About the author

Jeff Chang (@JeffChang30) is a growth technical leader at Pinterest and angel investor. If your startup is looking for an angel investor who can help with all things growth, please send over an email!

Introduction

Signup and login conversion is the art of building new user intent and converting intent to authenticated users. Since every user starts out as a new user, this part of the user flow is very important to optimize. In order to actually grow your service quickly, signup rates must be high. In this post, I’ll talk about how to start growing signups and logins into a large growth channel. To summarize, in order to set yourself up for success, you must understand unauth user mix, understand how signups and logins interact, and build data tools. Then, I’ll talk about a few different areas of optimization you can try.

Setting yourself up for success

Understand unauth user mix

In my experience, the best way to think about unauth users is to realize that there is a wide distribution of product comprehension and intent, so we should create flows that perform decently well with all users and very well with the most common users. The user type distribution varies a lot product to product. Here is a list of different types of users that can come across your unauth surfaces, and the general strategy that should be applied to each:

First-timer: Users who don’t know anything about the product -> educate them about the product
Skeptic: Users who have heard of your product, but do not have enough intent to sign up yet -> sell them on the value proposition of your product to increase their intent
Ready-to-signup: Users who have heard of the product and are ready to sign up -> make it easy to find the sign up flow and go through it
Dormant user: Users who have already signed up in the past, but are not core users -> help them remember their login credentials and avoid duplicate accounts
Returning core user: Users who have already signed up and want to log in -> make it easy to find the log in flow and go through it

There is a limited amount of product surface, so you have to find the right balance between these strategies to optimize for your user mix. Of course, your user mix will also change over time, so the optimal strategy will change over time.

Example: Thinking about user mixes on the home page

At new companies, the home page is usually a large driver of signups and logins, so let’s talk about the home page. If you look at some sample home pages such as Pinterest, Facebook, or Linkedin, you’ll notice that their home pages are very simple, and don’t tell you much about the product. It may be tempting to just copy what these high growth companies do, but this probably won’t work for most new products. This is because most people already know about what these big companies are and have probably signed up for them in the past, so most users fall into user categories dormant user and returning core user. If your product mainly receives these kinds of users to the home page, this is the kind of design you want: very prominent way to log in, and few distractions.

You’ll notice that we do have a red tooltip that links to an education page, and that’s because we want to also cater to the few first-timers that may visit our site, even though they may not be the primary focus.

If your user mix also contains a lot of first-timers, this probably won’t work and you’ll need to have way more education. One product that has a nice education flow on the home page is tumblr:

Experiment with signups and logins together

One common mistake that occurs with new growth teams when growing signups is that they are grown without thinking about logins. At Pinterest, we didn’t start optimizing logins until a few years after we started optimizing signups, and realized that we missed out on some huge opportunities. Also, by optimizing just signups, there was a large amount of duplicate signups that were occurring. So, every unauth experiment should measure both signup and login changes. Sometimes, you will see results like signups up and logins down, or logins up and signups down, and if the magnitude is similar it usually suggests a duplicate account issue. Usually, you would want to go with the experimental group that has more logins but less signups to reduce duplicate signups.

Building the right data tools

Compared to SEO, you don’t need a special AB experiment framework to successfully run signup or log in experiments; a normal AB experiment framework will work, as long as you measure both signups and logins. For signups and logins, the two most important tools you will have to build are landing surface segmentation charts and real time monitoring charts. I’ll go over how to make great monitoring charts in the next blog post and focus on the segmentation charts. For surface segmentation charts, you should have sessions, signups, logins, and conversion rates available for segments such as referrer, landing page, and country, so it will look something like this:

This is a very powerful tool for various reasons. First, it helps you keep track of historical changes that you can correlate to product changes or external events. For example, let’s say a certain page broke in a certain country. Without this charge, all we see is a signup drop, but it would be difficult to debug. By toggling the segmentation, we can quickly determine what country and page broke, and fix the bug. Next, it helps you easily determine opportunity sizes. For example, if somebody had an idea to increase conversion on a certain page, and the certain page didn’t receive much traffic, then maybe that idea shouldn’t be prioritized.

Areas of optimization

Simplify the flow

Users who are going through signup flows might not have a lot of intent because the product is new to them. The slightest roadblock could possibly cause them to give up. One common problem here is too many steps. Frequently, products need a lot of information from the user to give the user a personalized experience, but sometimes people don’t feel comfortable giving information away, or feel like the signup process is too long. If your service requires a lot of fields just to operate correctly, consider breaking up the fields into multiple pages and make the user feel like they’re making progress. Another common problem is including a captcha. This causes some users to drop off because it is too much work and some captchas are actually pretty difficult. Instead, consider post registration filtering or requiring an email confirmation from a valid email.

Experiment with all copy

Every single line of copy in your sign up flow should be copy tested. I can pretty confidently say that over 80% of copy in Pinterest unauth has been copy tested, and the only ones that haven’t are usually in new features.

Use social signups if you can

As you can tell from the previous screenshot, Pinterest uses social signups. In general, adding social signups will increase your overall conversion rate. However, there are some issues that you will have to deal with if you use social signups such as account duplication and decreased security. It’s good to have your social buttons to act as both signup and login buttons. When you get a response from Facebook or Google after they authenticate with those services, you should check if that user has an account already, and sign them up with those credentials if they don’t, or log them into their account if they do. Because social signup is so important to Pinterest, we even work with both Google and Facebook to suggest and beta test new authentication features. One then-new feature that I personally beta tested is the personalized “continue as” Facebook login button, which increased Facebook signups by over 30%. You definitely should try if you are using Facebook authentication.

Test different upsell timings

There are a lot of different timings that you can test with your signup upsell. You can pop it up immediately (although that probably won’t convert well), or after showing some content, or way down the line after the user has seen X articles. When testing timings, you have to understand how user intent changes over time as they browse through your product. For example, for a lot of users, they do not have enough intent to sign up right when they land on your page. That’s why modals that pop up immediately usually do not perform. As the user browses your products and learns more, intent can increase, but the longer you wait, the more users that leave. So, you want to find the sweet spot where users have enough intent but not that many of them have dropped off yet. This varies product to product, so I advise testing a variety of upsell timings.

Conclusion

Signup and login conversion is an area of optimization that can have tremendous impact on your growth, and is vital to every growing company. In this post, we just covered the thought process behind user mixes, basic data tools, and a few areas of optimization, but because this is such a large area, more will be covered in later blog posts. Having worked on signup/login conversion for years, it’s interesting to me that the space continues to change and we continue to learn new things that increase conversion rate. It’s a constant learning process. I hope from this post you got a general sense about how to approach signup and login conversions!

Takeaways

Understand your unauth user mix
Build a conversion segmentation chart
A few optimization opportunities are: Simplify the flow, experiment with all copy, use social signups, and test different timings

Want some advice with optimizing your signup and login rates? Email me at jeff@growthengblog.com

Scaling new growth opportunities series

Part 1: Intro to growth engineering

Part 2: SEO - Get started 10x'ing your traffic

Part 3: Signup and login basics

Part 4: Growth monitoring done right

Part 5: Start your journey to the top of the app store - mobile acquisition basics

Signup and login basics - learn how to get started with conversion optimization