How a car rental service can increase sales by 2.8 times amid fierce competition and rising tourism prices
How to increase the profit of an interregional car rental service in a highly competitive environment? We analyze successful contextual advertising strategies that helped triple the number of orders. What to do if domestic tourism in Russia is growing, and demand for car rental is declining? Focus on the premium audience and promote expensive cars.
In this article you will learn:
On how to adapt a car rental advertising campaign in Russia to the decline in demand for tourism services.
How to use Yandex.Direct's new tool – the Unified Performance Campaign (UPC). And why the experiment with the Yandex Advertising Network (YAN) was not as successful as expected.
How to reorient your strategy to a premium audience, increase your average bill and boost sales by 2.8 times.
How to use the new Clickable Promo Code tool in your ads.
What is the value of constantly testing new tools and adapting to changes in the market?
The car rental service has stopped growing due to increased competition in the niche. Advertising campaigns in Yandex.Direct did not help due to low rates for conversions
At the end of February 2024, we were approached by a client engaged in car rental throughout Russia. He needed to focus on the following areas: Crimea, St. Petersburg/Leningrad region and Kaliningrad. Competition for car rental was actively growing, more and more services were appearing. And sales were declining.
The client believed that his main target audience in the mid-price segment was families with children who preferred active recreation. And he built his marketing strategy based on this hypothesis. He tried to launch contextual advertising on his own, but there were no good results.
The company had a number of advantages: a large fleet of its own vehicles (300+ cars), a convenient online booking system on the website, allowing customers to instantly book a car.
We conducted an audit and identified errors in the advertising settings. Several campaigns did not gain conversions, and automatic strategies were not trained because low rates were set. Yandex algorithms did not gain conversions at such prices.
Also, the budget settings were set incorrectly. For example, for 10,000 rubles they wanted to receive at least 10 conversions.
Meanwhile, the actual conversion price was 3,000 – 6,000 rubles, i.e. the budget was enough for 2-3 conversions. Algorithms will never learn like that. Of course, everyone wants to get orders cheaper. But it is better not to underestimate the advertising budget so that campaigns do not stop due to a lack of conversions.
There was another major mistake in advertising – the lack of control and adaptation of strategies. They set up the tools once and forgot about it. With this approach, the budget was often spent on irrelevant impressions.
Focused on a batch strategy for data collection and algorithm training, but failed to increase conversions
We optimized the budget by connecting micro-conversions (user actions before placing an order) and combining them with macro-conversions (placing an order).
The following micro-conversions were tested:
Time on site (1, 2 and 4 minutes);
View vehicle card;
Pageview depth.
We identified a correlation between microconversions and final conversions, looked at which actions most often lead the user to a purchase. Viewing the car card worked best, this was proof of engagement.
We added card viewing to the advertising goals. We signaled to the algorithms that we wanted to attract people who perform this action. But, of course, we wanted to get those who rent cars even more. We indicated the importance of this goal by setting the conversion value: this macroconversion was more expensive than microconversions.
The amount and value were calculated based on the ratio of micro conversions to orders. For example, out of 10 card views, we have 1 order. If we want the application to cost no more than 4,000 rubles, then the card view should cost no more than 400. And for macro conversion, we set the price at 3,500 rubles so that the system understands that this goal is more important to us. We estimated how much 10 conversions would cost and set the appropriate budget. And then we started testing the hypotheses.
At the start we decided to test the following strategies:
We worked on brand queries. The client had not used them before, but they are important, because people searching for our brand are most likely to place an order. In order to better control this conversion traffic, we allocated branded queries to a separate campaign.
Segmented search campaigns. We made different advertising campaigns for geolocations where you can rent a car (Crimea, St. Petersburg, Kaliningrad), and each of them was shown throughout the Russian Federation. We also divided different segments into different advertising campaigns, for example: advertising campaign Search Brand St. Petersburg, advertising campaign Search car rental St. Petersburg, advertising campaign Search priority St. Petersburg. We made different advertising campaigns for these segments because they differ in the degree of users' readiness to place an order.
Used a batch strategy. Each of the search campaigns had less than 10 conversions per week, so the algorithms were not trained. We used the Package strategy for all three campaigns, and in this case, conversions from all three campaigns were used to train the algorithms.
We turned off non-converting regions to get more conversions. But it didn't work. Since automation now plays a big role in training Yandex algorithms, the fewer restrictions, the better. To keep the campaign from fading over time, it's better not to cut geotargeting coverage and not limit algorithms.
At the end of the month, the results were disappointing: we managed to reduce the cost of a lead, slightly increased orders, but the data for full training of Yandex algorithms and automation began to fade, impressions began to decrease.
We started looking for other ways to solve the problem. And the next step was to give free rein to Yandex algorithms to automatically select the advertising audience.
We divided search campaigns by car rental locations, launched EPC and received 2 times more orders
The campaigns were initially divided by car rental locations (Crimea, St. Petersburg, Kaliningrad). For each location, there were separate campaigns with geotargeting for the whole of Russia. We decided to change this strategy: we combined all the locations into one campaign and divided them into groups: rental in Crimea, in Kaliningrad and the Leningrad region. And we made 2 campaigns with different geotargeting settings: Moscow and Russia.
We decided to enhance the effect of two advertising campaigns on Moscow and the Russian Federation by using a new type of advertising campaign — EPC (Unified Performance Campaign). Usually, we immediately test new Yandex Direct tools to get the most out of advertising. It is no secret that Yandex secretly promotes its new formats. It is later, as the tool penetrates the market, that activity decreases, when more and more advertisers appear and competition increases.
The EPC is flexible in its settings: you can choose the display location, use more targeting and create more interesting scenarios. We set up displays on Search for all rental locations. And we launched the EPC in parallel with the old search campaigns to compare what works better. Since in any case, Yandex.Direct gives preference to more effective campaigns.
The search was chosen due to the specifics of the customer journey. In our case, car rental is situational consumption, when the user has specific tasks (business trip or vacation) and a conscious, formed, hot demand. The renter purposefully chooses a car for himself using specific queries in the search bar.
As a result, the automatic strategies of EPC worked better than the usual Search and attracted the majority of conversions. The total number of rented cars increased by 2 times compared to the previous month.
And in June 2024, demand for rentals fell and sales too. We experimented in YAN, but received 35% fewer orders
In summer, demand for car rentals usually grows, because the hottest time of year begins — the vacation period. But 2024 brought an unpleasant surprise — bookings, on the contrary, began to fall. It was necessary to urgently find a new source of traffic. We doubted, but still decided to use YAN, although Yandex Advertising Networks are not a tool for hot demand and have not previously taken off on similar projects.
We tried not only those semantic segments that we used in the search, but also tried to expand the audience by showing ads in networks to those who have not yet searched for car rental, but are our potential clients – looking for plane or train tickets to the locations we need. We used the strategy of payment for conversion, so as not to spend money on those who are not our target audience.
In this case, the hypothesis did not work. There were conversions from YAN, but there were very few of them and they were not 2 times more expensive than conversions from Search.
The total number of car rental orders was steadily declining. And the cost per conversion was up 24% compared to the previous month. So we started developing a new advertising strategy.
We found a more profitable niche and started promoting expensive cars. We attracted the democratic target audience with clickable promo codes. We returned the previous level of sales
We started to discuss the situation in the tourism and recreation industry. Here are the disadvantages we saw: limited choice of tourist places, tourism prices continue to rise, the solvency of the middle class is falling. People are starting to save more on vacations, so the demand for renting affordable cars may decrease. And those who are ready to rent a car are looking for cheaper offers. To turn a minus into a plus, we began to attract the affordable target audience with lucrative bonuses.
For example:
In August 2024, Yandex Direct introduced a new tool, “Clickable Promo Code.” Previously, promo codes in ads were text. Now, using the “Clickable Promo Code,” you could go to a specified landing page with the terms of the promotion. Of course, we immediately used this new tool.
But we didn't limit ourselves to working with a democratic target audience. We thought: if we can't significantly increase bookings, then we need to increase the average check. This way we can increase the client's sales. And for higher checks, we needed to attract a different audience – a more solvent one. After all, our client's car fleet also included expensive cars. But such new advertising offers needed their own landing pages, which they didn't have.
Initially, the site had landing pages for rental locations displaying a list of cars that were rented there.
We added sorting by car class to the interface. We created landing pages for each location with premium cars. VIP audiences began to be brought to them.
We prepared landing pages, advertisements and started training Yandex algorithms to bring us premium target audience. To do this, we downloaded the audience that ordered luxury cars earlier. And based on this segment, we created Look-a-like audiencewhich was adjusted upward. The system found users who were similar in behavior to the loaded audience and showed them our ads.
So we returned the client's sales level to the previous level. And increased the average car rental bill. Previously, it was 26,500 rubles, and after the launch of the promotion of expensive cars, it increased to 35,600 rubles.
Project results: car rental orders increased 2.8 times in 6 months
We have achieved growth in car rental orders, despite high competition in the niche and an unexpected drop in demand during the high season. Over the first 6 months of 2024, we increased requests from 62 to 175 per month. At the same time, the cost of an order was reduced from 6,941 rubles to 3,787 rubles.
In conclusion, we would like to give a brief summary of the contextual advertising strategy. In any market situation, no matter what happens, you need to look for a way out by testing different hypotheses. You cannot set up campaigns once and forget about them. It is important to test even those hypotheses that were previously unsuccessful. Out of 10 tests, you may have 7 erroneous ones and only 3 successful ones. But if you do not try, you will not get high results.
And you can't ignore the trends of Yandex Direct. Yandex is developing, constantly changing its algorithms. To keep campaigns from fading away, it's better not to limit them too much. Strategies for splitting campaigns by geo, product categories – are starting to become obsolete.
The algorithm system learns when it receives a lot of data. If there is little data, then problems begin – orders decrease. Our task is to throw as much data as possible into the campaign furnace to bring in the maximum number of orders. That is why we need to experiment.
If you need to increase your conversions, order our deep audit of contextual advertising for 0 rubles on makodigital.ru
How will this be useful?
We will find technical errors in campaign settings and segment them by danger level: minor, significant and critical.
Let's check the statistics and find the reason for the increase in the cost of circulation and the fall in profits.
Important! We will recommend positioning, content and development (conversion, usability, functionality) for the site.
We will develop a contextual advertising strategy with new growth points.
