The most common reason a Meta Ads account underperforms isn't the platform — it's the setup. Three mistakes account for most of the wasted spend we see when we audit a new account.

First, targeting that's too broad. "Everyone in Bengaluru aged 18-65" isn't an audience, it's a guess. Every campaign should start from a specific problem a specific person has, and the targeting should follow from that.

Second, no creative testing. Running one ad creative until it gets tired is the single biggest source of rising costs we see. We typically run 3-4 creative variations per campaign and let the data decide, not our opinion of which one "looks better."

Third, boosting posts instead of running campaigns through Ads Manager. A boosted post optimises for engagement. A properly built campaign optimises for the outcome you actually want — leads, calls, or purchases. They are not the same tool, even though they look similar.

Fix these three and most accounts see cost-per-lead drop within the first couple of weeks of a rebuild.

One more mismatch worth checking: campaign objective versus what you actually want. A campaign optimised for "traffic" will happily bring cheap clicks from people who never intended to buy anything — it did exactly what you told it to do. If the goal is leads or sales, the campaign objective needs to say so explicitly, or the algorithm optimises for the wrong outcome by design, not by accident.

Audience overlap is a quieter budget leak. Running five ad sets that all target roughly the same people means those ad sets end up bidding against each other in the same auction — you're competing with yourself and paying more for it. Checking for overlap before launching a new campaign, not after costs climb, saves real money.

Creative fatigue is real and underestimated. The same ad shown to the same audience for weeks stops earning attention, and cost-per-result quietly climbs even though nothing else changed. A rough rule of thumb: if an ad has been running more than three to four weeks without a refresh, it's worth testing a new version before assuming the audience or offer is the problem.

Most small businesses hear "SEO" and think of hundreds of blog posts and backlinks. Start somewhere much simpler: your Google Business Profile.

For any business with a physical presence or a local service area, an optimised Google Business Profile — accurate categories, complete business hours, real photos, and a habit of responding to reviews — usually drives more qualified traffic than months of blog writing.

After that, on-page basics: does every page have a unique title tag and meta description that actually describes what's on it? Does the site load in under three seconds on mobile? These sound boring, but they're the foundation everything else sits on.

Only after both of those are solid does content strategy start to matter — and even then, one genuinely useful page that answers a real question your customers ask beats ten generic ones every time.

SEO is a compounding channel, not a quick win. Expect the real movement to show up around month three to six, not week two.

Treat any agency promising "page 1 rankings in 30 days" as a warning sign, not a selling point. Search engines reward consistency and genuine relevance over time; anything promising to shortcut that process is either describing paid ads (a different channel entirely) or using tactics that risk a penalty down the line.

Before touching content, most small businesses have easier wins sitting unclaimed: a fully filled-out Google Business Profile, with accurate hours, real photos, and a habit of responding to reviews, often moves the needle on local search faster than a blog post will — because it's exactly what Google shows first for "near me" searches.

It also helps to separate two different jobs that both get called "SEO." Technical SEO is making sure a site can actually be crawled and loads fast — the plumbing. Content SEO is having pages that genuinely answer what people are searching for. A technically flawless site with nothing worth reading won't rank, and neither will great content sitting on a site search engines can't properly read.

A property launch generating hundreds of leads means nothing if the sales team can't work through them fast enough to matter. The gap between "leads generated" and "units sold" is almost always the nurture journey, not the ad spend.

The campaigns that actually convert pair paid lead generation with a structured follow-up sequence — email, WhatsApp, SMS, and an RM call, timed so a lead never goes more than a day or two without a touchpoint in the first two weeks.

Visual quality matters more in real estate than almost any other category. Drone footage of the actual site, walkthrough videos, and real (not stock) photography consistently outperform generic renders in both engagement and lead quality.

And launches benefit from genuine urgency — real price-appreciation milestones, real unit availability — used honestly. Manufactured scarcity gets noticed and erodes trust fast in a category where trust is the entire sale.

Retargeting is where a lot of otherwise-good campaigns leave money on the table. Most site visitors don't convert on their first visit to a project page — they're comparing options. A retargeting sequence that follows up with the specific project or unit type they looked at converts meaningfully better than sending everyone back into the same generic top-of-funnel ad.

Video walkthroughs have moved from "nice extra" to expected. A buyer deciding whether a property is worth an in-person visit increasingly checks for a walkthrough video first — listings without one get skipped over in favour of ones that show the space properly, even if the underlying property is comparable.

Pricing transparency in the ad itself, not just on a landing page three clicks deep, filters for serious buyers rather than casual browsers. It feels counterintuitive to lead with price, but hiding it just delays the same conversation and wastes a call on someone outside the actual budget range.

Two numbers matter more than any other metric in a report: cost-per-lead, and what happens to that lead afterward. Everything else is context.

Impressions and reach tell you how many times an ad was shown — useful for diagnosing a campaign, not for judging whether it worked. A campaign can have excellent reach and terrible results, and vice versa.

Cost-per-lead tells you what you're actually paying to get someone interested. Conversion rate — how many of those leads become customers — tells you whether the leads were any good in the first place. A cheap lead that never converts is more expensive than an expensive one that does.

If a report leads with likes, followers, or reach, ask for the leads and cost-per-lead numbers instead. That's the conversation that actually matters.

One nuance worth knowing: a lead credited to "Google" in a report often saw two or three other touchpoints first — an Instagram ad, a WhatsApp message, a friend's recommendation — before searching and clicking. Single-channel attribution is a simplification, not the full story, which is exactly why channels should be judged as a system working together, not ranked against each other in isolation.

Impressions and reach without context are close to meaningless on their own — a number can look impressive while telling you nothing about whether the right people saw it or whether any of them cared. Ask what happened *after* someone saw the ad, not just how many people saw it.

Return on ad spend (ROAS) is worth understanding even in rough terms: it's simply revenue generated divided by what was spent to generate it. A campaign with a 4x ROAS returned four rupees of revenue for every rupee spent. It's not a perfect metric — it doesn't account for margin — but it's a far more honest conversation starter than "how many people saw our ad."

Google's own research has long shown that bounce rate climbs sharply once a mobile page takes more than a few seconds to load — and most visitors never get past that first impression to see how good the rest of the site actually is.

The second problem is what greets people once the page does load. If the first screen doesn't make it obvious what you do and what to do next, visitors decide within seconds whether to stay or leave — and most of that decision happens before any scrolling.

The third problem is designing for desktop first. The majority of traffic for most small businesses now arrives on a phone, so a site that looks great on a laptop but cramped or slow on mobile is optimising for the wrong device.

Fast load times, a clear first screen, and mobile-first design aren't nice-to-haves — they're the difference between a visitor who explores and one who's already gone.

Trust signals matter more than most redesigns account for. A visible phone number, real testimonials, and clear proof of who's behind the business all reduce the hesitation a first-time visitor feels before taking action — often more than another round of visual polish would.

Form friction quietly kills more conversions than bad design does. Every additional required field is another small reason to abandon — asking for name, phone, and what they need is usually enough to start a conversation; asking for a company size, budget range, and three other details up front feels like an interrogation before any trust has been built.

A page trying to do everything at once — sell a service, collect an email, promote a blog post, and push a phone call, all above the fold — usually accomplishes less than a page with one clear, obvious next step. Pick the single action that matters most on that specific page and make everything else secondary.

The first frame does most of the work. On a feed where people scroll fast, an ad has a fraction of a second to earn a second look — which means the opening image or first second of video matters more than almost anything else in the creative.

Polished, obviously-staged stock photography tends to read as an ad and get scrolled past. Creative that looks like it belongs in the feed — real settings, real product use, less "shot on a professional set" — consistently earns more attention.

Contrast and clarity beat cleverness. A message that's instantly readable at a glance outperforms a beautifully designed layout that takes a second to parse, because that second is rarely given.

The agencies that consistently win aren't the ones with one perfect "hero" ad — they're the ones testing several creative directions at once and letting performance data decide, rather than personal taste.

Design for sound off. The large majority of video ads are watched muted on a feed, at least initially — if the message only lands through voiceover or dialogue, most viewers miss it entirely. Captions and a visual hook that works without audio aren't an afterthought; they're the difference between an ad that communicates and one that just plays.

Showing the result in the first frame, rather than building up to it, consistently outperforms the "slow reveal" structure ads are often storyboarded around. Feeds don't reward patience — an ad that opens on the transformation, the finished product, or the outcome gives people a reason to keep watching instead of asking them to earn it.

Testimonial-style ads that look like they came from a real customer, not a studio, routinely beat polished brand-produced spots on cost per result. The rougher, more "someone actually filmed this on their phone" look isn't a budget compromise — for a lot of products, it's the better-performing creative choice on its own merits.

Creator-style UGC (user-generated content) ads — the ones that look like a real person talking to camera rather than a polished commercial — are consistently outperforming traditional brand video in most feed placements right now, largely because they don't read as advertising at first glance.

AI tools are genuinely useful here, but mostly in the unglamorous parts: faster scripting, quick voiceover variations, and cutting iteration time from days to hours. Used well, AI speeds up the process of testing more creative ideas, not replacing the idea itself.

The mistake we see most is treating "AI video" as the strategy rather than the tool. An AI-generated ad with no clear hook, offer, or audience in mind performs no better than a bad ad made the old way — it's just faster to produce.

The winning combination is usually a real creative strategy, executed faster with AI-assisted production, not AI replacing the strategy itself.

Authenticity is also becoming a genuine differentiator as AI content gets more common in every feed. Audiences are getting better at sensing when something feels synthetic, and brands that stay transparent about what's real footage versus AI-assisted tend to keep more trust over time than ones that blur that line.

In practice, the useful AI workflow looks less like "generate a finished ad" and more like a set of accelerators bolted onto a normal creative process: faster first-draft scripts to react to and edit, quick voiceover variations to test tone, and rapid iteration on a hook before committing to a full shoot or edit.

Human review before publishing isn't optional. AI-assisted output still needs a person checking that the claims are accurate, the tone matches the brand, and nothing reads as slightly off in a way an audience would notice even if they couldn't say exactly why.

Trying to run every platform at once, with the same content on each, is one of the most common ways businesses waste a social media budget. Different platforms serve different intents, and content built for one rarely performs as well when copy-pasted to another.

Instagram rewards visual, lifestyle-driven content and works best for brands people want to browse. LinkedIn is built for B2B trust-building and works when the goal is credibility, not virality. YouTube rewards genuine explanation and demonstration — it's where long-consideration purchases get researched.

The better question isn't "which platforms should we be on," it's "where does our actual audience already spend time, and what are they there to do." Two platforms done well consistently outperform five done half-heartedly.

In the Indian market specifically, WhatsApp is consistently underused as a marketing channel despite converting better than almost anything else once someone's already interested — a click-to-chat ad or catalog message often closes a sale that a form submission never would, simply because it meets people where they're already comfortable talking.

It also helps to think in terms of funnel stage, not just platform. Instagram and short-form video tend to do the heavy lifting for awareness — introducing people to a brand they didn't know existed. Google and search-driven channels tend to catch people already close to a decision, actively looking for a solution. Treating every platform as if it should do the same job leads to judging a channel by the wrong metric.

Before committing meaningful budget to a new platform, a small test — a few weeks, a modest spend — tells you more than any amount of theorising about where the audience "should" be. Some assumptions about where a specific audience spends time turn out to be wrong in practice, and it's cheaper to find that out early.