Most of the wasted spend on Google Ads in India comes from a handful of avoidable mistakes: no conversion tracking before launch, one giant campaign with 200 unrelated keywords, targeting all of India on a ₹25,000 budget, and switching to smart bidding before there's enough data for it to learn. Fix those four and a small budget starts behaving like a much bigger one.

The principles below apply whether you have ₹20,000 or ₹1 lakh per month. Budget changes the scale (how many campaigns, how wide the geography), not the fundamentals. This guide is structured so you can find the section for your exact budget band.

What Google Ads Actually Costs in India

Cost per click (CPC) is set by a live auction, so any "average CPC" number is only a loose starting point. It swings with your Quality Score, the exact keywords, your city, and how many competitors are bidding. Use ranges to sanity-check your plan, never as a guarantee.

As a rough guide, high-intent commercial keywords in competitive verticals (legal, insurance, real estate in metros, B2B SaaS) sit at the expensive end, often a few hundred rupees per click. Local services (plumbers, electricians, tutors) and general e-commerce keywords are usually cheaper. The only number that matters is your own CPC after two weeks of running, which you can read directly in the Keywords tab.

The practical implication: a ₹20,000/month budget at a ₹50 CPC buys roughly 400 clicks; at ₹150 CPC it buys about 130. If your typical landing-page conversion rate is 3-5%, 130 clicks may produce only 4-6 leads in a month. That math should decide your strategy before you build anything. If the numbers say a single keyword theme is all you can afford, run only that.

How to Split the Budget Across Campaign Types

These splits are starting points, not rules. The logic: at small budgets you cannot afford to spread thin, so put almost everything into Search where intent is highest, and reserve a slice for remarketing because it is the cheapest conversion you will get.

  • ₹20,000-40,000/month: roughly 70% Search, 20% Performance Max, 10% Remarketing. Honestly, at the very bottom of this band you can skip Performance Max entirely and run Search plus Remarketing only.
  • ₹40,000-80,000/month: roughly 60% Search (brand + non-brand split), 20% Performance Max, 15% Remarketing, 5% Display only if you have a real awareness reason.
  • ₹80,000-1 lakh+/month: roughly 50% Search across multiple campaigns by intent and geography, 25% Performance Max, 15% Remarketing, 10% YouTube only if you already have video assets.

One caution on Performance Max: it is a black box and it will happily spend on cheap, low-quality placements that look like conversions but are not. On small budgets, keep it small until your Search campaigns are profitable and you have clean conversion data feeding it.

Setting Up Campaigns That Convert

Campaign Structure

The most common small-budget mistake is one campaign called "Services" holding 200 keywords. You lose all bid control and your most expensive keyword drains the budget meant for your best one.

Structure by intent theme instead. For a website-development startup in Bengaluru on ₹30,000/month, a workable split:

  • Campaign 1, High-intent local (~₹15,000): "website development bangalore", "web design company bangalore". These convert best, so they get the most budget.
  • Campaign 2, Broader discovery (~₹10,000): "website development india", "web design agency india". Wider, less qualified, lower bids.
  • Campaign 3, Brand terms (~₹5,000): your own brand name. Cheap, very high converting, and it stops competitors from showing above you on your own name.

Separating these lets you set bids by intent. A local high-intent search deserves a higher bid than a vague national one, and you cannot do that if they share a campaign.

Keyword and Match-Type Strategy

On a small budget, lean toward phrase match and exact match, not broad match. Broad match on a tight budget routinely burns money on loosely related searches before you can react. Add exact-match high-intent terms (city + service, "[service] price", "[service] near me") and let phrase match catch close variants.

Negative keywords are not optional. Add these to every campaign on day one, before a single click:

jobs
job
salary
vacancy
career
internship
intern
free
cheap
tutorial
how to
course
pdf
download
sample

Then check the Search Terms report (Insights and reports → Search terms) every few days for the first month and keep adding irrelevant terms. The Search Terms report shows what people actually typed, which is different from your keywords. This is where most wasted spend hides.

Writing Responsive Search Ads

Each Responsive Search Ad lets you add up to 15 headlines and 4 descriptions, and Google mixes them. Give it raw material to test, but keep every line truthful. Claims you cannot back up ("50+ brands trust us", "money-back guarantee") may help an early click but they damage trust, can trigger disapprovals, and waste the click if the landing page does not deliver.

A reliable headline structure:

  • Some headlines that include the exact keyword (so the ad matches the search and helps Quality Score), e.g. "Web Design Company in Bengaluru".
  • Some headlines with a concrete, true differentiator: your real turnaround time, a real specialisation, a genuine local presence.
  • Some headlines with a clear call to action: "Get a Free Quote", "Book a Call", "WhatsApp Us Today".

Pin a headline only when you need a specific line to always appear (for example a required disclaimer). Over-pinning kills Google's ability to test combinations. Add at least two or three genuinely different descriptions and let the system learn which combination earns clicks.

If you want to use a number in an ad ("X projects delivered", "X years in business"), use your real number or do not use one. Made-up proof is exactly the kind of thin, untrustworthy content that backfires.

Track Conversions Before You Spend a Rupee

This is step zero, not an optimisation step. Without conversion tracking you are flying blind: you might be buying 300 clicks and zero customers and have no way to know. Set this up before the campaign goes live.

What to track for a typical Indian business:

  • Website form submissions (contact, quote request, callback)
  • Phone calls (call-from-ads and calls from the website)
  • WhatsApp clicks: for most Indian SMBs this is the primary lead action, so treat it as a real conversion, not an afterthought
  • Purchase completions (e-commerce)

The cleanest setup without a developer: install Google Tag Manager, add your Google Ads tag and GA4 through it, and fire conversion triggers on the thank-you page or button click. On WordPress, the official Site Kit by Google plugin connects Search Console, Analytics and Ads in a few clicks. Turn on Enhanced Conversions in Ads (Goals → Conversions → Settings). It uses hashed first-party data to recover conversions that cookie loss would otherwise drop, which matters more every year.

For a WhatsApp click-to-chat button, the click itself is the event you tag. A typical link looks like this, with the number in international format and no plus or spaces:

<a href="https://wa.me/919876543210?text=Hi%2C%20I%20saw%20your%20ad%20and%20want%20a%20quote">
  WhatsApp Us
</a>

In Tag Manager, add a click trigger on that link and map it to a Google Ads conversion. Now WhatsApp leads show up next to form leads, and smart bidding can actually optimise toward them.

Bidding: Earn the Right to Automate

Smart bidding (Target CPA, Target ROAS) is genuinely good, but it needs data to learn. Google's own guidance is to give automated strategies a meaningful conversion history before judging them; a common working threshold people use is roughly 15-30 conversions in the trailing 30 days. Below that, the algorithm is guessing.

  • ₹20,000-40,000/month: Start on Manual CPC or Maximize Conversions with a tight budget while you gather data. Manual CPC gives you control when every click counts. Move to Target CPA only once you are consistently logging conversions weekly.
  • ₹40,000-80,000/month: Once you have a steady conversion volume, switch to Target CPA. Set the target from real economics: if a customer is worth ₹5,000 to you and you want to keep ~30% margin, a target CPA around ₹1,500 is a sane starting point; then adjust based on what actually happens.
  • ₹80,000-1 lakh+/month: For e-commerce with transaction values flowing into Ads, Target ROAS works (want 3x return, set 300%). For lead-gen services, Target CPA is usually the better fit because "value" is harder to pass back accurately.

Whatever you pick, do not change bid strategy and budget on the same day, and do not touch it during the "Learning" status. Every major change resets learning and wastes spend.

Reading the Search Terms Report

After two to four weeks, the Search Terms report tells you what to do next:

  • Terms that converted: add them as exact-match keywords and consider raising bids.
  • Terms with clicks but no conversions: tighten match type, or pause once they cross roughly 20-30 clicks with zero conversions. On a small budget, be ruthless here.
  • Irrelevant terms: add as negatives immediately.
  • Long-tail terms that worked: these often convert better at lower CPC. Promote them into their own tightly themed ad group.

Common Mistakes Indian Startups Make

Targeting all of India on a small budget

Targeting the whole country on ₹25,000/month spreads spend so thin that no location gets enough data to optimise. A web-design startup in Pune should win "web design pune" and nearby metros first, prove it converts, then expand outward. Set location targeting to "Presence: people in or regularly in your targeted locations" rather than the default that includes people merely interested in your area. The default quietly serves your ads to other countries searching about India.

Ignoring Quality Score

Quality Score (1-10) directly affects how much you pay and where you rank. A high score can win a better position for less money than a competitor with a weak score. You can read the components (Expected CTR, Ad Relevance, Landing Page Experience) in the Keywords tab by adding those columns. Improve it by keeping ad groups tightly themed (one intent per ad group), echoing the keyword in the ad, and making sure the landing page matches the ad's promise and loads fast on mobile. See Google's Quality Score documentation for the exact definitions.

Skipping ad assets (extensions)

Ad assets (formerly extensions) make your ad physically larger and give people more reasons to click, at no extra cost per click. Add at minimum:

  • Sitelinks: deep links to Services, Pricing, About, Contact.
  • Callouts: short, true benefit phrases ("Free Consultation", "Local Team", "GST Invoice Provided").
  • Structured snippets: list your actual service lines or product types.
  • Call asset: essential for local service businesses where people phone before they buy.
  • Location asset: link your Google Business Profile so the ad can show your address and map pin for local searches.

Neglecting remarketing

Most visitors do not convert on the first visit. Remarketing reaches the ones who showed interest and left, and because that traffic is already warm it is usually your cheapest conversion. Build these audiences in the Audience Manager:

  • All visitors, last 30 days
  • Visitors of specific service or product pages
  • Cart abandoners (e-commerce)
  • Past customers (for upsell, kept separate so you do not pay to re-acquire them)

Cap frequency to a sensible level so the same person is not seeing your ad ten times a day. Note that smaller remarketing lists can take time to reach the minimum size required to actually serve, so do not panic if a brand-new list shows no impressions on day one.

Strategy by Budget Band

₹20,000-30,000/month

Do one thing well. Pick the single keyword theme most likely to produce business and concentrate there. For a coaching institute that might be "[exam] coaching [city]" or "[subject] tuition near me".

  • One Search campaign, 20-30 tightly grouped keywords, phrase/exact match.
  • One remarketing campaign on last-30-day visitors.
  • Manual CPC, watched closely.
  • Goal for month one: gather a real conversion history before expanding anything.

₹30,000-60,000/month

Now you can separate intents and protect your brand.

  • One Search campaign for core high-intent keywords.
  • One Search campaign for broader discovery terms.
  • One remarketing campaign.
  • A brand campaign if competitors might bid on your name.
  • Move to Target CPA or Maximize Conversions once volume supports it.

₹60,000-1 lakh/month

Structure by service line and, if you serve several cities, by geography.

  • Brand campaign.
  • Non-brand Search campaigns split by product or service line.
  • One Performance Max campaign, fed by clean conversion data.
  • Multiple remarketing audiences.
  • YouTube only if you genuinely have video assets. Do not start making videos just to spend budget.

Which Metrics Actually Matter

Pick metrics by business model, and always tie them back to revenue, not vanity clicks:

  • Lead-gen services: Cost per lead, then lead-to-customer rate, then true cost per acquisition (CPA = cost per lead ÷ close rate). A cheap lead that never closes is not cheap.
  • E-commerce: ROAS and CPA measured against average order value, plus conversion rate by campaign.
  • SaaS/Tech: Cost per trial, cost per demo request, and cost per qualified lead (MQL): track the step that predicts revenue, not just sign-ups.

Scale a keyword or campaign when it is consistently beating your target CPA with a healthy conversion count and a stable Quality Score, and you can actually service more customers. Pause or rework when CPA runs well over target for a couple of weeks straight, a keyword piles up clicks with no conversions, or Quality Score sits stuck at the bottom. Give each decision two to four weeks of data. Week-one numbers are noise.