BSSA #66 - The Wide Event & Mixpanel Tips 🚀

Hello everyone! 😁

I have a special announcement!

This year, Shopify announced that Shopify Unite would be virtual. It’s usually the best time to network with others in the ecosystem.

So I decided to host my own Shopify Event! 🚀

On Saturday 16th of September, from 6 pm to 11 pm, I am hosting The Wide Event, a Shopify event for partners and merchants.

150 people are expected! And we are doing it in Paris (but it will still be an English event!)

The exact location is still secret but it’s a famous place where I always wanted to host an event! 👀

And Shopify will also share this event as I’ve been in touch with them!

I can’t wait to meet you all!

The tickets will be for sale soon and I will send a special email when it’s live. 🎟️

150 people is not a lot and tickets will probably sell quickly but I wanted to start small!

We are also looking for sponsors for that event!

Only 1 slot left for sponsors!

If you sell products or services for Shopify partners, it’s your chance! (Also if you sell to merchants but it will be mostly partners at this event).

• Get your brand in front of 150 people

• Talk about your product or service on stage

If you’re interested, reply to this email and I will send more details!

Mixpanel Tips for a specific use case

Last week I was looking at July’s numbers for WideBundle. We had many installs during that month but not that many new paying users.

So I was wondering why?

First step: Look at the number to validate that problem

I had the feeling something was wrong, we had installs but not many paying users, so I had to prove that.

Because guessing is not enough.

So I listed the different numbers to show that:

• Number of installs during that month

• Churn that month

• Conversion Rate during that month

Basically all the metrics of the funnel after installation.

Installs were higher so I was right. Churn was the same as last month. And then I checked the conversion rate, it was lower!

The conversion rate was still high, but lower than previous months. Now I know the wrong metric.

Second step: Make assumptions

Once you have your number that is too low (or too high based on what you’re looking at) you need to make assumptions.

What could be the potential reasons? The more reasons you can find, the better.

I found these:

1. The new settings we added in July are too complicated to understand so people uninstall

2. It’s summer so conversion rates are lower

3. The new acquisition channels we started attract the wrong target

Third step: Validate these assumptions

Once done, you need to test each assumption by looking at the numbers on Mixpanel.

For each you have to list the things you have to check on Mixpanel that could validate or not these assumptions.

Here we are for the 3 assumptions:

1. We’ll look at the % of people who saved the settings and the time it took our users to save them. We’ll then compare that number with previous months.

2. We’ll look at previous years to find the same pattern and track the numbers again in August and September to see if it’s back to normal.

3. We’ll create cohorts: one group with users who came through these new channels and one with others. We’ll then compare the conversion rate to see if the new acquisition channel has a lower conversion rate. We can also compare the conversion rate with previous months

Fourth step: Iterate

If you can’t validate your assumptions because everything seems fine, you need to find and test new assumptions.

When you find the reason, you can make some changes (iteration) and then test it again in the coming weeks!

That’s it for today’s email!

I hope you enjoyed it! Feel free to share my newsletter with someone else: https://news.matdesousa.com/

And let me know what you want me to talk about in the next email!

See you next week,

Mat