Advertising of India's Largest Wedding Portal on Instagram

Our agency was contacted by one of the biggest wedding portals in India — wedding.net. Their main request was to attract new customers from Instagram in India’s capital, Delhi NCR. Their target audience was couples planning a wedding who needed to pick a suitable venue. When analyzing the client's business processes, we determined the strong and weak points of their sales funnel. Considering the wide variety of restaurants in the capital — more than 2,000 — choosing one for a wedding is obviously not an easy task, and an expert’s help is welcome. The wedding.net model is especially enticing because their team of professionals is ready to help couples for free, taking into account all their requests. Why for free, you ask? Simple: they receive their fee from the venue that gets chosen.

Wedding.net is well adapted to all types of devices, but when you open it on your phone and see the huge choice of venues, you might just scroll through some of them and leave without calling or requesting an estimate. After discussing it with us, the client decided to create a separate landing page, well adapted for mobile devices, with a short but comprehensive description of their service for picking a wedding venue for free. At the bottom of the page would be a mini questionnaire asking about the main parameters. This way, customers can clearly see what kind of help they can receive, and if someone leaves a request, our client will already know what the couple is looking for in a venue. Next, the client's manager would contact this potential customer for additional information, provide their recommendations, and compile a list of possible venues. When the customer decides on a venue, this manager will help them with booking and signing a contract. Win-win!

After discussing all aspects of the future landing page, we started on its creation. It should be noted that we began with several potential design schemes with different amounts of information. The first design used only the questionnaire, with minimal info about the service. The conversion rate from this variant turned out to be smaller than that obtained by the design made according to the classic principles of creating a marketable landing page:

• The first screen has a header that sells.
• The page urges you to complete an action (fill the form, subscribe, call, etc.).
• Product or service information is easy to read and understand.
• A landing page is created with a particular target audience in mind — people who have similar needs for fulfilling a certain demand.
• The visual look of a landing page requires special attention.

The potential conversion rate of the first variant was about 6%, whereas for a landing page that also had a selling text that described why this service is useful and what benefits you can get from it, the conversion rate amounted to about 18%.

Below is the final look of our landing page as seen on a smartphone, since all our traffic comes from Instagram.

Here we discuss steps we took to get consistent results in the number and the price of leads.

In the first stage of the process, we set up analytics, added a Google Analytics counter and Pixel FB for the main site, as well as set up conversion events in Facebook Events Manager to collect statistics, aiming for the future transition to conversion strategies in promotional campaigns.

Next, we went through possible audiences. We got 16 main variants: from the standard targeting of users with an interest in weddings to a remarketing audience for the main portal and its lookalike.

Next, we tested placements and formats of advertisements. For this, we picked several audiences that we considered to be potentially the most effective: those who visited wedding.net in the last 45 days and 2% lookalike of those who viewed contact information of venues. The audience was restricted to Delhi NCR in the campaign options.

After testing available placements on several audiences, we confirmed our assumption that the format of the feed and stories is still preferable. Their CTR (Click through Rate)  was several times higher than for other placements.

Then came one of the most important stages of the testing — the test of format and contents of our advertisements.

We made various creative ads in different formats: images, carousels, slideshows, videos. Each format was tested for its visual component, the contents of its text, and its call to action. People showed the best reaction to ads with a carousel and a slideshow.

The test gave us information on which types of advertisements were the most effective. Next, we tested all previously created audiences using the best ads from their feed and stories. When it comes to audiences that show the best results, usually it’s people who already interacted with the site or at least know about their products or services. It was the same in our case too. A lookalike seed based on the visitors of the main site gave the cheapest and the highest quality leads. The audience size was big enough such that there was little concern about them burning out initially.

As a result, we achieved an effective promotional campaign with a budget that could be gradually increased without worrying that the client’s funds were spent on something fruitless. As mentioned earlier, at the time of the launch, we planned to eventually transition the ad campaigns optimization strategy to conversions. To do so, we needed to collect enough target actions recorded in Pixel FB, and the existing promotional campaign enabled precisely that. The result of the campaign that worked with the 2% lookalike seed was 211 leads with an average cost of $2.24 per lead.

The algorithms of Meta’s neural network are some of the best in the world. For this reason, when you have enough recorded conversions, you should definitely test your conversion strategy without giving the system any harsh restrictions. Such an approach often yields better results than the periodic update of “burned out” audiences.

Next, we proceeded to the final stage of the setup. For the main conversion point of the campaign’s optimization, we chose the step when, after viewing an offer, a client starts to fill in a questionnaire — the first question, essentially. In it, we ask about the planned type of a celebration. Our audience is the whole Delhi NCR. The restrictions are kept to a minimum: we exclude only wedding specialists (photographers, entertainers, videographers, organizers). Thus, from 22 million accounts in Delhi NCR, we are left with about 4 million, which is, of course, still a bit too much for our subject matter. Here we should get help from Meta’s automatic algorithms that don’t like being restricted, especially for small audiences that burn out quickly.

After choosing the target conversional action for which we have a lot of accumulated data, we tell algorithms that we expect a maximum number of conversions we need for the budget spent. Provided that all requirements are observed, at that step, the possibility to improve key figures should be quite high.

To sum up, let’s briefly go over the results. Several promotional campaigns in such a format have been ongoing for more than a year now and have been consistently effective. We periodically update ads and the text in them — this way, the audience won’t get bored from the same creatives. The cost of an order sent was cut down from $2.24 to $1.5–$1.7. Here are the campaign statistics for May 2022 as an example. Right now, two almost identical campaigns are running in Delhi; the difference in the content of their ads is minimal. Both use the reply to the first question as a conversion event in their optimization strategies. As you can see, algorithms work almost identically in both campaigns. This is clearly visible in the cost of a lead — the difference is just one cent :)

Finally, we want to point out how important it is to use a structured approach and staged testing of the main elements of any advertising campaign. By spending a small amount of funds on test campaigns at the start, we determined the best combinations of targeting and creatives, as well as cut down the cost of a lead to an acceptable amount. After collecting enough statistical data regarding the target action, we switched campaigns to optimizing this conversion, which allowed Meta algorithms to find people potentially interested in the service. And that, in turn, allowed us to cut the cost of a lead by additional 30%.

Our team is ready to conduct a free analysis of your existing Instagram campaign, find points of growth, and draw up a staged action plan to improve your business performance. Just submit a request!