Social Clicks Don't Convert. Page or Promise?

The complaint arrives in roughly the same words every time. The reel did fifty thousand views, we got eight hundred clicks, and not one enquiry. Then comes the conclusion the agency usually agrees with, because it sells more work: the creative needs to be better, the targeting needs tightening, we should test more hooks.

Sometimes that is true. More often the creative did its job perfectly — it made eight hundred people interested enough to tap — and something after the tap lost them. The question worth answering before you spend another rupee is a narrow one: did they leave before they read anything, or did they read and decide no? Those are two completely different faults with two completely different fixes.

The two failures look identical in your reports

In most dashboards, both show up as the same thing: clicks in, no conversions out. Zero is zero. You cannot tell from the number whether your page took nine seconds to appear on a 4G connection in Ghaziabad, or whether it loaded beautifully and the first line said something the viewer was not expecting.

So you have to go and look. Three checks, in this order, because the first one invalidates the other two if it fails.

Check one: open your own ad on your own phone, on mobile data

Not on office Wi-Fi. Not on desktop. Turn Wi-Fi off, open Instagram or Facebook, find your ad or your link-in-bio, and tap it the way a customer would. Count seconds out loud.

You are testing something specific. A tap inside Instagram or Facebook does not open Chrome — it opens an in-app browser, a stripped-down browser window inside the social app. It is slower, it handles some scripts badly, and it is where most of your social traffic actually lives. A page that behaves on desktop Chrome can behave quite differently there.

Watch for:

How long until you can read the headline. Google's own Core Web Vitals guidance treats Largest Contentful Paint — the time until the biggest visible element has drawn — as good at 2.5 seconds or under, and poor past 4. A hero photo straight off a phone is often four megabytes; on patchy mobile data that is the entire problem on its own.
Things jumping about. If you go to tap a button and the layout shifts under your thumb, that is Cumulative Layout Shift, and Google's threshold for good is 0.1. You will not need a tool to notice it — you will mis-tap.
What covers the screen first. Cookie banner, newsletter popup, WhatsApp widget, language switcher. Three overlays before a sentence of copy is a page that is actively fighting its own visitors.
Whether the phone number is tappable and whether the WhatsApp link actually opens WhatsApp from inside the in-app browser. This one breaks silently and often.

If any of that fails, stop. You do not have a promise problem. You have a page problem, and no amount of new creative will fix it. That is a build and performance job, not a marketing one.

Check two: compare the same page against search traffic

This is the cleanest test available to a small business and almost nobody runs it. In your analytics, look at one page and split it by how people arrived: organic search versus social.

If people arriving from Google convert on that page at a sane rate, and people arriving from Instagram do not, the page works. The page is fine. The mismatch is between what the ad promised and who turned up. If neither source converts, the page is the suspect regardless of what your creative says.

There is an honest caveat: search traffic is warmer by nature. Somebody typing "orthodontist in Laxmi Nagar" has already decided they want one. Somebody who stopped scrolling at your reel has decided nothing. You should expect social to convert worse. What you should not expect is a page that converts at a reasonable rate from search and at effectively nil from social — that gap is a signal, not noise.

Check three: read your ad and your page side by side, out loud

Literally. Ad caption on one screen, landing page on the other. Read the ad's claim, then read the page's first heading. If a stranger could not tell they belong together, you have found it.

The common versions in Indian small-business advertising:

The ad says "starting ₹999". The page shows packages from ₹4,999 and the ₹999 thing is a line in a table three scrolls down.
The ad is a specific offer — festive menu, admission deadline, end-of-season clearance — and the link goes to the homepage. The visitor now has to find the offer themselves. They will not.
The ad shows one product beautifully. The page is a category listing of ninety products, and the one from the ad is not visible.
The ad speaks Hinglish and sounds like a person. The page is corporate English about "synergistic solutions". Same company, two different voices, and the second one feels like a different shop.

The promise problem is usually a temperature problem

When the page is technically fine and the message does match, there is a third thing going on, and it is the subtlest: you are asking for too much, too soon.

Search is pull. Social is interruption. The person who tapped your reel was watching a cousin's wedding video thirty seconds ago. They are curious, not committed. A page that opens with a nine-field enquiry form asking for company name, GST number and budget range is asking a curious person to behave like a committed one.

The fix is almost never "better copy". It is a smaller next step. Instead of book a consultation, offer see the price list, check if we deliver to your pincode, WhatsApp us a photo and we'll quote. You are not lowering your standards; you are matching the size of the ask to the temperature of the visitor, and then warming them up where warming actually happens — in a system that remembers who enquired and follows up, rather than on a single page that gets one shot.

If the ad asks someone to stop scrolling and the page asks them to fill in a form, you have skipped a step. Most dead social funnels are missing exactly one step, not one idea.

The attribution trap, before you blame anything

Some of your social conversions are real and simply not being recorded. In-app browsers, privacy settings and people who switch devices all break the thread between a tap and a purchase. A very normal Indian buying pattern looks like this: sees the reel on Instagram, does not tap, Googles your brand name two days later on a laptop, enquires. Your report credits Google. Your social spend looks dead.

Two cheap counters. First, put a "how did you hear about us?" field on your enquiry form and actually read the answers — it is unscientific, it is also the only data nobody can strip out. Second, watch your branded search volume. If searches for your business name rise in the weeks you run social, something is working even if the clicks look worthless. Our note on matching the right scoreboard to the right goal goes further into that.

What not to buy while you work this out

Plenty gets sold into this exact anxiety. Some of it is work we could happily invoice for and still would not recommend at your traffic level.

A/B testing programmes on small traffic. To tell two versions of a page apart with any confidence, you need a serious number of visits per version. Eight hundred clicks split two ways will produce a "winner" that is just noise, and you will then build your strategy on it. Below a few thousand visits a month to a single page, judge the page by the three checks above, not by a split test.
More creative variations, when check one failed. Fresh hooks pointed at a nine-second page buy you more disappointed people.
Landing pages on a separate domain. The builder-tool pages that live on someone-else's-url.com strip away your navigation, your reviews, your other pages and the familiarity of your own address. For a business whose customer is deciding whether you are legitimate, that is a bad trade.
Heatmap and session-recording subscriptions bought before you have watched a single session yourself. They are genuinely useful tools. They are also widely sold as a substitute for looking, and they add weight to the very page you are trying to speed up.
Boosting the post that already performed. It performed on reach. Reach was never the thing that broke.

The order to fix things in

Make the page fast and clean on mobile data. Compress the hero image, cut the overlays to one, check the WhatsApp and phone links from inside the in-app browser. Nothing else counts until this is done. This is build work.
Point every ad at a page that repeats its promise in the first heading. Offer ads get offer pages. One product ad, one product page. Never the homepage.
Shrink the ask. One action, visible without scrolling, sized for a curious stranger.
Add the "how did you hear about us" field and read it for a month before you judge the channel.
Then, and only then, work on creative — with the content and the destination treated as one job rather than two departments blaming each other.

What to do this week

Take fifteen minutes. Phone off Wi-Fi, open your own ad, tap through, and time it. Then open your analytics and compare search versus social on that one landing page. You will know by the end of it which of the two problems you have, and you will know it better than any dashboard could tell you.

If the answer is the page, that is fixable and the fix is usually boring — image sizes, one less popup, a heading rewritten to match the ad. If you would rather someone else did the looking, tell us which ad and which page and we will go through the same three checks on it.

Questions we get asked

How many clicks do I need before I can say social isn't working?

Far more than most people assume. A few dozen clicks tells you nothing at all — conversion rates for small businesses mean you could easily get zero enquiries from a hundred visits purely by chance. Before you draw conclusions, make sure the page itself passes the basic mobile checks, because otherwise you are measuring a broken page rather than a channel.

Should I send social ads to my homepage or a dedicated landing page?

To a dedicated page on your own website, not to the homepage and not to a page hosted on a third-party builder's domain. The page should repeat the ad's exact promise in its first heading, and it should keep your normal site navigation and trust signals, because social visitors are still deciding whether you are a real business.

Why does my page load fine for me but slowly for customers?

You are almost certainly testing on office Wi-Fi, on a device that has already cached your images, and in a full browser. Your customers tap from inside Instagram or Facebook, which opens a stripped-down in-app browser, usually on mobile data. Test the way they do: Wi-Fi off, tap from inside the social app, count the seconds.

Is it worth running A/B tests on my landing page?

Not until you have substantial traffic to that single page. Split testing needs a large number of visits per version before the result means anything, and on small volumes you will get a 'winner' that is pure chance. At low traffic, fix the obvious things — speed, message match, size of the ask — and judge by whether enquiries rise over months, not weeks.

My enquiries come from Google even when I'm spending on Instagram. Is the social spend wasted?

Possibly not. A very common pattern is that someone sees you on social, does nothing, then searches your business name days later on a different device — and the sale gets credited to search. Add a 'how did you hear about us' field to your enquiry form and watch whether searches for your brand name rise while you are running social.