Google Ads services
Google Ads is a set of very different advertising products sharing one interface. Which of them belongs in your account depends on what you are trying to achieve — qualified enquiries, ecommerce sales, or demand for something people are not yet searching for.
I work with businesses on that decision and everything that follows it: how campaigns are structured, what counts as a conversion, whether the numbers can be trusted, and when increasing spend is justified. That covers accounts serving customers in India and in overseas markets, where competition and cost per click behave very differently.
Goal decides the campaign — not the reverse
Anyone can spend a budget on Google. The platform will deliver clicks reliably from the first day. What is not automatic is whether those clicks come from people who might plausibly buy, whether the resulting enquiries are worth having, and whether the cost of acquiring one leaves anything behind.
Most of the consequential decisions happen before a single ad is written: which searches to compete for, what action to optimise towards, how campaigns are grouped so the system has enough data to learn from, and what the account is allowed to pay for a customer. Ad copy matters, but it is downstream of all of that.
Traffic volume is the easiest metric to move and the least useful on its own. Broadening match types or lowering bids will raise clicks in a week, and a business can double its paid traffic while its enquiry count stays flat.
The metrics that matter sit further down: how many of those visits became meaningful business actions, what each one cost, and whether the customers behind them were profitable. That chain — from search, through click, through conversion, to revenue — is what a Google Ads engagement should be accountable for. Everything above it is a means.
Accounts usually get built around whichever campaign type the person setting them up is most comfortable with. The better order is objective first, then how customers behave around that objective, then what action represents value, and only then which campaign format serves it.
Need more qualified leads?
Start where intent already exists. Someone typing a service and a problem into Google has declared what they want; the job is being present for the searches that indicate a real buyer and absent for the ones that only look similar. Lead volume is the easy part — lead quality is where accounts are won or lost, and that depends on which queries you allow.
Need more ecommerce sales?
Product-led advertising works from your catalogue rather than a keyword list, which shifts the centre of gravity from ad copy to product data. Price, availability and how products are described all become advertising inputs, and campaigns can only promote what the feed describes accurately.
Need to build demand?
Some products solve problems people have not named yet, so nobody is searching. Video and feed-based formats reach those audiences before intent exists. Expect slower results and harder attribution — this is an investment in future demand, and judging it on last-click conversions will always make it look like a failure.
Need to re-engage visitors?
Most first visits do not convert, and re-engaging people who already showed interest is the cheapest conversion available. It is also the most over-credited: this activity harvests demand the rest of the account created, so it should be measured separately rather than allowed to flatter the average.
Each campaign type has one role it performs well. The mistake is not choosing the wrong one — it is expecting one campaign to do a job it was never designed for, then concluding the platform does not work.
| Campaign type | Best suited for | Primary role | Typical business objective |
|---|---|---|---|
| Search | People actively searching for your products or services | Capture existing demand | Leads, sales, qualified traffic |
| Shopping | Online retailers with a product catalogue | Product-led demand capture | Ecommerce sales |
| Performance Max | Businesses with measurable conversion goals and reliable data | Goal-based optimisation across Google's channels | Sales, leads and other supported goals |
| YouTube | Brands with video that can hold attention | Reach and consideration | Awareness, views, consideration and supported conversion goals |
| Display | Visual reach and re-engagement across websites and apps | Display advertising | Awareness, consideration and remarketing |
| Demand Gen | Creating interest before people start searching | Demand creation and consideration | Awareness, consideration and conversions |
| Remarketing | Previous visitors and engaged audiences | Re-engagement | Return visits, leads or sales |
| Local Service Ads | Eligible local service businesses in supported categories | Local lead generation | Qualified enquiries |
Two rows deserve a caveat. Performance Max works across Google's inventory with limited manual control, so it depends heavily on the quality of what you feed it — conversion data, product information and creative assets. And Local Service Ads are only available in certain categories and locations, so eligibility has to be confirmed rather than assumed. Most accounts end up running two or three of these together, with each doing its own job rather than competing for the same clicks.
The question a campaign has to answer is not whether it can generate conversions, but whether it can generate them at a price the business can afford to pay. These five variables determine that, and four of them live outside the ad account.
The relationship between them is what makes benchmark figures unhelpful. A target ROAS or cost per lead quoted without margin context describes nothing: a 30% margin business and a 70% margin business selling at identical prices can support acquisition costs that differ by more than double.
This is also the most common reason a technically well-run account disappoints. The campaigns do what they were told, the reported metrics look respectable, and the business still cannot find the profit — because nobody established what a customer was worth before deciding what to pay for one.
Different objectives need different measures, and each list below runs from the metric the platform reports most readily down to the one the business actually cares about. The further down you can measure, the better the decisions get.
The gap between leads and qualified leads is where most lead-gen accounts are misjudged. Feeding sales outcomes back from a CRM is what closes it.
Return on ad spend measures revenue against spend and knows nothing about cost of goods, shipping or returns — which is why the last line matters more than the third.
This is the hardest group to measure honestly. Demand creation rarely converts on last click, so holding it to the same standard as Search guarantees it looks like waste.
Underneath all three sits the same requirement: the conversions being counted have to correspond to things that actually happened. Google Ads, Google Analytics 4, Tag Manager and your CRM or ecommerce platform each hold part of that picture, and they will not agree exactly. Reconciling them against real orders or enquiries — rather than picking whichever number is highest — is what makes the rest of the reporting worth reading.
More budget applied to a working account produces more results. Applied to one with an unresolved problem, it produces the same problem at greater cost — and adds pressure to justify the outcome. Six things worth confirming first.
Do the recorded conversions represent actions that matter, and does the count match what the business received? Automated bidding spends against these signals, so an inflated or duplicated conversion does not merely misreport — it directs money towards whatever produced it.
Read the search terms report, not the keyword list. The queries actually being matched are frequently broader than intended, and scaling budget scales that mismatch proportionally.
Does the page confirm what the ad promised within a few seconds? A conversion rate problem at the page level is the cheapest thing on this list to fix and the most commonly attributed to the campaign instead.
Budget split across too many campaigns leaves each with too little conversion data for the bidding system to learn from. Fragmentation feels like control and often costs performance.
Know what a lead or a customer is worth before committing more money to acquiring them. Without that figure there is no way to tell whether a rising cost per acquisition is a problem or an acceptable trade for volume.
Product campaigns advertise what the feed describes. Missing attributes, stale prices and disapproved items all quietly limit what can be shown, which reads as poor campaign performance rather than a data problem.
The platform is the same for everyone. What changes is which campaign types earn their place, what counts as a conversion, and how long the feedback loop takes between spend and evidence.
Advertising runs on catalogue data, so the product feed becomes an advertising asset rather than an operations detail. Transactions are measurable and the feedback loop is short, which makes optimisation faster — but it also means an average that looks healthy can hide products selling below their true cost. That work has its own page: Google Ads for Ecommerce.
High-intent search is usually the whole opportunity, and volume is limited by how many people need the service this month. That makes lead quality the dominant variable — a service business can afford a higher cost per enquiry than it expects, provided the enquiries are from people who can actually be served profitably.
Long buying cycles break the normal feedback loop: the click and the closed deal can be months and several people apart. Optimising towards form fills alone tends to attract researchers rather than buyers, so the account needs sales outcomes fed back from the CRM to learn what a good lead looks like.
Proximity does much of the qualifying, and the searchable market is finite. Geographic targeting, location-aware ad copy and eligibility for local advertising formats matter more than campaign sophistication — there is only so much demand within a serviceable radius, and the goal is capturing it rather than expanding it.
The first job is not growth but learning whether the acquisition maths works at all. Small, deliberately structured tests answer that faster and far more cheaply than scaling on optimism. Getting a defensible cost per customer before raising spend is the difference between a channel and an expense — more on Google Ads for Startups.
Each campaign type has a dedicated page covering how it is set up, structured and improved.
Six causes that produce the same complaint. They need different responses, which is why "Google Ads isn't working" is never one problem.
Five stages that repeat rather than finish. The last one feeds the second, which is what separates an advertising strategy from a launch.
The business model, how customers actually decide, what the offer is worth and which commercial objective the advertising serves. This is not preamble — every later decision depends on it.
Campaign architecture, measurement design and acquisition targets derived from the economics, agreed before anything is built.
Implement the strategy in a new account, or restructure an existing one where the foundations are sound enough to keep.
Compare advertising data against outcomes the business recognises, and treat disagreements between the two as findings rather than noise.
Decide what changes and what scales on the evidence — then return to planning, because the first plan was built on assumptions that real data has now replaced.
Seventeen years in digital marketing, working across ecommerce catalogues, service businesses, B2B accounts with long sales cycles, and local businesses — in the Indian market and for overseas markets including the UK, US, Canada and Australia.
The useful part of that range is not the count. It is that the same tactic behaves differently depending on where it is applied, and having seen both cases makes the difference obvious rather than theoretical. Broad match on an ecommerce account with thousands of weekly conversions is a reasonable bet, because the system has enough data to learn quickly. The same setting on a B2B account generating twelve leads a month is a way to spend a quarter's budget discovering what you already suspected.
International accounts sharpen the same point. Cost per click, competitive density and buyer expectations differ enough between the Indian market and the UK or US that a structure which works well in one can be uneconomic in the other — and a target imported from the wrong market will make a healthy account look like a failing one.
What that experience is for is judgement about which decisions matter for your situation. I work on accounts directly rather than through a team, so the person forming that judgement is the one you can question about it.
Google Ads buys visibility immediately, which is its advantage and its risk. It will expose an unresolved problem faster and more expensively than any other channel.
None of the second column is permanent, and most of it is a few weeks of work rather than a reason to abandon the channel. The mistake is starting spend while they are outstanding and concluding from the result that Google Ads does not work for your business.
Tell me what you are trying to achieve and where the advertising currently sits — more qualified enquiries, better returns on product campaigns, or a first campaign you have not launched yet. It helps to know what you are running now, what a customer is worth, and what is not working. I will tell you what I would change and whether Google Ads is the right place to start.