OnlineTours Success Story: How to Achieve a 15% Email Conversion Rate

7
896

About OnlineTours

OnlineTours is the first and largest online travel agency in Russia. The company’s website onlinetours.ru was launched in 2012, and in its first year, was used to provide travel and tourism services to 30,000 people. In 2016, the website received 100,000,000 visitors per month. The company partners with more than 130 tour operators and offers tours for every taste to over 80 countries. Customers can pay for their chosen tours on the website, in the company’s offices, or by courier. The company also has a call center, enabling customers to book a tour or get their questions answered over the phone.

Goal

The services provided by the company are rather expensive. Moreover, buying a trip is not what a person does very often. This is why customer retention and re-engagement are among the company’s primary goals, along with attracting new customers.

‘‘

“In a year, year and a half or so, customers might forget where exactly they purchased a trip. We strive to ensure that we’re the company our customers remember whenever they need a travel service again.”

Konstantin Pobedkin,
Co-Founder at OnlineTours
Konstantin Pobedkin

Looking for a way to increase conversion rates and sales, the company decided to run two experiments and compare what’s more effective: offering services via phone calls or sending out emails.

Solution

Before proceeding to the two experiments, let’s first take a look at the analytics tools OnlineTours implemented in their course of business.

Google Analytics is used to track user interactions with the website: clicks, pageviews, internal searches, scroll depth, time between sessions, number of viewed photos, visits across different devices. With OWOX BI Pipeline all this data is also collected in Google BigQuery in near real time.

The data about all transactions made by a customer, including call center transactions, is sent to Google Analytics via the Measurement Protocol. After that, the data is sent to Google BigQuery using OWOX BI Pipeline.

The data about tours and tour prices for the last 14 days is stored in the company’s own ClickHouse-based warehouse.

In addition, the company adds tokens to every email in their email marketing campaigns. This allows for tracking opens, clicks and traffic for each email recipient. About 10% of the website traffic is identified using token based authentication.

Experiment 1. Cold calls

If a user has performed certain actions on the website, it may be assumed that this user is planning a trip. It’s worth helping him make a purchase decision by making a call and offering the most appealing options for what he’s looking for.

To create a database of call numbers, the company’s marketers identified a number of characteristics by which a customer can be added to the selection. For example, customers are considered willing to purchase a tour if they:

  • used search 5 times and viewed more than 15 tours,
  • used search 10 times in a week and viewed more than 15 tours,
  • used search 10 times, viewed 20 tours, looked through photos or scrolled down to read reviews in 10 tours.

Next, the company selected only those customers whose phone numbers were known. These were the people who already used OnlineTours to travel, or requested dynamic packaging. The company then created personalized offers based on the data about the resort, hotel and accommodation type that each particular customer was interested in. The offers were handed over to managers who then called the customers and offered to buy tours best suited to their requests.

OnlineTours received mostly positive feedback and reminded the customers of themselves. Unfortunately, the company couldn’t generate a sales increase, as the conversion rate for cold calls was 3% on average. Moreover, call center operators spent too much time talking to the customers. The experiment turned out to be unsuccessful, and the company decided to run another one.

Experiment 2. Personalized emails

For the second experiment, OnlineTours also analyzed website behavior of the customers, but the main means of communication was switched to email.

Every day, 40 million package tours are added. They differ by destination, departure date, duration, hotel, room, meals, number of people in a group, and tour operator. Moreover, tour parameters and prices can change throughout the day. As a result, up to 220 million unique packages are available on the website. How could the company offer tours that would be of interest to a particular user?

Let’s say for example, if a website visitor first searches for “Greece 5 star all inclusive” and then starts looking for cheaper hotels, it may be assumed that he’s looking for the best value on a limited budget. In the company’s experience, customers can decide to spend more money than initially planned, if they’re offered what matters most for them. The company decided to email their customers with such offers, by picking up tours that have recently become much cheaper. Such search parameters as department time or duration are considered to be of secondary importance to the customer and can be slightly adjust to include the offer.

Let’s say for example, a customer is planning to visit Greece with three family members for a 10-14 night stay in the end of July. He visits the website and looks for “all inclusive” hotels by the sea, ready to spend up to 80 thousand rubles. The data about price changes in Greek hotels over the last 14 days is stored in the company’s database. Marketing specialists at OnlineTours can easily select tours where prices have decreased the most and pick up the closest to the department dates and the vacation duration for which the customer searched. Although the offer is not an exact match, the customer is very likely to rearrange the dates if he learns how much the price has decreased and how lucky he is to get this offer.

Technically, the experiment is conducted as follows:

  1. A list of email subscribers is created by querying the website database.
  2. Google BigQuery is used to create a list of active website users and analyze their searches.
  3. Scripts are run at regular, pre-set time intervals, to match the data from both lists and see what the subscribers were looking for on the website.
  4. Next, the data about tours and prices in ClickHouse is queried to pick up tours that match the specific searches of each customer.
  5. The obtained data is used to create personalized emails and send them out via ExpertSender.

Results

By sending out emails with automatically generated offers, OnlineTours managed to achieve the following success indicators:

  • Email Open Rate — 85%.
  • Email Click Rate — 65%.
  • Email Conversion Rate — 15%.

Unlike the experiment with the cold calls, the company didn’t have to involve call center resources. With a powerful business intelligence system at their disposal, OnlineTours continues testing different hypotheses, and we’re all looking forward to publishing more success stories from the company :)

P.S. How do you improve email conversion rates? Share your ideas and ask your questions in the comments below.

You might also like