case study: higher rebuy rate and lower costs

7 Nov ‘16 is an international flower delivery service that was founded in 1999. The chain boasts more than 700 flower shops in Russia and abroad. The company has served more than one and a half million customers from over 150 countries.

Your company focuses not only on «we will sell for you», but also on your good, powerful product and willingness to customise to meet client demands.

Arkady Balandin, Chief Analytical Officer at

Mindbox’s primary goal was to centralise all emails in one place and set up a sequence of triggers with complex logic and testing capability. Initially, part of’s emails were sent via a third-party system, and another part via an outdated CRM and various scripts that they had written in-house.


The data were stored separately, messages could overlap, and independent email triggers sent customers similar messages. This was quite inconvenient. The information about customer site interactions collected by the tracker and information about orders placed by telephone or paid for offline were stored in separate, unconnected databases.



In just two months, we transferred the entire existing email and SMS function to Mindbox, and it was already more profitable than the previous tools we had been using. In the first month, the following triggers were transferred to Mindbox:

  • Abandoned cart email. If the user left items in their cart on the site, they get a reminder
  • Reactivating clients who have left and, based on their data, are more inclined not to return
  • Abandoned view and category scenarios conducted A/B testing, with 50% of customers continuing to use the old system.

The testing showed that these triggers even work slightly better with Mindbox (by 1%, a statistically significant result), while Mindbox costs less to use.

The further, the better.

67’s analysts developed an algorithm to collect unique information about their customers.

The data make it possible to more accurately predict the customer’s next actions. We wanted to use this knowledge to help the customer place a new order. Customer parameters can be instantly created in the Mindbox interface and put to immediate use in business.

We built a series of triggers that are personalised for each customer based on their parameters.

These triggers can be broken down into three basic categories:

  • Next point of contact (we send the customer an email on a date when we predict the customer might place a new order)
  • The customer’s behavioural patterns suggest they might leave and require additional stimulus to return
  • The customer has a strong tendency to make purchases and requires special attention

Next point of contact

We set up several triggers that run when a particular client has a combination of these parameters at any given time. Google Analytics shows a conversion rate of 30%.

Compared with the control group, there is a statistically significant increase in sales.

We got rid of the old CRM system for email and SMS. The IT staff that this freed up can now work on other profitable company projects.

Customer reactivation

Some customers might not place a new order, even though they fully comply with a given segment. There can be various reasons: the client’s financial situation may have changed, or there could be other parameters that neither we nor are technically able to collect. In such cases, the customer reactivation trigger should run. Customer reactivation emails really make a lot of people change their minds – the conversion rate is 24.6%.


Recommendations for next purchase

Since’s product catalogue changes almost daily, an automatic product recommendation system based on «if they bought product A, they will probably also buy product B», which is usually used in such situations, is not always effective. It’s important to give the customer information about only the products that are popular and available when they receive the email.

After A/B testing, the automated recommendations in emails were replaced with collections generated based on current product stocks. A unique bouquet of flowers was developed for each segment.

Manual messages is a unique project:

  • Manual messages are almost never used
  • Most messages are automated and planned 5 years ahead (conditions can be fine-tuned and tests can be run at any moment)
  • The remaining manual messages nonetheless rely on detailed segmentation of the client database, which makes them essentially indistinguishable from the trigger-based messages.

The standard practice of «let’s send the entire customer database the same email each week» has been replaced by dividing the database into detailed dynamic segments. Customer interactions on the site, as well as offline, prompt triggers for whether it’s necessary to send them seasonal emails, such as «we’ve updated our collection».


We have integrated with the store, gathered under one roof all necessary data about customers and orders, including parameters that are unique to We have launched triggers, added product recommendations based on customer segment and automated the manual messages.

Feedback from

To better understand what it’s like to use our system, we briefly interviewed Arkady Balandin, Chief Analytical Officer at

What do you think is the most important aspect of working with Mindbox?
The ability to implement all planned scenarios for increasing email and SMS sales.

Why Mindbox?
Mindbox offered a product that was both flexible and ready to use, but is customisable to meet our needs.’s business processes are often unique and require unique functionality.

How quick was the integration for you?
We integrated with Mindbox fairly quickly. As’s project manager for the Mindbox integration, I noted that our programmers did not have to do very much. In just two months, we transferred the entire existing email and SMS function to Mindbox, and it was already more profitable than the previous tools we had been using. Then we created triggers fairly quickly together with Mindbox’s managers that we had not planned to implement until the distant future before the integration.

What changed at after the integration?
Our developers have been freed from 99% of the obligation to support the old messaging tools. Our overall efforts to refactor the code made it possible for us to delete a number of irrelevant scripts from the project. We got rid of the old CRM system for email and SMS. The IT staff that this freed up can now work on other profitable company projects.

How long did it take for the Mindbox integration to pay for itself?
Even accounting for the time it took to introduce the tracker and develop APIs, as well as the cost of paying our developers’ salaries during the integration, the project fully paid for itself in two and a half months.

What do you like about Mindbox’s out-of-the-box functions?
We currently have more than 20 active triggers running. For example, we definitely don’t want a client to get 10 emails from various triggers in five minutes (and get lost in them). It’s important to understand the order in which the triggers can run, as well as how they can block each other. It’s particularly nice to see the trigger conditions in the Mindbox product in the form of a logical schematic that can literally be printed out on paper.

How do you like the tech support? 🙂
It’s important that messages and triggers can be created by just writing a detailed technical task. Our personal manager at Mindbox takes care of the rest of the work. Live communication is sometimes just as important as the powerful functionality of the product itself.

What are your future plans with Mindbox?
Conduct A/B testing to get even more accurate settings for each trigger. Of course, profit generated will always be a criterion. Perhaps we will begin to use other Mindbox products to maximise the personalisation of our customer relations.

«Your company focuses not only on ‘we will sell for you’, but also on your good, powerful product and willingness to customise to meet client demands.

We know our customers and what functionality we want to use. By working with you, we can convert this knowledge into sales».

Arkady Balandin, Chief Analytical Officer at

Mindbox project team:

  • Filipp Volnov, Project Manager
  • Semyon Mikanyov, Lead Manager
  • Leonid Gureev, Technical Manager

Tell us a little about yourself

We’ll respond within 24 hours

Partnership request

Typically we’re answering within 24 hours