Saulo Marti (Marketing Director at Olist, previously 8fit, Hibooks, Blinkist) talks with Esther Shatz (VP Product Marketing & Consultancy at Storemaven) about setting up the right KPIs, deciding on channels, what a successful testing structure looks like and the power of paywalls.
Analyze the 3 key components of growth and decide what you want/can do:
- Resources that you have (money, people, etc.)
- Speed of growth you need (sometimes you don’t care too much about efficiency and more about outpacing competition)
- Efficiency
You’re never going to be able to do everything at once: spend less, grow more, grow faster. It’s just not going to happen.
The network you want to test is where your target audience is, so find your ICP (Ideal Customer Profile). Where do they live? What do they use? What are they doing? For most early stage consumer subscription apps, the most ROAS-friendly channel is probably Instagram (or Google depending on the company): what are the chances to crack something nobody else has?
Stick to the basics, learn the fundamentals of your messaging and the creative direction that works for you. Test on one of the main channels first, then you can test/scale on other channels.
Make sure that everyone understands what a hypothesis document is and what it looks like. Whatever test you run has to be able to be described in a hypothesis document: what test you want to run, why you want to run it, how do you run it exactly, the variables tested, how you’ll measure success or failure, what’s the outcome and what are the next steps if it’s positive or negative.
If you don’t know what to do after a positive experiment, don’t even run the test. This kills a lot of tests. Example: testing a creative with a green vs. blue background, then what’s the next step...Everything green? Testing a creative with someone looking left vs. right, then what’s the next step...Everyone looking to the right? It doesn’t scale.
Don’t complain that your product team doesn’t want to test or prioritize something. Make sure you explain the outcome of the test and how it could impact the company’s north star metric. If you did that already, then this needs to be discussed/figured out at a higher level.
Keep in mind Darius Contractor’s psych’d framework to increase funnel conversion: you attribute + and - “psych” points to every step in the journey. For products where the value is not immediately there (Blinkist, Audible, even Netflix), getting people to start a free trial helps them engage and gives them a chance to like the product.
Having hard paywalls as early on in the process as possible helps you take advantage of the highest psych point possible. Asking people to pay is what most frustrates them, and if you do it too late (long onboarding, etc.) you’re missing out and not capitalizing on impulse.
People blame the apps for their lack of commitment: language, fitness, etc. (anything that requires discipline). This is why you need to generate as much commitment as possible early on, and nothing works better than putting in your credit card.
Analyze the 3 key components of growth and decide what you want/can do:
- Resources that you have (money, people, etc.)
- Speed of growth you need (sometimes you don’t care too much about efficiency and more about outpacing competition)
- Efficiency
You’re never going to be able to do everything at once: spend less, grow more, grow faster. It’s just not going to happen.
The network you want to test is where your target audience is, so find your ICP (Ideal Customer Profile). Where do they live? What do they use? What are they doing? For most early stage consumer subscription apps, the most ROAS-friendly channel is probably Instagram (or Google depending on the company): what are the chances to crack something nobody else has?
Stick to the basics, learn the fundamentals of your messaging and the creative direction that works for you. Test on one of the main channels first, then you can test/scale on other channels.
Make sure that everyone understands what a hypothesis document is and what it looks like. Whatever test you run has to be able to be described in a hypothesis document: what test you want to run, why you want to run it, how do you run it exactly, the variables tested, how you’ll measure success or failure, what’s the outcome and what are the next steps if it’s positive or negative.
If you don’t know what to do after a positive experiment, don’t even run the test. This kills a lot of tests. Example: testing a creative with a green vs. blue background, then what’s the next step...Everything green? Testing a creative with someone looking left vs. right, then what’s the next step...Everyone looking to the right? It doesn’t scale.
Don’t complain that your product team doesn’t want to test or prioritize something. Make sure you explain the outcome of the test and how it could impact the company’s north star metric. If you did that already, then this needs to be discussed/figured out at a higher level.
Keep in mind Darius Contractor’s psych’d framework to increase funnel conversion: you attribute + and - “psych” points to every step in the journey. For products where the value is not immediately there (Blinkist, Audible, even Netflix), getting people to start a free trial helps them engage and gives them a chance to like the product.
Having hard paywalls as early on in the process as possible helps you take advantage of the highest psych point possible. Asking people to pay is what most frustrates them, and if you do it too late (long onboarding, etc.) you’re missing out and not capitalizing on impulse.
People blame the apps for their lack of commitment: language, fitness, etc. (anything that requires discipline). This is why you need to generate as much commitment as possible early on, and nothing works better than putting in your credit card.
Analyze the 3 key components of growth and decide what you want/can do:
- Resources that you have (money, people, etc.)
- Speed of growth you need (sometimes you don’t care too much about efficiency and more about outpacing competition)
- Efficiency
You’re never going to be able to do everything at once: spend less, grow more, grow faster. It’s just not going to happen.
The network you want to test is where your target audience is, so find your ICP (Ideal Customer Profile). Where do they live? What do they use? What are they doing? For most early stage consumer subscription apps, the most ROAS-friendly channel is probably Instagram (or Google depending on the company): what are the chances to crack something nobody else has?
Stick to the basics, learn the fundamentals of your messaging and the creative direction that works for you. Test on one of the main channels first, then you can test/scale on other channels.
Make sure that everyone understands what a hypothesis document is and what it looks like. Whatever test you run has to be able to be described in a hypothesis document: what test you want to run, why you want to run it, how do you run it exactly, the variables tested, how you’ll measure success or failure, what’s the outcome and what are the next steps if it’s positive or negative.
If you don’t know what to do after a positive experiment, don’t even run the test. This kills a lot of tests. Example: testing a creative with a green vs. blue background, then what’s the next step...Everything green? Testing a creative with someone looking left vs. right, then what’s the next step...Everyone looking to the right? It doesn’t scale.
Don’t complain that your product team doesn’t want to test or prioritize something. Make sure you explain the outcome of the test and how it could impact the company’s north star metric. If you did that already, then this needs to be discussed/figured out at a higher level.
Keep in mind Darius Contractor’s psych’d framework to increase funnel conversion: you attribute + and - “psych” points to every step in the journey. For products where the value is not immediately there (Blinkist, Audible, even Netflix), getting people to start a free trial helps them engage and gives them a chance to like the product.
Having hard paywalls as early on in the process as possible helps you take advantage of the highest psych point possible. Asking people to pay is what most frustrates them, and if you do it too late (long onboarding, etc.) you’re missing out and not capitalizing on impulse.
People blame the apps for their lack of commitment: language, fitness, etc. (anything that requires discipline). This is why you need to generate as much commitment as possible early on, and nothing works better than putting in your credit card.
Notes for this resource are currently being transferred and will be available soon.
Background in ecommerce, then went into growth. Worked at Blinkist, Hibooks, 8fit.
Set your KPIs not only to your (realistic) SMART goals, but also to the stage at which your company is as well as B2C vs. B2B
[💎@07:05] Analyze the 3 key components of growth and decide what you want/can do:
[💎@07:53] You’re never going to be able to do everything at once: spend less, grow more, grow faster. It’s just not going to happen.
Growth hacks might work if you’re a product-led growth company. Companies that truly grow organically can do it because of how great the product is, or virality and hitting a critical mass (e.g Clubhouse).
If you don’t have an infinite budget, how do you test new channels (programmatic, Quora, etc.)?
[💎@10:56] The network you want to test is where your target audience is, so find your ICP (Ideal Customer Profile). Where do they live? What do they use? What are they doing? For most early stage consumer subscription apps, the most ROAS-friendly channel is probably Instagram (or Google depending on the company): what are the chances to crack something nobody else has?
[💎@11:37] Stick to the basics, learn the fundamentals of your messaging and the creative direction that works for you. Test on one of the main channels first, then you can test/scale on other channels.
Saulo uses a framework and pitches it to all companies he advises. Maybe not the best, but it is the simplest.
[💎@13:16] Make sure that everyone understands what a hypothesis document is and what it looks like. Whatever test you run has to be able to be described in a hypothesis document: what test you want to run, why you want to run it, how do you run it exactly, the variables tested, how you’ll measure success or failure, what’s the outcome and what are the next steps if it’s positive or negative.
[💎@14:29] If you don’t know what to do after a positive experiment, don’t even run the test. This kills a lot of tests. Example: testing a creative with a green vs. blue background, then what’s the next step...Everything green? Testing a creative with someone looking left vs. right, then what’s the next step...Everyone looking to the right? It doesn’t scale.
Not everyone should be data-driven, but everyone should be data-informed. There are some things that you do not need to test: for example you know that if you remove the paywall and don’t charge for a subscription, you’re going to get more conversions. But the next step would then be “how do you hit your revenue/ROAS goals?” or “how do you upgrade those free users down the line?”. You don’t need to run that initial test of charging/not charging and lose time.
Saulo was fortunate that he would usually have good control on what he could do. But sometimes you do have some conflict.
[💎@18:45] Don’t complain that your product team doesn’t want to test or prioritize something. Make sure you explain the outcome of the test and how it could impact the company’s north star metric. If you did that already, then this needs to be discussed/figured out at a higher level.
“A lot of growth people fail because they fail to explain why it’s important to run a test”
If you don’t have a growth mindset across all departments, you shouldn’t have a growth department. It’s your job to instill that in the other teams. A lot of the time the Chief Growth Officer’s job is to make people understand why it’s important to seek growth in alternate routes through testing, being data-informed, etc.
Example:
With a paywall you’re forcing a user to make a commitment to actually use the product through a free trial and engage with it. This means more people stick around.
[💎@28:40] Keep in mind Darius Contractor’s psych’d framework to increase funnel conversion: you attribute + and - “psych” points to every step in the journey. For products where the value is not immediately there (Blinkist, Audible, even Netflix), getting people to start a free trial helps them engage and gives them a chance to like the product.
[💎@30:13] Having hard paywalls as early on in the process as possible helps you take advantage of the highest psych point possible. Asking people to pay is what most frustrates them, and if you do it too late (long onboarding, etc.) you’re missing out and not capitalizing on impulse.
“If you didn’t have a paywall early on in fitness apps, no one would pay for it”
[💎@31:38] People blame the apps for their lack of commitment: language, fitness, etc. (anything that requires discipline). This is why you need to generate as much commitment as possible early on, and nothing works better than putting in your credit card.
You still need to work on the product of course, but putting the paywall early capitalizes on the impulse.
Exceptions are apps like Spotify: if you’re a heavy user, there is pretty much nothing other apps can throw at you to make you leave. All your playlists are in there, etc.
Talks about Netflix, Disney, etc. and growth/retention play based on the shows they launch or acquire.
You need to think about features that are either retention-driven or acquisition-driven. Example: recipes in 8fit are retention-driven, and that’s not what people sign up for (they want to lose wait). But they help keep some users around.
One tip to someone starting in Growth: learn to ask questions.
Favorite resource: Reforge but it’s expensive. He loves Reid Hoffman’s podcast: Masters of Scale.
Background in ecommerce, then went into growth. Worked at Blinkist, Hibooks, 8fit.
Set your KPIs not only to your (realistic) SMART goals, but also to the stage at which your company is as well as B2C vs. B2B
[💎@07:05] Analyze the 3 key components of growth and decide what you want/can do:
[💎@07:53] You’re never going to be able to do everything at once: spend less, grow more, grow faster. It’s just not going to happen.
Growth hacks might work if you’re a product-led growth company. Companies that truly grow organically can do it because of how great the product is, or virality and hitting a critical mass (e.g Clubhouse).
If you don’t have an infinite budget, how do you test new channels (programmatic, Quora, etc.)?
[💎@10:56] The network you want to test is where your target audience is, so find your ICP (Ideal Customer Profile). Where do they live? What do they use? What are they doing? For most early stage consumer subscription apps, the most ROAS-friendly channel is probably Instagram (or Google depending on the company): what are the chances to crack something nobody else has?
[💎@11:37] Stick to the basics, learn the fundamentals of your messaging and the creative direction that works for you. Test on one of the main channels first, then you can test/scale on other channels.
Saulo uses a framework and pitches it to all companies he advises. Maybe not the best, but it is the simplest.
[💎@13:16] Make sure that everyone understands what a hypothesis document is and what it looks like. Whatever test you run has to be able to be described in a hypothesis document: what test you want to run, why you want to run it, how do you run it exactly, the variables tested, how you’ll measure success or failure, what’s the outcome and what are the next steps if it’s positive or negative.
[💎@14:29] If you don’t know what to do after a positive experiment, don’t even run the test. This kills a lot of tests. Example: testing a creative with a green vs. blue background, then what’s the next step...Everything green? Testing a creative with someone looking left vs. right, then what’s the next step...Everyone looking to the right? It doesn’t scale.
Not everyone should be data-driven, but everyone should be data-informed. There are some things that you do not need to test: for example you know that if you remove the paywall and don’t charge for a subscription, you’re going to get more conversions. But the next step would then be “how do you hit your revenue/ROAS goals?” or “how do you upgrade those free users down the line?”. You don’t need to run that initial test of charging/not charging and lose time.
Saulo was fortunate that he would usually have good control on what he could do. But sometimes you do have some conflict.
[💎@18:45] Don’t complain that your product team doesn’t want to test or prioritize something. Make sure you explain the outcome of the test and how it could impact the company’s north star metric. If you did that already, then this needs to be discussed/figured out at a higher level.
“A lot of growth people fail because they fail to explain why it’s important to run a test”
If you don’t have a growth mindset across all departments, you shouldn’t have a growth department. It’s your job to instill that in the other teams. A lot of the time the Chief Growth Officer’s job is to make people understand why it’s important to seek growth in alternate routes through testing, being data-informed, etc.
Example:
With a paywall you’re forcing a user to make a commitment to actually use the product through a free trial and engage with it. This means more people stick around.
[💎@28:40] Keep in mind Darius Contractor’s psych’d framework to increase funnel conversion: you attribute + and - “psych” points to every step in the journey. For products where the value is not immediately there (Blinkist, Audible, even Netflix), getting people to start a free trial helps them engage and gives them a chance to like the product.
[💎@30:13] Having hard paywalls as early on in the process as possible helps you take advantage of the highest psych point possible. Asking people to pay is what most frustrates them, and if you do it too late (long onboarding, etc.) you’re missing out and not capitalizing on impulse.
“If you didn’t have a paywall early on in fitness apps, no one would pay for it”
[💎@31:38] People blame the apps for their lack of commitment: language, fitness, etc. (anything that requires discipline). This is why you need to generate as much commitment as possible early on, and nothing works better than putting in your credit card.
You still need to work on the product of course, but putting the paywall early capitalizes on the impulse.
Exceptions are apps like Spotify: if you’re a heavy user, there is pretty much nothing other apps can throw at you to make you leave. All your playlists are in there, etc.
Talks about Netflix, Disney, etc. and growth/retention play based on the shows they launch or acquire.
You need to think about features that are either retention-driven or acquisition-driven. Example: recipes in 8fit are retention-driven, and that’s not what people sign up for (they want to lose wait). But they help keep some users around.
One tip to someone starting in Growth: learn to ask questions.
Favorite resource: Reforge but it’s expensive. He loves Reid Hoffman’s podcast: Masters of Scale.
Background in ecommerce, then went into growth. Worked at Blinkist, Hibooks, 8fit.
Set your KPIs not only to your (realistic) SMART goals, but also to the stage at which your company is as well as B2C vs. B2B
[💎@07:05] Analyze the 3 key components of growth and decide what you want/can do:
[💎@07:53] You’re never going to be able to do everything at once: spend less, grow more, grow faster. It’s just not going to happen.
Growth hacks might work if you’re a product-led growth company. Companies that truly grow organically can do it because of how great the product is, or virality and hitting a critical mass (e.g Clubhouse).
If you don’t have an infinite budget, how do you test new channels (programmatic, Quora, etc.)?
[💎@10:56] The network you want to test is where your target audience is, so find your ICP (Ideal Customer Profile). Where do they live? What do they use? What are they doing? For most early stage consumer subscription apps, the most ROAS-friendly channel is probably Instagram (or Google depending on the company): what are the chances to crack something nobody else has?
[💎@11:37] Stick to the basics, learn the fundamentals of your messaging and the creative direction that works for you. Test on one of the main channels first, then you can test/scale on other channels.
Saulo uses a framework and pitches it to all companies he advises. Maybe not the best, but it is the simplest.
[💎@13:16] Make sure that everyone understands what a hypothesis document is and what it looks like. Whatever test you run has to be able to be described in a hypothesis document: what test you want to run, why you want to run it, how do you run it exactly, the variables tested, how you’ll measure success or failure, what’s the outcome and what are the next steps if it’s positive or negative.
[💎@14:29] If you don’t know what to do after a positive experiment, don’t even run the test. This kills a lot of tests. Example: testing a creative with a green vs. blue background, then what’s the next step...Everything green? Testing a creative with someone looking left vs. right, then what’s the next step...Everyone looking to the right? It doesn’t scale.
Not everyone should be data-driven, but everyone should be data-informed. There are some things that you do not need to test: for example you know that if you remove the paywall and don’t charge for a subscription, you’re going to get more conversions. But the next step would then be “how do you hit your revenue/ROAS goals?” or “how do you upgrade those free users down the line?”. You don’t need to run that initial test of charging/not charging and lose time.
Saulo was fortunate that he would usually have good control on what he could do. But sometimes you do have some conflict.
[💎@18:45] Don’t complain that your product team doesn’t want to test or prioritize something. Make sure you explain the outcome of the test and how it could impact the company’s north star metric. If you did that already, then this needs to be discussed/figured out at a higher level.
“A lot of growth people fail because they fail to explain why it’s important to run a test”
If you don’t have a growth mindset across all departments, you shouldn’t have a growth department. It’s your job to instill that in the other teams. A lot of the time the Chief Growth Officer’s job is to make people understand why it’s important to seek growth in alternate routes through testing, being data-informed, etc.
Example:
With a paywall you’re forcing a user to make a commitment to actually use the product through a free trial and engage with it. This means more people stick around.
[💎@28:40] Keep in mind Darius Contractor’s psych’d framework to increase funnel conversion: you attribute + and - “psych” points to every step in the journey. For products where the value is not immediately there (Blinkist, Audible, even Netflix), getting people to start a free trial helps them engage and gives them a chance to like the product.
[💎@30:13] Having hard paywalls as early on in the process as possible helps you take advantage of the highest psych point possible. Asking people to pay is what most frustrates them, and if you do it too late (long onboarding, etc.) you’re missing out and not capitalizing on impulse.
“If you didn’t have a paywall early on in fitness apps, no one would pay for it”
[💎@31:38] People blame the apps for their lack of commitment: language, fitness, etc. (anything that requires discipline). This is why you need to generate as much commitment as possible early on, and nothing works better than putting in your credit card.
You still need to work on the product of course, but putting the paywall early capitalizes on the impulse.
Exceptions are apps like Spotify: if you’re a heavy user, there is pretty much nothing other apps can throw at you to make you leave. All your playlists are in there, etc.
Talks about Netflix, Disney, etc. and growth/retention play based on the shows they launch or acquire.
You need to think about features that are either retention-driven or acquisition-driven. Example: recipes in 8fit are retention-driven, and that’s not what people sign up for (they want to lose wait). But they help keep some users around.
One tip to someone starting in Growth: learn to ask questions.
Favorite resource: Reforge but it’s expensive. He loves Reid Hoffman’s podcast: Masters of Scale.
