not getting impressions — check the first-page bid estimate — before trying to reactivate it.
Levers 6–10: Bidding Strategy Fixes
Bidding strategy selection is where accounts leave the most money on the table — either through premature automation or excessive manual control that ignores available signals.
Lever 6 — Switch from Manual CPC to Target CPA when you have conversion data: Manual CPC gives control but costs you the ability to adjust bids in real time across dozens of user signals simultaneously — device, location, time, search history, audience membership. Target CPA uses all of these at once. The prerequisite is volume: 30–50 conversions in the past 30 days before switching.
Lever 7 — Set a realistic Target CPA based on 30-day data: The number you enter as your Target CPA directly affects delivery. Set it too low and the algorithm restricts impressions aggressively. Set your starting target 10–20% above your actual current CPA — let the algorithm calibrate for two weeks, then tighten the target gradually. Entering an aspirational number based on what you'd ideally like to pay, rather than what the account is actually delivering, causes underperformance in the first weeks and leads teams to abandon smart bidding prematurely.
Lever 8 — Bid adjustments for device, location, and time: Even with smart bidding campaigns you can apply bid adjustments as explicit overrides. If your conversion data shows mobile converts at 50% the rate of desktop, a -30% mobile adjustment tells the algorithm your floor. For Kerala service businesses: a law firm in Thiruvananthapuram may see 3x the conversion rate from Trivandrum city centre searches compared to rural district-level queries — set location modifiers accordingly.
Lever 9 — Portfolio bidding across complementary campaigns: Portfolio bid strategies pool conversion data across multiple campaigns, letting the algorithm hit a shared CPA target more efficiently than individual campaign targets allow. Group campaigns that share conversion actions — for example, separate campaigns for web development, SEO, and digital marketing services all converting through the same lead form — under one portfolio strategy.
Lever 10 — Smart bidding seasonality adjustments for Indian festivals: During Onam, Diwali, Vishu, and Eid, conversion rates often spike 20–40% for consumer-facing businesses. Standard smart bidding algorithms take 7–14 days to adapt to volume shifts. Use the Seasonality Adjustment feature under Tools to tell Google to expect a 25–30% uplift in conversion rate during the festival period. Remove the adjustment immediately after it ends. This is particularly relevant for Kerala e-commerce, jewellery, and hospitality businesses.
Levers 11–15: Quality Score Improvements
Quality Score is Google's proxy for relevance, and each point increase translates directly into a lower cost-per-click. An advertiser at QS 8 pays approximately 20% less per click than a QS 6 advertiser at equivalent position — without changing a single bid.
Lever 11 — Ad copy relevance to keyword intent: The headline of your Responsive Search Ad should directly mirror the searcher's intent. If someone searches "SEO services Kochi," the first pinned headline should contain those exact words — not a generic brand tagline about being "your digital growth partner." Google's Expected CTR component of Quality Score rewards this specificity. Build separate RSAs per keyword theme rather than one per ad group trying to serve all intents.
Lever 12 — Landing page load speed: Google's Landing Page Experience column in the Quality Score breakdown shows "Below Average" when the page takes more than 3 seconds to load on mobile. Each additional second of load time costs approximately 7% of your conversion rate. Use PageSpeed Insights to measure your landing pages, then address the common culprits: uncompressed images, render-blocking scripts, and slow server response times. This single fix often produces the largest Quality Score improvement of anything on this list.
Lever 13 — Expected CTR: pin your strongest headline to Position 1: In RSAs, Google tests combinations across all your assets. If your asset report clearly shows one headline consistently generates a higher CTR, pin it to Position 1. Leaving all positions unpinned works for discovery, but once you have evidence that a specific headline outperforms, pinning it reinforces the Expected CTR signal over time.
Lever 14 — Ad extensions that boost click-through rate: Sitelink, callout, and structured snippet extensions increase your ad's visual footprint and drive CTR improvements that feed back into Expected CTR and Quality Score. Every campaign should have at minimum: 4 sitelinks with unique descriptions, 4 callout extensions, and 1 structured snippet set. For lead generation campaigns, add a lead form extension to capture interest without requiring a landing page click at all.
Lever 15 — Single-keyword ad groups for high-value terms: For your 5–10 highest-spend keywords, consider single-keyword ad groups with one exact match keyword and one RSA whose every headline is built around that term. This produces the tightest possible keyword-to-ad relevance and typically achieves Quality Scores of 8–10. It requires more maintenance but is justified for terms driving significant monthly spend.
Levers 16–20: Audience and Targeting Refinements
Audience layers overlaid on search campaigns are one of the most underused capabilities in Indian PPC accounts. Most accounts pay the same CPC for a first-time visitor as for someone who has already visited their pricing page or submitted a partial inquiry form.
Lever 16 — Customer Match upload for CRM lists: Upload your existing customer or lead database to Google Ads as a Customer Match audience. Apply it with a positive bid adjustment of +30–50% on your search campaigns. These users already know your brand; when they search for your service category again, you want to win that impression more aggressively. Customer Match also enables the exclusions in Lever 18.
Lever 17 — In-market audience bid adjustments: Google's in-market audience segments classify users by their recent browsing behaviour into purchase-intent buckets. Add relevant segments — IT Services, Business Software, Home Improvement, Legal Services — to your campaigns in Observation mode. After 2–3 weeks of data, apply +15–25% bid adjustments to any in-market segment that shows a meaningfully lower CPA than your campaign average.
Lever 18 — Excluding converted customers from acquisition campaigns: If your goal is new customer acquisition, add your Customer Match list of existing customers as an exclusion audience. This prevents spend on people who have already converted — budget that should go toward genuinely new prospects. For service businesses in Kerala, exclude clients signed in the last 6 months; for subscription businesses, exclude all active subscribers.
Lever 19 — Remarketing for abandoned checkout in Indian e-commerce: Users who reached your checkout page but did not complete a purchase are among the highest-intent audiences in your account. Create a remarketing list from the checkout URL and exclude the thank-you/order-confirmation URL so purchasers are removed automatically. Apply this list to a dedicated RLSA campaign with 30–40% higher bids and ad copy that addresses the friction point — free shipping, pay-on-delivery, or easy returns.
Lever 20 — Location targeting refinements for Kerala businesses: Radius targeting — say, 15 km around Kochi city centre — is often more precise than targeting the entire Ernakulam district. For service businesses with physical locations, radius targeting ensures budget goes to the catchment area of actual visitors. Compare conversion rates and CPAs for radius-targeted versus city-level campaigns over 30 days before deciding which to scale.
Levers 21–25: Account Structure and Budgets
Lever 21 — Shared budgets across complementary campaigns: If you have two or more campaigns with related goals — one targeting Kochi and one targeting Trivandrum for the same service — a shared budget lets Google automatically route more spend toward whichever is converting more efficiently on any given day. This eliminates the problem of one campaign exhausting its daily budget at noon while another under-delivers.
Lever 22 — Ad scheduling based on conversion data by hour and day: Pull your conversion data segmented by Hour of Day and Day of Week using the Segments menu. For most B2B service businesses in Kerala, conversions cluster between 9 AM and 7 PM on weekdays, with a sharp drop on Sundays and after 9 PM. Scheduling campaigns to reduce bids by 60–70% during overnight hours eliminates spend on low-intent off-hours traffic.
Lever 23 — Separating brand and non-brand keywords: Brand keywords and non-brand keywords behave entirely differently: brand terms carry high Quality Scores, low CPCs, and high conversion rates; non-brand terms carry lower scores, higher CPCs, and lower conversion rates. Mixing them hides the true CPA of your acquisition spend. Separate them into distinct campaigns with distinct budgets so each can be evaluated and optimised independently.
Lever 24 — RSA headline testing with the Combinations report: The Combinations report within your RSA asset detail shows which headline pairings Google tested and which performed best. Use this data monthly to retire underperforming headlines and replace them with variations of your top performers. An RSA regularly updated with insights from this report typically achieves 15–20% higher CTR than one with stale copy.
Lever 25 — Cross-campaign negative keyword lists: When two campaigns share keywords — even accidentally — they bid against each other in the same auction, inflating your CPCs. Create a Shared Negative Keyword List for each campaign that explicitly excludes the keyword territory of other campaigns. Your brand campaign should have all non-brand terms as negatives; your non-brand campaigns should have all brand terms as negatives. This boundary enforcement reduces internal competition and lowers average CPC across the account.
A Real Account Audit: Before and After
A Kerala-based IT services company — providing software development, ERP consulting, and web application projects to businesses across South India — came in with a ₹2 lakh/month Google Ads budget and a CPA of ₹8,000 per qualified lead. Their account had been running for 14 months with the same structure, managed intermittently.
The audit took 90 minutes. Here is what it found: 68% of spend was on Broad Match keywords, and the search terms report showed phrases like "free software development course," "ERP software price list PDF," and a direct competitor's brand name — none of which matched the company's service offering. The average Quality Score across active keywords was 4.8. Their landing page loaded in 6.1 seconds on mobile. Three of their four ad groups had ad copy with no mention of the keyword terms being bid on. There were no audience layers, no remarketing, and no device adjustments despite mobile CPA being 3.1x higher than desktop. Both their brand campaign and their primary non-brand "IT services Kerala" campaign contained overlapping keywords, causing internal auction competition on every search.
Over six weeks, the team worked through the levers systematically: structural changes first (match type cleanup, negative keyword lists, brand/non-brand separation), then Quality Score work (landing page speed brought to 2.4 seconds, ad copy rewritten to mirror keyword intent, extensions completed across all campaigns), then bidding (Target CPA enabled with a starting target of ₹9,500, then tightened to ₹7,000 over three weeks), then audience layers (Customer Match upload of 340 existing contacts, mobile bid adjustment of -40%, abandoned-inquiry remarketing list).
After six weeks: CPA dropped from ₹8,000 to ₹4,800. Monthly spend held at ₹2 lakh. Qualified lead volume increased from approximately 25 to 41 per month. The improvement came almost entirely from eliminating wasted spend — not from finding new audiences or increasing budgets.
Frequently Asked Questions
How many conversions do you need before switching to Target CPA bidding?
Google's actual recommendation is 30–50 conversions in the past 30 days before enabling Target CPA bidding. Below that threshold, the algorithm has too little signal to make reliable bid decisions and will oscillate between over-spending and under-delivering. If your account generates fewer than 30 conversions a month, run Maximise Conversions without a CPA target for another 4–6 weeks to build the data set, then introduce the CPA target constraint once you have sufficient history.
Should Indian businesses use Google Ads in English or regional languages?
English dominates search volume for B2B, IT services, and premium consumer categories across India — including in Kerala. Malayalam campaigns can be effective for hyperlocal services like groceries, repairs, or community events, but conversion rates in Malayalam search campaigns tend to be lower because transactional intent queries are still typed in English even by Malayalam-first speakers. The practical approach: run English as your primary spend and test a Malayalam ad group at 10% of budget for 60 days. If the CPA is within 20% of your English campaigns, scale it; if not, redirect that budget to English with tighter location bid modifiers instead.
What's a good Quality Score target for competitive Indian keywords?
Aim for 7 or above for most competitive terms — at score 7 you're paying roughly 28% less per click than an advertiser at score 5 with equivalent Ad Rank. For your own brand keywords you should consistently hit 9–10; anything below 8 on brand terms signals a landing page load issue or ad relevance problem worth investigating immediately. A score of 10 on non-brand keywords is rare in competitive Indian markets like IT services, real estate, or legal services — but achievable on long-tail, highly specific terms where you're the only credible bidder.