Julien Le Coupanec“ Do you need help to grow your company? I have seen hundreds of creative ideas to get new customers and to increase retention. Call me on Clarity today and get your answers now. ”

About this Experiment


The idea is to activate our new email subscribers by onboarding them through a series of emails. When a user subscribes to your service (or in my case, to my blog), they might not know where to start. To solve this problem, we are going to send them the information in a certain order and at a certain time. This process is called building a drip campaign. If we can make them do the important actions with the appropriate sequence, it will move them from one step to another. This will increase retention as they start using our service more intensively and develop a habit (so, you have to understand the whole funnel of what triggers make someone active).

To save us time, we will use Autopilot (you can do the same with Mailchimp Automation but it is more expensive and less user-friendly). I let you dive into their service to know more about every feature they offer (you can send SMS, Slack messages, Notifications and other interesting actions so that you engage them on the right channel). The main thing to remember is that no coding skills are required (building your customer journeys is as simple as drawing on a white board: see the picture below) and that it is completely automated.

After you subscribe to my newsletter, I send you a series of articles to read. The idea is to make you read my best pieces of content in the right order.

After you subscribe to my newsletter, I send you a series of articles to read.
The idea is to make you read my essays in the right order.

Julien Le Coupanec“ As my emails are automatically pushed to Mailchimp, I use Zapier to seamlessly add them to my Autopilot list. Besides,drip campaigns are great to build a relationship with your users. But this means that you have to send your emails in a plain-text format. Avoid all marketing type emails when you are an early-stage startup (and maybe even after), as it is so less personal and the click rate is much lower. Just make sure to keep your emails short with a clear visible-link at the end. Here is the first email I send in this drip campaign. ”

Email Autopilot Drip Campaign

Metrics and Tracking


Autopilot offers a built-in reporting system in their dashboard so we do not have anything to do other than looking at it. They let you access various metrics like the open rate, the click rate, the clicks per unique opens and the number of people who unsubscribed. In our case, we will mostly care about the click rate and the number of people who unsubscribed. To conclude this experiment as a success, we will have to reach a click rate above 15% and an unsubscribe rate below 3%. If this is the case, we will optimize our system and invest more time in it (read this essay about why should focus on something until it saturates).

Julien Le Coupanec“ I do not really care about how many people are opening my emails (even though it could be an interesting metric if I want to optimize my titles), I only look at the opening rate when I am afraid to fall in spam (an opening rate below 10% is a clear sign that something went wrong). For me, the click rate remains a much more interesting indicator to assess the performance of this experiment. ”

Lesson Learned


So, did we succeed to reach a click rate above our objective?


Click rate is above 23% and almost 1 out of 2 people who opened our emails has clicked on the link.


This experiment was run on 300 subscribers which lead us to a 2.66% unsubscribe rate.


This experiment was run on 300 email subscribers during 4 months. Here are the number of clicks and opens per day. As I did not write any new article for the last three months, less people were giving their email address and were added to the list. Thus, the graph is decreasing as less emails were sent every day. The click rate remains greater than 20% though.

The click rate and the unsubscribe rate are pretty good and the size of our sample (300) is high enough to affirm the performance (conclusion: this is definitely a retention weapon). We will keep this experiment running and when someone will give us his email address, he will be automatically added to our welcoming drip campaign. He will become more familiar with the content I write and it will be easier to make him do higher investments in the future (like sharing an article, asking him to upvote my content on a HackerNews-like website or simply make him buy a product).

Here are a few optimizations we can make to go further.

  • We will capture as many emails as possible when new visitors drop on our content. We could implement exit-intent popups like SumoMe ListBuilder or simply WelcomeMat.
  • We could optimize our opening rate by testing different titles, like adding an emoji at the end (which I did) or create more curiosity. But, we will always keep our emails short and in plain-text.
  • We could use Mixpanel or Amplitude to analyze our cohorts and figure out more in depth what happens on our product with people who received this welcoming drip campaign.
  • We could add that we are available to answer questions at the end of our emails to increase our response rate. This will enable us to build an even stronger relationship with our subscribers (resulting here, again, in a higher commitment in the future).
  • Personalization in drip campaigns usually leads to huge improvements. People always tell to go niche and it is true. The product itself must be super niche but if you can get the marketing to be super niche too, you will basically have the same effect. So, don’t be afraid to ask for additional pieces of information if it helps to send better emails.

Feel free to add a comment to tell me what you think about this, ask me questions or to give us more optimization ideas.

Thanks to Nada Rifki for reading drafts of this experiment.

23 y/o College Dropout. Former Growth Hacker at TheFamily. Video Editor at Growth Hacker TV.

  • Super article comme d’habitude Julien!

    Petite question, est-ce que tu as A/B testé ta signature manuscrite? Si oui, quels ont été les résultats?


    • Nop je n’ai pas testé mais je pense qu’en mettant une signature textuelle on augmente legèrement le taux de clic 🙂

  • tu as fait une welcome sequence plutot pas mal.

    Pour l’aspect que texte je suis pas certain personnellement. Je pense que tu peux faire un email un peu HTML mais pas trop. Après tout s’A/B test ;-). et ca dépend du service et de la population.

    La question a se poser sur une séquence d’acceuil c’est qu’est ce qui est important pour générer de l’activation/rétention.
    Sur un blog, c’est facile, c’est la lecture d’autres articles et le partage. Voir l’abonnement feedly.

    Je sais plus trop quel gros blog fait ca mais dans le welcome email, il met le lien vers ses 10 articles les plus populaires. Et c’était une erreur car il vaut mieux envoyer plusieurs emails avec à chaque fois que un article.
    Le seul soucis, c’est que un mec qui arrive sur ton blog et qui lit quelques articles, tu risque de lui pousser un article qu’il a déjà lu mais bon c’est pas très grave…

    En onboarding, on retrouve souvent les mêmes actions:
    – activer l’email pour vérifier que l’email est bon
    – ajout de photo (pour les réseaux sociaux et les annuaires)
    – explication de la value proposition/le concept (pour un blog pas necessaire)
    – partage/referral qui peut arriver early dans le welcome (la viralité early ca marche)
    – enrichissement de profil (réseaux sociaux surtout)
    – commencer à participer (selon ce qui défini un profil actif)

    Bref, on explique, on active et on invite à entrer dans la partie active des users….

    Un article intéressant sur le sujet:

    • Merci pour l’article Olivier! Oui tout a fait d’accord pour le lien avec les 10 articles les plus populaires. Ce n’est pas valable uniquement pour l’emailing mais le mieux c’est de toujours faire un email = une action. Sinon les gens sont confus et cliquent beaucoup moins.

      Pour ce qui est du HTML, je trouve que les taux de clics sont toujours inférieurs à du plain-text (surtout pour un blog où tu veux créer une relation de proximité entre l’auteur et ses lecteurs). Donc j’ai tendance à être radical et à ne jurer que par ça mais oui il y a des contre-exemples selon le service et la population. Si tu essayes de vendre un truc plutôt visuel (Afrostream ou Netflix par exemple), ça va être très difficile de ne pas utiliser de photos. C’est d’ailleurs pour ça que leurs campagnes publicitaires à la radio fonctionnent bien moins qu’à la télévision ou dans le metro.

  • Guy Marion

    Great post Julien – thanks for sharing!

    Another thing to try is to include a data enrichment email early in the drip, such as “are you most interested in learning more about: a) SEO or b) product growth or c) integration/online channels.” (topics could also be related to product interests, SMB vs. Enterprise, B2B vs B2C, etc.)

    Add a UTM to each of these links, then create a smart segment in Autopilot that is specific to each of these UTMs / interests. It’s another easy way to incorporate their preferences into your content drips, and drive better results through #personalization.

Next Post