Book24.ru is an online bookstore that offers its customers more than 50 thousand items. The store officially represents many Russian publishers – EKSMO, MIF, and AST among others. Book24 had an existing and rather nice bonus program in place.
In the first success story, we covered how, thanks to the personalization of Book24’s newsletters, they managed to earn an additional 3 million rubles.
Since the start of 2019, the share of the email channel relative to the company’s total revenue had nearly doubled thanks to a transfer of triggers from their previous system over to Mindbox and the launch of new mechanics. We have moved and launched more than 30 automatic campaigns together with the client since the start of 2019.
The result of the success story is an increase in the share of revenue from the email channel by means of using the last click algorithm, the graph is redrawn according Book 24’s Google Analytics. We managed to increase revenue on account of launching more than 20 new mechanics, transferring up to 10 triggers from their previous system to Mindbox, and personalized mailings.
The email channel share doubled
Opinion of Book24’s Director
With Mindbox we enjoy the fact that their guys are improving together with ours.Even if our newly emerging goals happen to exceed their skills, we know for sure that after a certain time our task will be solved.
I would also like to praise the interest and care of the managers with whom you work with directly. It’s nice when your question becomes a general question.
We have launched over 20 new mechanics since the start of 2019. For example, here is a list of the most interesting ones, from our point of view and of Book24:
- Abandoned basket: a chain of 3 letters with a decreasing discount
- Informing customers about a personal discount changing
- Informing customers that a lot of bonuses have been accumulated
- Regular collections of books for those who read letters but don’t make purchases
- Regular collections of books for those who have never shopped
- Regular collections of books from a favorite category
- A request for feedback on an order
- Favorable offer to those who added goods to favorites, but did not buy them
- Welcome chain that introduces the store and its capabilities
Below we go in depth about a few of these mechanics.
New mechanics with points
There is a bonus program on the Book 24 site – where customers are awarded points for purchases that can be spent on their next orders. Marketers of the project decided to embed information in their letters on points accumulated by customers, to experiment with motivational point mechanics.
Ran a rest mailing with points
The first newsletter involving points was launched manually in order to evaluate indicators. We carried out an import of customer balances and substituted the number of points on the account in the header and text of the mailing.
The first newsletter involving points was launched manually in order to evaluate indicators. We made an import of customer balances and substituted the number of points on the account in the header and text of the mailing.
|Sent||Orders||Conversion to orders||Last click attribution revenue|
|560 553||594||0,11%||904 762 r.|
Added balance and discount to newsletters
The first change we introduced was to add an account balance score and a personal discount to the newsletter header. This will allow the client to always be aware of the current state of his bonus account.
Launched reminder letters about points on customer accounts
To remind customers that points have accumulated in their account, we launched trigger mechanics, which send out a letter with a balance:
|Sent||Orders||Conversion to orders||Last click attribution revenue|
|44 785||550||1,25%||867 895 rub.|
Launched letters on increased personal discounts
The Book 24 project has a laddered discount system – depending on the total amount of orders, the client can receive a constant discount between 4 to 10%.When the client moves to the next level, we automatically send him an email with the size of his new personal discount.
|Sent||Orders||Conversions to orders||Last click attribution revenue|
|3426||42||1,29%||72 457 rub.|
Compare points with discounts in abandoned shopping carts
Book 24 marketers wanted to check what’s more interesting for their customers: discounts or points? To find out, they launched a test in the «Abandoned shopping cart» letter, in which the first option offered a 20% discount on goods left in a cart, while the second option offered users 2000 points.
A/B test results for discounts and points in an abandoned cart
|Option||Sent||Click Rate||Conversion to orders|
|Option 1 – letter with a 20% discount||14 242||11,14%||22,03%|
|Option 2 – letter with 2000 points||14 032||15,02%||29,3%|
The letter option, in which we proposed to accrue 2000 points to pay for goods in the cart, won based on click rate ,with a margin of over 3%.The differences are statistically significant at 95% confidence level. There is also a statistically significant difference in conversions from the recipient group to orders (calculated for the period of 7 days after the letter) – at a confidence level of 99%.
Using the Mindbox scoring and segmentation system
The Book 24 team of analysts has created a model that allows them to for example, select people that are interested in detective novels. Inside Mindbox, it was decided to send the newsletter not only to these people, but to a wider selection – built using Mindbox filters based on viewed items.
As a result, we wanted to see how much additional revenue «expansion» would bring. Perhaps the indicators are good enough to include additional customers in the mailing list.
We received two mailing levels:
А) A selection of Book24 analysts – 4,938 people.
B) Selection using filters Mindbox (there are order, viewing categories or specific books from the category «Detectives») – 38 689 people.
At the same time, groups 1 and 2 intersected for 4 110 people. That is, 83% of Group A customers were included in the Mindbox sample.
Here are the indicators obtained from the letters:
|Sample||Sent||Amount of orders based on the last click algorithm||Last click conversion to order|
|Book24 Analysts||4938||31 309||0,49%|
|Based on Mindbox filters||38 689||135 361||0,26%|
Although this is not an A/B test, we can nevertheless compare it using a calculator: differences in conversion are statistically significant at a confidence level of 97%. It seems that the selection of Book24 analysts turned out to be more effective in terms of conversions.
The results of the mailing were also compared with the third group in deciding whether or not the newsletter should be sent to the Mindbox group. It did not intersect with groups A and B:
C) Selection using Mindbox-2 filters (there are orders, category browsing of any books, including detective novels)
Indicators of the same mailing for group B:
|Sample||Sent||The amount of orders by the last click algorithm||Last click conversion to order|
|Based on Mindbox-2 filters||27 573||42 572||0,11%|
We see that the interest in mailing by conversion rate is lower than that of customers who were interested in detective stories. According to the calculator, the difference between groups B and C in terms of conversions are statistically significant at the 99% level.
By such an indirect method, it was estimated that expanding on the sample of analysts using Mindbox filters, which include clients with an interest in detective stories, is useful enough to send newsletters to such clients. Further expansion – to customers who were simply interested in books – does not lead to a lot of additional revenue. In order to not send a newsletter with info on detective novels to those who are not interested, we decided to not use the third sample.
Book24 Head of Direct Marketing
Head of marketing opinion
Our new email marketing strategy involved major changes to our approach, content, and visual components of the letter.
We completely reworked the letter template, made it more informative. Segmentation was up next. Here, designers felt the bulk of the work. They selected template elements in a such a way that the letter sent to clients with an interest in science fiction would differ fundamentally from the letter sent to the «sentimental prose» or «children’s literature» segment. We built our segmentation not only based on the choice of the segment itself, but also on the visual part.
Of course, in addition to this, corresponding content is substituted into the letter. Such as recommendations from the category of interest for a user, a thematic selection from the blog, and more. Currently this is being done manually to consolidate results. But we already have a plan to automate that during the second quarter.
According to results of the first quarter, revenue from the email channel more than doubled. And surely, this is not the limit.
I like that there are no restrictions in Mindbox. Any ideas, even the most usual ones, at first glance that is, are realized quite smoothly and easily. Account managers deserve special attention. They are part of our team, and sometimes several full-time marketers replace me.