PERFORMANCE MARKETING & PPC AGENCY IN DELHI NCR

The one that agrees a measurable number with you before it starts, puts the ad accounts in your name, and can explain last month’s decisions in plain language. Rankings of agencies won’t tell you that.

The worry behind this question is real. Delhi has hundreds of agencies, many with confident claims and guaranteed numbers on their homepages, and plenty of founders have a story about a wasted year. So judge on evidence, not adjectives.

Ask to see a result for a business like yours, with the timeframe and how it was measured. Ask who will manage your account each week, and speak to them. Ask how they handle fake leads, and how they connect ad results to your CRM or store. Ask what happens to the accounts if you leave. And ask for a sample weekly report from a current client, with names removed.

Good answers are specific and slightly cautious. They talk about tracking, creative, testing and budget moves. Weak answers talk about guaranteed ROAS and how many leads they can get you.

My rule: choose the agency that explains its decisions, not the one that promises outcomes.

No, not honestly. ROAS depends on things no agency controls: the auction, your competitors’ budgets, your prices, your stock, your website and your sales team. A guarantee means someone is choosing a metric that makes the promise easy to keep.

You’ll see Delhi PPC pages advertising guaranteed ROAS, and I understand the appeal when you’ve been burned. Look closely at how they’d measure it. Platform-reported ROAS counts the conversions the platform claims credit for, which can include sales that would have happened anyway, and for Indian ecommerce it often ignores cash-on-delivery returns. A guarantee measured that way can be met while your bank balance says otherwise.

What you can ask for is a clear number agreed up front, measured from your own data, with a timeline. For ecommerce that’s usually return on spend after returns, or revenue against total spend. For lead generation it’s cost per qualified lead, defined with your sales team. Then ask for weekly reporting and a monthly review where decisions are explained.

My rule: a guaranteed ROAS is a reason to ask harder questions, not a reason to sign.

Make your forms harder to abuse and teach the platform what a real lead looks like. You can’t stop someone submitting junk, but you can stop paying to attract more of it.

This isn’t paranoia. Advertisers in Delhi NCR real estate have reported forms flooded with fake submissions, sometimes most of their leads. Bots and click farms hit cheap, frictionless forms hardest, and when Meta is told that any form fill is a success, it keeps finding more people like the junk.

Start with friction that filters. Use a higher-intent form type, add a qualifying question, and verify phone numbers. Watch for patterns, like bursts of submissions at odd hours or repeated numbers. Then connect leads to your CRM, mark which ones turned into real conversations, and send those outcomes back to Meta through the Conversions API. Optimise towards qualified leads rather than raw submissions.

If a channel stays polluted, move the conversation somewhere harder to fake: click-to-WhatsApp ads, calls, or landing pages with verification. Keep records of the suspicious activity and report it to Meta.

My rule: judge every campaign on cost per qualified lead. If you only count form fills, you’re paying for the junk.

Put three things in writing before you start: the number you’ll be judged on, who owns the accounts, and how you can leave. Most wasted years I hear about failed on those, not on talent.

A recent thread about Delhi NCR agencies captured it well: a year of fees, a lock-in, leads that went nowhere, and no clear way out. The pattern repeats because the contract rewarded activity, not results, and the client couldn’t see what was really happening inside the accounts.

So set it up differently. Agree one number up front, measured from your CRM or store. Insist the ad accounts, pixel and analytics are in your name with you as admin, and that spend goes on your card. Ask for a written plan for the first 30 and 90 days. Get a weekly report showing what changed and why. And keep the exit simple: a sensible first term, then a notice period, with all data and creative staying with you.

Then hold the review. At day 90, compare the agreed number with month zero. If it hasn’t moved and nobody can explain why, leave without guilt.

My rule: never sign anything you can’t walk away from with your data intact.

It ranges widely. Indian agency guides quote retainers from around ₹20K a month to ₹200K and more, plus the ad spend itself. The number on its own tells you very little.

The confusion comes from the label. “Social media marketing” can mean writing captions and posting, or it can mean running Meta and Instagram ads with creative testing, tracking and weekly budget decisions. Those are different jobs with different costs and very different effects on sales. Ask any agency which one they mean.

If your goal is sales or leads, you’re buying paid social, and the questions are the same as for any performance work. Who makes the creative, and how often is it refreshed? What does the campaign optimise for? How are results measured against your CRM or store? Who owns the accounts? Is the ad spend on your card or theirs?

The trap is paying for activity: a calendar of posts, a monthly deck, follower growth. None of that is bad, but none of it is the same as revenue.

We focus on paid social for performance. Every engagement starts with an audit, the retainer is scoped in writing, and ad spend stays on your own card.

My rule: decide whether you’re buying posts or customers before you compare prices.

Usually because the ad or the landing page trips Meta’s review on claims, targeting or missing details. Real estate gets closer scrutiny than most categories, and it’s usually fixable.

Advertisers in Indian real estate regularly report lead ads being rejected. The common causes I see: claims like guaranteed returns or assured appreciation, prices or offers that don’t match the landing page, landing pages with no clear business identity or contact details, and auto-generated creative full of exaggerated text. In some countries Meta also places housing ads in a special category with restricted targeting, so check what applies to your account before you build audiences.

There’s a legal layer in India too. Under the RERA Act, advertisements for a registered project are expected to carry its RERA registration number. Leaving it off is both a compliance risk and a trust problem with buyers.

When an ad is rejected, read the specific reason in Account Quality and fix that issue instead of resubmitting the same ad. Keep the copy factual: location, configuration, price band, possession timeline and the RERA number.

My rule: write property ads you’d be comfortable showing a regulator, and most rejections stop.

Yes, and it should be clearly separate. You pay the ad platform for media, ideally directly on your own card, and you pay the agency a fee for managing it. If those two blur together, you can’t see what you’re really paying.

The risk founders worry about is hidden markup. Some agencies pay Google or Meta themselves and send you one combined invoice, which makes it easy to inflate the media line. Others take a percentage of spend, which rewards them for spending more. Neither is automatically dishonest, but both make value harder to judge.

The clean setup is simple. The ad accounts are in your name and billed to your card, so the platform invoices you directly, with GST, and you can claim input tax credit. The agency invoices its fee separately, for clearly defined work: tracking, structure, creative, testing, budget decisions and reporting. You can see every rupee of spend inside the platform yourself.

That’s how we work. Ad spend stays on your card, and our fee is a separate, fixed retainer agreed in writing.

My rule: if you can’t see your exact media spend in your own ad account, ask why.

Give a new agency about 90 days to show the work and early signals, and up to six months to judge the full result. Less than that rarely gives the platforms enough data. Much more without progress is just hope.

The honest answer depends on volume. Ad platforms need data to learn. Meta talks about roughly 50 optimisation events per ad set in a week to leave its learning phase, and Google’s automated bidding needs steady conversions too. A high-volume ecommerce account learns in weeks. A B2B account with a few leads a week takes longer.

Here’s what I’d expect. By day 30: tracking fixed, account restructured, creative pipeline running, and a baseline agreed. By day 90: weekly tests decided, budget moved towards what converts, and your agreed number trending against month zero. By day 180: scaling into what your own data confirms.

The red flag isn’t slow results. It’s not being able to see what was done and why. If the reports are vague, act early.

My rule: judge the work at 30 days, the direction at 90, and the result at 180.