Case Study: Info Business Funnel Promotion

It was proposed to experiment, rewrite the text, and test different conversion buttons alongside informational ones. During the analysis of the advertising campaign, we saw that the AI audience was used broadly and with various interests. If the audiences performed well, we wouldn’t have understood which interest yielded positive results. Therefore, money would have been spent on unnecessary actions.

Our entire subsequent strategy was aimed at using a maximum of two broad interests that we set for testing.

During analytics analysis, we saw that a person makes a decision on the same day. As soon as they encounter a creative, they should register. The 7-day verification strategy or 1 day after clicking was changed to 1 day after clicking or viewing. This way, we realized that the audience would burn out, and it would be necessary to change creatives as often as possible.

The client agreed with this, and we launched our first advertising campaign. On average, our audiences were not large, approximately 250-500 thousand people. In 2018, Facebook had not yet significantly lowered the cost per thousand impressions, so the audience was reduced as much as possible.

At the same time, 50% of the entire new audience we launched for testing was “Look-a-like.” Initially, the “Look-a-like” audiences were targeted, meaning they first came from people who made payments and then less targeted ones from those who interacted with us in some meaningful way, such as sending a message on Instagram.

There were many audiences, and they performed well for several days since the project’s budgets allowed it.

The funnel was in the format of a lead magnet, where we offered a system for entrepreneurs to scale their businesses. The user had to fill in three fields: name, phone number, and email address. All three were mandatory. During funnel analysis, it was suggested to reduce the number of fields to two (name and phone number), but during the advertising campaign launch, we did not see a huge increase in conversion rates. At the same time, clients did not have a single interaction channel. The client decided to abandon this test, and we continued advertising on the landing page with three fields. Additionally, when a person received a series of emails, they were directed to a thank-you page, where they were also invited to subscribe to Telegram, where they received a checklist.

After receiving this series of emails, they tested different offers and email sequences to convert this traffic. Our task was only to drive traffic, so we were indirectly involved in this aspect. The client’s team handled all internal work. After several months, analytics became crucial. Before our involvement, there was no digital tracking of advertising results. We broke down each audience into unique identifiers. The first thing we did was remove UTM tags. Previously, they were used only for internal tracking. Our task was to use UTM tags to understand which target customers would purchase our product. We standardized the questions with UTM tags, and once we had enough data collected through these tags, we proceeded.

About six months passed, and we began identifying who our target customers were. Each audience had a unique identifier. For each audience, we collected the total budget spent and the total revenue generated from this funnel. Since the funnel was a lead magnet, the downside was a long decision-making process. Sometimes, a client took up to 4 months to decide, which was bad for the advertising campaign because the people we targeted 4 months ago might have already been excluded from our audiences.

When we presented the first results and they were positive, the client proposed scaling the advertising campaign. As a result, we decided to merge Facebook and Instagram, as it was part of Facebook’s next update, and we needed an audience of more than 1 million to reduce traffic costs.

We merged placements and launched ads, but we lost in efficiency. Before the launch, we noticed some traffic coming from Facebook, and its ROMI was slightly higher than that of Instagram. When we merged these audiences, we gave Facebook the ability to choose sources freely, either Facebook or Instagram. If it saw more leads coming from Instagram, it directed more traffic there. Our funnel became more expensive, with a $5 per lead cost on Facebook and $1.2 on Instagram. Additionally, the challenge was that at least twice a week, creatives burned out, although sometimes a creative performed well for a long time. We searched for stock materials and sent them to video production for editing new videos.

Ultimately, the project lasted for a long time. This case had a positive impact on our long-term activities in the information business sector.

The client was willing to test hypotheses using core approaches, immediately discarding ineffective ones while scaling and optimizing successful solutions.

As a result, we considered the promotion successful. Our client believes that we successfully handled the tasks assigned to us.