Email marketing for the Korean cosmetics store

What did we have at the start

Quite a large client database of contacts - 4 500 people. It has been collecting for 1-2 years. Work on the database has not been carried out before.

We checked the original base for validity and liquidity. We had to clear it of low-quality emails. As a result, only 3228 email addresses were active.

At the time of the beginning of cooperation, the website added the possibility of registration and a pop-up that offered a discount for registration. Then the user received a confirmation letter of registration and a letter after placing an order. No other work was carried out on the chanel.

What was done

Welcome-chain

Since there was no need to implement a subscription form, we immediately started compiling a welcome chain for new users. As a result, it consisted of 4 letters:

1. A letter with a bonus for registration and how to use it.
2. A letter with the loyalty program and its terms.
3. A letter with the benefits of the campaign.
4. Invitation to subscribe to the social media campaign.

Fig. 1 — Welcome email chain

Welcome chain metrics:

Fig. 2 — Welcome chain metrics

The user could register either through the button on the website or directly from the shopping cart by checking the registration checkbox. Given this specificity of collecting contacts and the fact that the discount was given only for the first purchase, we divided all new contacts into 3 segments:

1. Registered and didn't buy.
2. Registered and bought.
3. Just bought.

A certain number of emails were sent to him based on which segment the user fell into. 

For the 1st group, we have added a letter with the discount terms and thanks for the registration.

For the 2nd, we started the chain straight from the loyalty program.

For the 3rd, we prepared an additional letter - an invitation to register and get a discount.

РFig. 3 — Letter with an invitation to register

Fig. 4 — Welcome chain shopping data

2 607 new contacts were collected for 4 months of work. And this despite the fact that the project came to us for management in early May, when the quarantine conditions were more stringent.

Welcome emails accounted for 57.5% of the total sales from the email channel. This is due to the fact that the target audience was already warm from the beginning and was expecting letters. Such an audience is potentially ready to buy. Therefore, the number of orders exceeds the indicators of promotional mailings by 1.5 times. 

The further task is to provide high-quality service and retain the Client, including through promo and RFM mailings.

Fig. 5 — Welcome chain diagram for those who registered but didn’t make a purchase

Total:

• Sent - 7 821 triggered letters;
• Opened - 939 people (12%);
• 161 people (17.2%) went to the website;
• Conversions - 22 orders (13.7%).

Promotional mailings

We launched promotional mailings from mid-June - after the multitemplate of the email was prepared, the database was cleaned and an account was registered in the mailing service.

Due to the fact that the project works in a popular and understandable niche for everyone, it was possible to play with the content and send not only promotional letters, but also letters with useful content.

Example:

Fig. 6 — Example of the sale email

There were 5 promotional mailings that were carried out in 2 months:

Fig. 7 — Open and conversion rates

The screenshot shows an increase in the open rate and the number of clicks. This is partly due to the fact that we conducted A/B testing of email subject lines and identified which ones work more efficiently. 

Besides the standard mailing list, we also tested emoji in the subject line. Previously, we already conducted a similar test for a clothing store and titles without emoji showed the best result there.

In this case, titles with emoji opposite won:

Fig. 8 — Emoji usage statistics in the subject line

Data:

• Sent - 42 912 letters;
• Messages were opened for the entire period - 2469 (5.8%);
• Came to the website - 425 (1%);
• Bought - 16 (3.8%).

Triggers

We are only running the abandoned cart triggers and RFMs at the moment, so it's hard to say anything about its effectiveness. The email after placing the order worked before, but we changed the template, and also added an email to the chain asking for feedback.

Original email template:

Fig. 9 — Original email template

Template after redesign:

Fig. 10 — Template after redesign

Fig. 11 — Abandoned cart email

Deliverables

Totals:

• Collected new contacts: 2 607;
• Sent: 50,733 letters;
• Opened: 3408 people (6.7%);
• Came to the website: 669 people (19.6%);
• Conversions: 38 purchases (5.7%).

The total contact database was 5,835 addresses in 3 months. That is the base has grown by 181.7% (1.8 times) from the very start.

We have adjusted the work of all email marketing tools and brought the channel to the same level of profitability as referral traffic. Given that the latter was dealt with long before the start of work with mailings. At the same time, the costs for email were significantly lower than for other types of promotion.

Fig. 12 — Traffic growth through the email channel

Conclusion

We can safe to say about rather high rates taking into account the fact that they did not work with the database at all and the contacts in it were 2-3 years old. We increased the open rate of emails by 2% by testing the content, subject lines and gradually increasing the number of mailings during the project.

Increased the contact database by 1.8 times and channel profitability by 2.56%. Thus, work with the new segment showed the highest results.