Étienne Garbugli
No matter how great your product is, it’s very likely that 40–60% of your free trial users won’t see your product a second time4. (Location 127)
Email, to this day, remains the best channel to build relationships and bring users back into your product. (Location 132)
To be successful, you have to go from list-thinking to database-thinking —understanding that contacts have different experiences and attributes. They may differ in terms of their signup dates, their levels of engagement, languages, preferences, spend, subscription plans, goals, worldviews, etc. (Location 157)
send the right message at the right time to the right people; go from list-thinking to a database and relationship marketing mindset; move away from one-off campaigns to workflows; and transition from “one-size-fits-all” to tailored and personal communications. (Location 162)
We looked for patterns worth targeting across signup cohorts (groups of users defined by their common signup period) and segments. This allowed us to identify engaged users, churners, users with a lot of one-off product purchases, and so on. We were able to find literally dozens and dozens of patterns worth targeting. Then we used proven upsell messaging to test the waters. (Location 194)
Users are only in one main state at a time: Onboarding OR Paid subscriber OR Churned. (Location 206)
In the train track approach, each user is on the most appropriate train track based on his or her behavior. At each “station”, we, the business, make an assessment of which communication program is best for their current behavior. When users change status, they go on a different track. As an example, video analytics company Wistia has three tracks just for trial users: onboarding Aha moment achievement conversion At LANDR, we had tracks for: onboarding paid subscribers onboarded, non-subscribers churners reengagement (Location 208)
All of these programs had clear start and end points, and their own unique objectives. For onboarding, it was activation, the moment when users became active users of the product. For paying subscribers, it was retention. For churners, it was re-subscription. (Location 213)
Users or customers sign up. They go through product value discovery. They are either convinced of the value of the product, or they’re not. If they are, then they upgrade, are retained, and (hopefully) end up recommending the product. If they aren’t, they cancel and they churn. (Location 226)
Signup: the initial account creation Activation: the Aha moment when the user perceives the product as valuable Conversion: when a user agrees to swipe their credit card and purchase the product, or when a colleague does it for them Re-Purchase or Retention: when users willingly pay for a second month, or a second year of subscription Referral: when users refer your product to other prospects (Location 231)
Breaking down the SaaS customer experience into these milestones helps create clear start and end points for each track. It helps to understand that the next milestone after signup is activation, and that the first milestone for a paid subscriber is retention. (Location 234)
The 6 Email Sequences You’ll Need (And the Key Emails within Them) (Location 239)
Signup, Activation, Conversion, Re-Purchase, and Referral. We know the milestones, now what? (Location 240)
Let’s look at these different sequences and how to handle them specifically. (Location 246)
product. If you do, the goal of your first sequence will be to convince these subscribers of the unique value of your product. For this sequence, you have to balance the value of what the prospects actually signed up for (content updates for example) against pushiness. (Location 249)
For this phase, your goals are to help establish product value and to ensure users achieve their desired outcome with your product. (Location 254)
If your users don’t convert and decide to stay on the free plan or give up on their trial, you can give them more reasons to upgrade with value expansion emails: ‘Hey, our product also allows you to do X’. It’s always a good idea to eventually try to close them again, approaching it from a different angle. We will see how to do that in the Upgrade, Upsell & Expansion Revenue deep dive later on. (Location 264)
Look at consumption, and when an account is within 20% of maxing out a key value metric—what and how you’re charging—send an email with an offer to upgrade their account to a more expensive plan. Experiment with discounting, and see how it goes. (Location 281)
To successfully get referrals, it’s often a good idea to run Net Promoter Score (NPS) surveys, standardized surveys assessing how likely users are to recommend your product. (Location 286)
Figure 7.3 – Key Goals & Emails (Location 292)
You can use the template at saasplaybook.co/traintracks to start defining and mapping your train tracks. (Location 295)
If you haven’t already done so, I strongly recommend organizing series of interviews to understand your product’s new signups. (Location 303)
To do so, look at signups from the past few months and create three buckets: Your best customers: the top 1% in terms of engagement and revenue The next best: customers ranked within the top 2–10% Your worst customers: the bottom 10% (Location 304)
Reach out to users in each of these groups. Schedule 20-minute discussions, either over the phone, or face-to-face if it is possible. You should do Switch interviews based on Jobs to be Done theory—the best interviews to understand the decisions leading up to using, or canceling a product. (Location 312)
Understanding Segmentation Data (Location 331)
There are four main ways to segment users: 1. By User or Buyer Persona (Location 334)
There is always a balance to be struck between asking for more information (more fields tend to mean less completions), and getting people to sign up and act quickly. It’s especially important to be aware of the friction that you may be adding by capturing information before a user is actually convinced of the value of your product. (Location 349)
In reality marketing emails are rarely the only ‘emails’ that users and prospects receive from these businesses, and in the eyes of users, all emails add up. For this reason, it is important to consider the totality of end user communications when assessing the current situation. (Location 373)
Go through your entire business. Make sure you have a clear understanding of all the ways users and prospects get contacted by your organization. Unless you replace them, anything you add will be on top of the communications already in place. Try to get a clear picture of the customer journeys: all the way from first-thought (awareness) to signup, and beyond. The customer experience doesn’t start at signup. It starts when prospects are first exposed to your brand, be it through ads, search results, or product reviews. (Location 382)
Product-Led Growth author Wes Bush recommends printing out each step and every action that users need to take in your product, and asking: What steps can be eliminated? What steps can be delayed? What steps are mission-critical? All communications, from receipts to support follow-ups, should help drive performance. It’s really important to evaluate everything in the context of your goals. (Location 396)
Here is an acceptable Minimum Viable Analytics Product (MVAP) for many SaaS businesses: (Location 411)
Creating Key User Segments (Location 462)
Before starting to write emails, you’ll want to create key user segments. Those could be: people who haven’t signed up for your product (if the required data is available); people who signed up today; people who signed up in the last seven days; people who signed up in the last seven days, but didn’t engage, or didn’t activate; people who signed up in the last 30, 60 or 90 days and activated; inactive users; users whose trial is about to end or just ended and that you would eventually like to convert; paid subscribers in their first month; paid subscribers retained for two months or more; subscribers on annual plans; users who you think would be willing to refer your product to others; subscribers who cancelled; subscribers who cancelled more than once; or signups per specific acquisition channel. (Location 465)
Go through random profiles in each of these segments and compare with the data from your database. Are those the users you’d expect to find in each of these segments? Any issues? You want to uncover issues with the implementation or your segmentation as early as possible. (Location 474)
Identify issues, adjust, and refine. (Location 478)
It’s always better to be the one setting the rules, than to have them defined for you. The rules that you will want to define will vary depending on who owns email marketing, and the responsibilities of each teams within the organization. (Location 485)
Sit down with the product, brand, marketing, analytics, finances, design, and support leaders. Get alignment and share back the rules for email marketing within the organization. You can use the template at saasplaybook.co/rules to get started fast. (Location 499)
Focus on learning the answers to these questions first: Is this the right email? Are we sending it to the right people? Does the timing work? Can the copy be improved? (Location 525)
What areas are most troublesome in your business right now? What metric are you expected to move with email? Is it: Engagement? Retention? Conversion? Revenue? Signups? (Location 535)
Researching Email Copy & Designs Here’s the trick to significantly improving your SaaS email marketing skills—you have to become a student of it. This means you should: Start collecting great email copy, CTAs, and designs. Understand the objective behind each and every email that businesses send. Try to understand the rationale behind copy, link, and design decisions. There are great websites like Really Good Emails11, Good Email Copy12, and Good Sales Emails.com13 that you can use for your research. These sites categorize email copy and designs by types. (Location 564)
Those include, among others: Drift MailChimp Pipedrive Shopify SurveyMonkey Trello Wistia Zapier (Location 572)
Alternatively, you can subscribe to paid services like Mailcharts14 or Mailody15. These services will help you track and understand your competitors’ email programs. (Location 585)
Email Sequence Pacing & Structure Ideally with emails, you’d like to send the absolute fewest number of emails, and achieve the maximum results. (Location 591)
To structure your email sequences, start with the end goal. Is it conversion? Activation? Referral? Reengagement? The sequence you’re working on should only have one main goal. This will help create focus. (Location 599)
You have to demonstrate how the readers will benefit. Don’t talk about features or your company. Try to use more “you” than “I” or “we” to make your copy more relevant to your user’s life. (Location 625)
like to use the nod test for this. You want users to nod in agreement as they read your email: (Location 629)
Figure 20.1 – Drift’s Feature Launch Email (Location 635)
Zapier – Upgrade Email Zapier helps automate API integrations. A day before the end of my 14-day trial, I received the following email: (Location 640)
Podia – Upsell Email Podia helps creators sell online courses, digital products, and memberships. As a subscriber on the monthly plan, I received the following email: (Location 649)
Illustrate product benefits and value with clear examples that relate to the unique situation of your users. (Location 668)
Use clear actions: Use a CTA that clearly establishes the next steps. Repeat it throughout the email, coming at it from different angles. Use the P.S. to attract the eye and to reinforce the action you want users to take (when appropriate). Keep your emails simple and your messaging scannable. It’s important for users to be able to get the email at a glance. Short and sweet often outperforms long and complex emails. You want a near-instant reaction from your readers. (Location 673)
A good email has to: capture attention through the subject line, personalization, or a story; build reader interest by demonstrating either the benefit or the problem; build desire to act by creating information gaps, time constraints, or the fear of missing out; and drive action through a well-timed CTA, telling users exactly what you want them to do. (Location 679)
These are really just the four steps of the AIDA model18 (Attention, Interest, Desire, and Action) applied to email copywriting. (Location 682)
But in reality, open rates are influenced by a combination of factors. Figure 21.1 – Open Rate Components The subject line, the preview text, the sender’s name, the sender’s email address, whether the email fell in the inbox or the promotion tab, your brand, the trust you’ve built over time, the number of emails you send/have sent, the timing, and the competition for your user’s inbox that day all factor into that user’s decision to open (or not) your email. (Location 690)
When it comes to subject lines, you want to evaluate: (Location 694)
As we have seen, good subject lines can trigger curiosity with mystery. This can be done by asking questions, or using cliffhangers. For example: (Location 703)
The average user only spends about eight seconds reading an email20. The subject line has to achieve two things fast: stand out in the inbox (e.g. I noticed it and want to open it), and properly set up the email copy. (Location 715)
Think of the open rate holistically. Look for ways to get more opens by: (Location 719)
Come up with dozens of ideas for subject lines and run them through Send Check It22, a free tool designed to help test subject lines. It’s the best predictor for subject line success based on the data we collected23. (Location 726)
If you are having difficulty coming up with good subject lines, you can download my subject line pack at saasplaybook.co/subjectlines, or use Neville Medhora’s Subject Line Generator Formula24. The Subject Line Generator Formula gives you over 100 patterns to explore. You can even input your own topic. (Location 729)
Email is great when the user isn’t currently using your product. It’s great to drive them back in, but when they are right there using your product, you can’t expect them to be checking their emails at the same time. (Location 742)
At LANDR, we often followed feature launch emails on-site with In-App messages. This helped to keep communications simple and goal-focused (one goal per message). The email was about getting people in the product, while the In-App message was about getting them to engage with the product. (Location 754)
Automation platforms like Intercom, ActiveCampaign and HubSpot generally allow you to combine messaging types. (Location 756)
For campaigns, a trigger, or entry condition will be required. You want to understand what qualifies a user to enter this campaign. (Location 769)
Was it: Sign up for a free trial? Trial expiration? Sign up for a paid plan? The cancellation of a subscription? A certain action that was taken for the first time (property changed from 0 to 1, or an event happening for the first time)? Did the user achieve a certain state (matched conditions or entered a segment)? Did the user do something a certain number of times (ever, or over a certain period)? or Did the user not do something over a certain period? (Location 770)
For timing, use the number of days since the signup date. Segment using the time since the last email when possible. Use clear filters, for example whether the user has received such and such campaign. Are there gaps in your automation data? Are you using first name or greeting personalization? Do all your users have those fields populated? (Location 795)
It was a simple experiment: We were testing a one-month upgrade to a paid plan with roughly 500 users. (Location 839)
The goal: Do you have a clear way to track performance? How will you evaluate campaign performance? (Location 858)
You can use the checklist at saasplaybook.co/presend to do your own spot-checks. (Location 860)
SaaS content marketer Stephanie Knapp recommends asking25: Does it have a clear objective? Is it customer-focused? Is it attention-grabbing or memorable? Does it make sense? Is it unique? (Location 866)
You will want to move away from aggregate results (e.g. 10,000 emails sent, 30% opens, 5% clicks) towards looking at trends and performance over time. How are your campaigns trending between dates? After the last changes you made? By segments? The stats you need to track for your campaigns are: (Location 900)
You can use the template at saasplaybook.co/reporting to get started fast. (Location 917)
Overall, you should also monitor: (Location 928)
To track deliverability and abuse reports, you can use free tools like Sender Score29, or Litmus Spam Testing30 and Email on Acid Spam Testing31. (Location 931)
You need to ensure that your unsubscribe rate remains under 0.5%32. If your unsubscribe rate is over that benchmark, check your users’ profiles to see how many emails they receive (transactional, one-offs, and automated combined), and assess the relevance of those emails. (Location 939)
As you learn about your business and your email program, the gaps will become more and more obvious. In what areas is your business struggling most: Retention? Revenue? Conversions? Getting feedback? Can email help achieve those goals? (Location 961)
In SaaS, since your users go through the same campaigns, any improvements you make at the start of the funnel improve the end of funnel as well: (Location 978)
Depending on the organization and the performance, we often saw three to four distinct groups of performance: (Location 1001)
Benchmarking against your own previous performance is the only way to get a clear picture of the potential of your email program. If one of your drip emails opens at 20%, but all of your other drip emails open at 40%, you can usually tell where the problem is. (Location 1009)
The sender’s name: A lot of experiments can be run around the sender’s name: Personal vs Company, Male vs Female, Common name (e.g. Matt) vs Atypical name (e.g. Jorane), etc. Having a clear picture of your target audience can help inspire tests for different campaigns. (Location 1132)
To help you get started, here are ten experiments you can try: (Location 1138)
I personally use Send Check It. It’s the tool that we’ve found to have the most accurate data49. It integrates with Slack and it’s very easy to use. (Location 1151)
What you are looking for is the percentage of users that accomplish the goal when they land on this page (or state) from all channels. Then you want to compare this percentage against the percentage by channel (e.g. email vs all other channels). (Location 1207)
I’m not going to go in details for app page flows, because these will vary depending on your app, but the following can apply to your site landing pages, pricing pages, and conversion flows: (Location 1225)
If we take a closer look at the flow: Users either get acquired through search, content marketing, or via cold emails. They sign up from the blog, the home page, the pricing page, or sometimes random pages like the Terms & Conditions. They can drop off at any one of the three steps of the setup process. (Location 1640)
A really good conversion rate for free-to-paid freemium products is 4%. That’s the conversion rate achieved by Dropbox and Evernote, while best-in-class products like Spotify get over 26%102. (Location 1688)
We first set up an upgrade email at the 14-day mark. To maximize conversions, we set up reminders a little later for the users that decided not to upgrade. We also added upsell emails for users that had churned in the past, depending on their cancellation reasons. Once those campaigns were in place, we began work on a much more elaborate upsell program to monetize past cohorts. We set out to identify the behaviors that often lead to subscription purchases: Are the people mastering five times a week more likely to upgrade? Are the people who use this or that feature more likely to upgrade? Do the people who use this file format upgrade? Looking at the data, we identified dozens of segmentation patterns worth targeting for upsells. We added each and every one of those to our experiment board. (Location 1705)
We came to the realization that behavioral targeting—by picking behavior patterns—wouldn’t allow us to scale the program. We eventually reworked the entire upsell program around the concept of train tracks introduced in Chapter #5. (Location 1718)
Freemium users now had their own train tracks. They received different upsell emails every few months. Between each attempt, we built up their perception of the product value. Each upsell email users received was based on their actual behavior. Every week the upsell program converted users across 20+ monthly cohorts. Sales and performance increased, and soon enough, ARPU per cohort was growing. To cap it off, we also used transactional emails to promote subscriptions: “You just mastered [ X ] tracks. You can get unlimited mastering at a better quality for just the cost of [ Y ].” Monetization is key in SaaS. It’s a way to confirm that your product is valued and valuable. And it’s a way to fund business growth. To successfully upsell users, you have to find the right time to sell a subscription and make an appropriate ask. (Location 1720)
Some people will upgrade just because you ask. With the number of free products on the market, it’s possible that some users never realized that your product had premium functionalities they could benefit from. (Location 1729)
Like Zapier, you can use loss aversion104 to your advantage. People are inherently afraid of losing something they already have. (Location 1732)
To make your email successful, suggest the appropriate subscription based on a user’s needs or usage. (Location 1735)
Stop upselling once users subscribe. Add them to your Retention sequences, or the most appropriate train tracks based on your program. Re-establish your product’s value for non-buyers. Are they still using your product? Work on overcoming any objections, show more (or different) product value, try again to upsell them using a different angle. (Location 1740)
Different types of users will respond to different types of pitches. These might include: quantification of value (e.g. you’ll save $700 per year); a hard push; discounts; or case studies (etc). (Location 1743)
Such signals can help you to understand who within your user base may be more likely to upgrade. Once you have identified some clear signals, you can start creating a program, for example: Day 122: Upsell attempt Day 182: Upsell attempt (+60) Day 242: Upsell attempt (+60) Day 302: Upsell attempt (+60), and so on (Location 1757)
1. Annual Subscriptions The easiest, and the most direct way to improve retention is to increase the percentage of annual subscribers. More annual subscribers means more cash up-front, and a higher customer lifetime value (CLV). (Location 1824)
This makes sense. Sam Levan, Co-Founder & CEO of MadKudu, explains98: “A free trial creates artificial purchasing urgency. But there isn’t anything magical about the last day of a trial—some customers continue to convert at their own rate based on incentives or their perceptions of value.” The first 30-40 days of your user cycle are crucial99. But you definitely want to follow up for the first 90 days after signup, well after the trial ends. As long as your onboarding emails add to the overall performance and usefulness of your series, you can keep adding and pushing for activations. (Location 1620)
Let’s look at it from Highlights’ perspective. At a high level, the funnel100 looks like this: (Location 1635)
The overall goal is to get users to activate—that is several clients, and several weekly uses—in under a week. Although this high-level funnel gives you a good overview, it can be highly misleading. There are actually many sub-steps a level below where usage breaks down. If we take a closer look at the flow: Users either get acquired through search, content marketing, or via cold emails. They sign up from the blog, the home page, the pricing page, or sometimes random pages like the Terms & Conditions. They can drop off at any one of the three steps of the setup process. When they do reach the home screen after those three steps, the results sometimes vary. (Location 1638)
Do users from different traffic sources behave differently? Are there countries, regions, or cities that perform better than others? Can users on mobile, tablet, or desktop complete the same tasks? Were there considerations in terms of site performance? Were there issues with payment, or anything else you haven’t foreseen? (Location 1648)