A skincare brand in Bandra was spending ₹80,000 a month on Google Ads. They were getting enquiries, but the cost per lead was ₹2,400 and they couldn't scale without losing money. Three months earlier they'd spent ₹1,20,000 and gotten fewer leads. Their previous agency's response: "Google Ads is getting more expensive, you need to spend more." That's not an answer. That's an excuse.
I spend two hours on their account. Cost per lead drops to ₹720. Same budget. Same product. What changed: the account structure was a mess, 60% of the spend was going to irrelevant searches, and the landing page had a 12-second load time on mobile. None of these were complicated problems. No one had looked.
Here's the exact process I use for a 2-hour Google Ads optimisation engagement.
Hour 1 — Diagnosis: finding where the money is going
I start with the search terms report, not the keywords tab. This is the most common mistake I see — people look at their keyword list and think they understand what their ads are matching. The search terms report shows you what people actually typed into Google before clicking your ad. These are not the same thing.
I export the search terms report for the last 90 days, filter to terms with at least ₹500 in spend, and sort by cost. Then I paste this into Claude with a one-paragraph brief about the business — what they sell, who their customers are, what cities they serve. I ask Claude to classify each term and draft a negative keyword list. This takes 5 minutes and typically surfaces 20–40 obviously irrelevant terms that have been burning budget for months.
Next: conversion tracking audit. I check whether conversions are tracking correctly — phone call clicks, form submissions, purchase events. In at least 40% of accounts I audit, conversion tracking is broken or misconfigured. If you're optimising for conversions and the conversion data is wrong, everything you do makes the account worse. Check this first.
Then I look at the quality scores. Keywords with quality scores of 3 or below are costing 2–3x more per click than they should. The fix is almost always one of two things: the keyword is irrelevant to the landing page, or the landing page is generic and doesn't match the specific keyword intent. Both are fixable.
Quick wins that take 20 minutes each
Negative keywords: Take the list Claude generated. Add them to a shared negative keyword list applied to all campaigns. Done. Estimated budget saving: 15–30% in most poorly-managed accounts — immediately.
Ad schedule: Pull the "Day and hour" performance data. Look for time slots with high spend and zero conversions consistently. A restaurant in Koramangala running ads at 2am. An immigration consultant running ads on Sunday when their office is closed. Turn those off. Set bid adjustments down on low-performing hours.
Location targeting: Check the locations report. Are you paying for clicks from cities you don't serve? A dentist in Andheri who added "Mumbai" as a location target is often paying for clicks from Navi Mumbai, Thane, and Mira Road — places they'd never get patients from at that distance. Tighten the radius or add location exclusions.
Device bid adjustments: Check conversion rates by device. If mobile converts at half the rate of desktop (common for high-consideration services like legal advice or B2B software), set a mobile bid adjustment of -25% to -40%.
Hour 2 — Structure and creative improvements
Campaign structure: I check whether the campaign is organised by theme, intent, or buying stage. Most small business accounts have one campaign with 50 ad groups or one campaign with 3 ad groups and 80 keywords in each. Neither is good. The goal is tight theme clusters — each ad group covers one specific topic, the ads in that ad group directly reference that topic, and the landing page directly matches that topic. Relevance all the way down.
Ad copy audit: I pull the current ads and look at what's actually running. For Responsive Search Ads, I check the asset performance ratings (Google shows you which headlines perform well vs poorly). I remove "Low" rated headlines and descriptions. Then I use Claude to generate 15 new headline options and 4 description options — I give it the top-performing existing headlines as style references, the business brief, and the target keyword themes. I review, select the strongest, and add them.
Landing page speed: I run PageSpeed Insights on the landing page the ads point to. If it's below 60 on mobile (which is shockingly common), I flag it as the highest-priority item in my recommendations, because no amount of ad account optimisation overcomes a landing page that takes 8 seconds to load on a Jio connection. This is often a separate engagement, but it's the single biggest conversion improvement available.
What I deliver at the end of 2 hours
A written audit summary with: the specific changes made (negative keywords added, bid adjustments set, bad assets removed), the estimated impact on cost per click and conversion rate, and a ranked list of further recommendations prioritised by impact vs effort. Clients get this as a Notion doc or a clean PDF. It's the kind of documented work that makes them want to retain you for ongoing management rather than treat the engagement as a one-off.
The going rate for this audit in India, from someone who can actually do it, is ₹15,000–35,000. That's 2 hours of work. The value to a client spending ₹1 lakh a month on ads — even a 15% reduction in wasted spend is ₹15,000 per month saved, every month. The audit pays for itself immediately. That's how you justify your rate without having to negotiate down.
What the client relationship looks like after the engagement
A good Google Ads audit creates a natural ongoing retainer. The issues I identified in the 2-hour session point toward work that takes longer — restructuring campaigns properly, A/B testing new ad copy variations, building out keyword lists methodically, optimising landing pages. That's where I try to convert a project engagement into a monthly relationship.
My pitch for ongoing management, if the audit produces clear results, is simple: "The account has stronger bones now, but this work continues to need attention. I can manage it monthly — the same level of attention, applied consistently — for ₹15,000–25,000/month. The alternative is you manage it yourself, which you now know more about than before, or you hire another agency." Some say yes. Some say they'll manage it themselves. The ones who manage it themselves usually come back in 3 months when they've made it worse again. Both are fine outcomes from my perspective.
The skill of converting a project to a retainer is the skill that creates stable freelance income. A project pays once. A retainer pays every month and gives you runway to plan. After the bottleneck-finding and audit approach is working, building retainer income on top of it is the highest-impact thing you can do for the business.
The conversion pitch that has worked most consistently: "The audit found ₹X in recoverable spend. Managing the account to actually recover that costs ₹Y per month. The break-even is [Z months]." When X is much larger than Y and Z is under 2, the client does the arithmetic themselves. You don't have to persuade — you just have to show them the numbers clearly.
Also see: Meta Ads vs Google Ads for Indian local businesses and How I find bottlenecks and build web apps that fix them.