He Tested 4,700 Paywalls. These Won. — Transcript

Full transcript

1. 0:00I spoke to someone who's designed 4,700
2. 0:03paywalls.
3. 0:03>> How important is paywall optimization?
4. 0:06>> As a starting point, it's honestly not
5. 0:08that important.
6. 0:09>> And the highest converting paywall isn't
7. 0:12always the one that you should use.
8. 0:14>> Unfortunately, it works.
9. 0:16>> So today, we are unpacking the biggest
10. 0:19growth lessons from thousands of paywall
11. 0:21experiments and why most founders are
12. 0:24optimizing the wrong thing.
13. 0:26>> For your first paywall, honestly, all
14. 0:28that you need to do is to have a
15. 0:29headline describing the product, a
16. 0:31couple bullet points describing the
17. 0:33product and what it does, and a button
18. 0:35to continue. I get in calls with some
19. 0:37people that think they need to do all
20. 0:39these crazy optimizations for their
21. 0:40placements and paywalls in the
22. 0:42beginning. You really don't. You just
23. 0:44need to make sure that you're getting
24. 0:45users into your app first, and then you
25. 0:46can optimize that traffic later.
26. 0:48>> When you run a paywall experiment, how
27. 0:50do you decide what's worth testing
28. 0:52first?
29. 0:52>> Design tests move the needle the most.
30. 0:55So, you try radically different designs
31. 0:57until you find something that blows
32. 0:59everything out of the water. You could
33. 1:01have a video paywall, you could have
34. 1:02bullet list paywall, a trial timeline
35. 1:04paywall, a long-form paywall. After
36. 1:07design, personalization moves the needle
37. 1:09the most. Calling out their name,
38. 1:11highlighting their goals. [music] So, if
39. 1:13they're male or female and it's a
40. 1:14fitness app, you can show like a strong
41. 1:16man on the paywall and it's
42. 1:18representative of the user's identity,
43. 1:20and then price packaging design. So, how
44. 1:23do you design the footer of the paywall?
45. 1:25If you have a yearly and a weekly
46. 1:27product, are they vertical product
47. 1:28groups? Are they horizontal? What do you
48. 1:30put in those product groups? How do you
49. 1:32present the price? Do you show it as
50. 1:34weekly for both options with a subtitle
51. 1:37that shows the actual real price? Or how
52. 1:39do you go about that? Ideally, you want
53. 1:42to default to a yearly product because
54. 1:44we see that that has the highest LTV.
55. 1:46You only want to show two to decrease
56. 1:49cognitive load on the user, but a yearly
57. 1:51product may not be the best for your
58. 1:53business. You can also start doing tests
59. 1:56based on what kind of device the user
60. 1:58has. If the user is on an iPhone 17 Pro,
61. 2:02they're on 5G and don't have low battery
62. 2:05mode, show them a higher price as
63. 2:07opposed to an a user that's on an iPhone
64. 2:0913 Pro on 3G and their low battery
65. 2:13indicator is working, maybe you want to
66. 2:15show them a lower price because they may
67. 2:17or may not convert at the higher price.
68. 2:19But, that's like one of the last levers
69. 2:21that you really pull because it's so
70. 2:23much work. It incrementally gets smaller
71. 2:25and smaller in what moves the needle.
72. 2:27>> What do you think are the metrics that
73. 2:29matter the most when you're evaluating a
74. 2:31paywall experiment?
75. 2:33>> That depends entirely on the business. I
76. 2:35used to use ARPU as my North Star
77. 2:37metric, but ARPU doesn't necessarily
78. 2:39give you all the answers. Your ARPU may
79. 2:41be high, you may have three or four
80. 2:43dollar ARPU, but the reason you may have
81. 2:45three or four dollar ARPU is because you
82. 2:46have weekly products, and so your churn
83. 2:48may be 90 to 95% per month. Even though
84. 2:52ARPU is high and it looks good, it
85. 2:54doesn't tell you the full picture. I
86. 2:55typically look at paywall percentage,
87. 2:57how many people are actually looking at
88. 2:59the paywall that installed the app. If
89. 3:01it's above 80%, you're doing good. If
90. 3:04it's 90% or above, you're crushing. I
91. 3:06look at trial to paid conversion rate,
92. 3:09LTV, D1, D7, D30, churn, and then based
93. 3:13on the kind of customer that they are,
94. 3:15I'll suggest different things. If you're
95. 3:17vibe coding an app and you just want to
96. 3:19build something and launch it and get as
97. 3:20many users as possible, then you're
98. 3:22going to be focusing on ARPU as the key
99. 3:24metric. But, if you're building like a
100. 3:26long-term business, you want to maximize
101. 3:28revenue while still balancing churn,
102. 3:30then optimize for LTV and retention.
103. 3:37We take your paywall and use that as a
104. 3:39control, and then we would design four
105. 3:41to five paywalls to go against it. For
106. 3:43all of our clients so far, we are able
107. 3:46to beat the control in the first week
108. 3:49for every app except two. It ended up
109. 3:52taking until the second or third
110. 3:55experiment to really see the
111. 3:56statistically significant win for them.
112. 3:59When you're doing experimentation, you
113. 4:00just have to be patient also. It's not
114. 4:03always going to work out.
115. 4:04>> How about the paywall touch [music]
116. 4:05points? How important is it to test
117. 4:07them?
118. 4:08>> All the different touch points are
119. 4:09extremely important. Onboarding
120. 4:11paywalls, gated paywalls that gate some
121. 4:13kind of feature, paywalls in the
122. 4:15settings page, paywalls for if the user
123. 4:18wants to cancel their account, you can
124. 4:20win them back showing them some kind of
125. 4:22discounted offer. Uh a paywall for when
126. 4:24they're about to delete the app. You'll
127. 4:26say, "Before you go or before you
128. 4:28delete, we'd love for you to try one
129. 4:30last time." When it comes to what is the
130. 4:32most used paywall, it's the after
131. 4:35onboarding paywall.
132. 4:36>> In fact, 78% of iOS apps on Mobbin show
133. 4:40the paywall after onboarding.
134. 4:43>> Now, do you create an account before or
135. 4:45after? That you have to test actually
136. 4:47because I have tested it multiple times
137. 4:50and I still have inconclusive results
138. 4:52and I think it's just because it's all
139. 4:53been in different app categories. I like
140. 4:55to create an account first and then show
141. 4:57the paywall. That way I can tie a user
142. 5:00ID to the person that's actually
143. 5:01subscribing. So, you have to think about
144. 5:03the entire life cycle of the customer.
145. 5:05It's literally the only part of the app
146. 5:07that makes you money.
147. 5:10Having a steep discount in your win back
148. 5:12campaign, I think is okay. I don't
149. 5:14really like it when apps immediately go
150. 5:17for the 80% off discount after
151. 5:19onboarding because you could be
152. 5:21cannibalizing revenue by showing them
153. 5:23such a discounted offer before they've
154. 5:25even had a chance to go back and see a
155. 5:27paywall a few more times. But, if
156. 5:29they're subscribed already and they're
157. 5:31thinking about canceling, showing them a
158. 5:33lower rate before you do a bunch of
159. 5:34price testing is the cool way for you to
160. 5:37keep your customers that are thinking of
161. 5:39churning anyway.
162. 5:40>> What's a paywall best practice you
163. 5:42completely disagree with?
164. 5:44>> Is the [music] spin the wheel paywall.
165. 5:46This is a predetermined spin the wheel
166. 5:48lottery file that may land on some kind
167. 5:50of discounted percentage. If you're
168. 5:53optimizing for the longer-term business,
169. 5:55that may not be a good paywall for you.
170. 5:57But, if you're optimizing for a weekly
171. 5:59price, and you want to maximize revenue,
172. 6:02unfortunately, it works. The issue is
173. 6:04that I think more and more users are
174. 6:06getting comfortable closing paywalls
175. 6:08because they know they're going to get
176. 6:09some kind of offer afterwards. For the
177. 6:11apps that want to build a really good
178. 6:14business, just remember that this is a
179. 6:15long-term game.
180. 6:16>> Throughout this conversation, Jonathan
181. 6:18kept coming back [music] to the same
182. 6:20question. What kind of business are you
183. 6:22trying to build? Neither of these are
184. 6:24right or wrong, but [music] they're very
185. 6:27different games. Next up, watch this
186. 6:29paywall video on why some of them work
187. 6:31[music] and some don't.