Case Study on Bus Transportation - Infobus

Intro

In November 2016 we took on the international bus ticket service Infobus 🚍 as a client. The company connects passengers and carriers, searches for all available routes, and also shows current prices and travel time. Users only have to choose a carrier, and the place and make an online purchase.

While Infobus was reducing the distance between cities, we had to bring the service closer to potential users. Namely, we had to raise brand awareness in Europe and increase bus ticket sales 🎫 through the online service.

Initial Data

• Client’s website – Infobus.eu
• Advertising account – Google Ads
• Client analytics system with our UTM tags

During our first communication with the client, we learned that PPC advertising had already been carried out by the customer independently, though the costs were 4 times higher than the return. We decided to create advertising campaigns from scratch and sent a brief to the customer. The information we received showed that Infobus covers thousands of bus destinations, which means that special attention needed to be paid to the structure of advertising campaigns. The estimated advertising budget was 300-500 thousand rubles per month, which we allocated to search ad campaigns, a display network, and dynamic remarketing.

Objectives

Since the main goal consisted of two objectives (brand awareness and sales growth), we divided it into two global tasks:

• To increase sales – create search campaigns for each trip direction.
• For brand awareness, enable dynamic remarketing and search advertising for brand queries.

Aside from global ones, we set out the current objectives:

• Update the dynamic remarketing data feed once a day so that users can see up-to-date prices
• Track advertising costs and ROI once a week in the analytics system

Plan of Action

Since the Infobus service has its roots in Czech and is quite in demand there and in Europe, we decided to create ads in Czech, English, and Russian.

But the main action was to optimize the structure of advertising campaigns:

• We created one campaign for each country with about 10 of them altogether.
• An ad group was created for each direction from each country. The number of groups within the campaigns ranged from 30 to 2000.

We created one campaign for each country, we got about 10 of them.
An ad group was created for each direction from each country. The number of groups within the campaigns ranged from 30 to 2000.

Here’s an example: an ad campaign for Czechia включает hundreds of groups, like 🚏 Prague – Warsaw, Prague – Moscow, Brno – Berlin, Brno – Katowice, etc. To avoid confusion with advertising return flights, we did the following:

• A campaign only refers to the country of departure. In other words, the route “Prague – Warsaw” is in the Czechia campaign, and “Warsaw – Prague” is in the Poland campaign.
• We cut back traffic using negative keywords. For example, on the list for the Moscow – Warsaw ad group, we add the exclusions “Warsaw – Moscow” and “from Warsaw” in the correct keyword match type.

All of the above settings were made manually using Google Ads Editor and Microsoft Excel. The ad for each route led to the corresponding landing page.

Implementation

Search Ads

After carefully setting up and checking all routes, the adjusted ones launched impressions at an average rate. With our number of requests, high rates would simply “drain” the budget 💸. The results of the first days of impressions helped to calculate the optimal rate and brought the first conversions, which we set up for buying tickets.

After getting enough conversions, the smart Target CPA strategy was connected. At that point, Google dictated who to show ads to, targeting users who were more likely to visit the site to buy a ticket. Since the user could visit the site several times before purchasing tickets, “Position-related attribution” was selected.

Dynamic Remarketing

We set up dynamic remarketing using a data feed in Google Ads. This showed that Infobus.eu visitors, who did not make a purchase right away but continued to compare competitors’ offers, saw our banners with the route of interest and the current price. This prompted them to buy from our client. The effectiveness of dynamic remarketing was proven by the large volume of inexpensive targeted traffic.

Since ticket prices are updated daily, we updated our feed every day to only show people relevant information.

Smart Display (Smart Campaigns)

We also tried out a smart display campaign based on Google machine learning to engage audiences. Smart Display can take on full control for bidding, targeting, and ad creation. However, the cost of conversion turned out to be high, so it was decided to discontinue smart display campaigns.

Tracking Efficiency

We used UTM tags in the client analytics system to see how much ads costs and how much money they make. This allowed us to see the most priority areas, adjust settings, and track conversion costs.

⚠ The effectiveness of advertising depended not only on the settings and analytics within the advertising campaign but also on the timely response to changes in the target audience. So in a period of growing demand for tickets from Ukraine to Poland, emphasis was placed on the corresponding campaign, which would bring conversions at a cost of 25-30 kroons.

Results

The work was carried out from November 2016 to May 2017. Already in the third month of work, advertising brought in twice as much as was invested in. Taking the previous period into account, the benefit went up fourfold, so we can talk about an increase of 600% 🚀.

The conversion price was up to 40-60 CZK. This result was made possible thanks to the right structuring for huge amounts of data, Google’s intelligent strategies, click attribution and dynamic remarketing.

Conclusion

Working on the Infobus project, we combined a creative approach with carefully considered actions. The smart strategy with Google Ads we had chosen fully met our expectations, but its effectiveness was based on painstaking manual preparatory work performed by our specialists. The results we obtained once again prove that we took the right actions.