Peggy Anne Salz (Lead Analyst & Founder at Mobile Groove) interviews Simon Lejeune (Head of User Acquisition at Hopper) shares the tools and channels that have allowed Hopper to stay top of mind with users as well as how marketing teams can combine acquisition and retention to break down the silos and super-charge ad creatives.
When you have a long sales cycle like it's the case for Hopper, it's important to have a couple of early metrics to tell if your users are qualified so you can shift your budget the right way. At Hopper it is the "watch" rate.
Hopper uses a fixed payback period of 12 months. They look back on cohorts and measure the average revenue per new user install during these 12 months in order to define a CPI target/limit. They then spend as much as possible within that limit.
Some things can work well from an initial engagement perspective but backfire on internet or social media. Example at Hopper: when they tried "fake Hawaii flight notifications" (backslash on Reddit).
Additional smaller channels can become a distraction: tracking, talking to reps, etc. Hopper now focuses on the biggest channels (FB/Instagram, Google UAC) and their efforts and energy there.
Hopper still leverages Pinterest/Twitter/Snapchat that all together bring an interesting amount of volume. It allows to diversify/spread when seasonality happens (to avoid peaks) and gives some leverage with the different channels (even the big guys).
For the automation of all their "flight deals" ads the (really technical) marketing team created "Hopper Ads Engine" where they can "fetch the top 500 flight deals for existing customers and generate creatives, campaigns and ad sets for X Y Z channel". The data and these creatives are then sent to the networks directly or to Smartly.io which generates the creatives (even dynamic video ads).
With Smartly.io Hopper can generate thousands of videos with just a few clicks that have a very high relevance because of the targeting and deals.
Try to understand where Facebook and Google are going and do work in the opposite direction: try to build on what they are lacking. Because FB/Google take over the optimization part of the business, Hopper focused on creatives instead of an optimization engine.
Hopper tries to build tools that reallocate budget between channels, not necessarily within a channel. Because FB/Google are never going to do that, it is another thing that is valuable for the long term.
When you have a long sales cycle like it's the case for Hopper, it's important to have a couple of early metrics to tell if your users are qualified so you can shift your budget the right way. At Hopper it is the "watch" rate.
Hopper uses a fixed payback period of 12 months. They look back on cohorts and measure the average revenue per new user install during these 12 months in order to define a CPI target/limit. They then spend as much as possible within that limit.
Some things can work well from an initial engagement perspective but backfire on internet or social media. Example at Hopper: when they tried "fake Hawaii flight notifications" (backslash on Reddit).
Additional smaller channels can become a distraction: tracking, talking to reps, etc. Hopper now focuses on the biggest channels (FB/Instagram, Google UAC) and their efforts and energy there.
Hopper still leverages Pinterest/Twitter/Snapchat that all together bring an interesting amount of volume. It allows to diversify/spread when seasonality happens (to avoid peaks) and gives some leverage with the different channels (even the big guys).
For the automation of all their "flight deals" ads the (really technical) marketing team created "Hopper Ads Engine" where they can "fetch the top 500 flight deals for existing customers and generate creatives, campaigns and ad sets for X Y Z channel". The data and these creatives are then sent to the networks directly or to Smartly.io which generates the creatives (even dynamic video ads).
With Smartly.io Hopper can generate thousands of videos with just a few clicks that have a very high relevance because of the targeting and deals.
Try to understand where Facebook and Google are going and do work in the opposite direction: try to build on what they are lacking. Because FB/Google take over the optimization part of the business, Hopper focused on creatives instead of an optimization engine.
Hopper tries to build tools that reallocate budget between channels, not necessarily within a channel. Because FB/Google are never going to do that, it is another thing that is valuable for the long term.
When you have a long sales cycle like it's the case for Hopper, it's important to have a couple of early metrics to tell if your users are qualified so you can shift your budget the right way. At Hopper it is the "watch" rate.
Hopper uses a fixed payback period of 12 months. They look back on cohorts and measure the average revenue per new user install during these 12 months in order to define a CPI target/limit. They then spend as much as possible within that limit.
Some things can work well from an initial engagement perspective but backfire on internet or social media. Example at Hopper: when they tried "fake Hawaii flight notifications" (backslash on Reddit).
Additional smaller channels can become a distraction: tracking, talking to reps, etc. Hopper now focuses on the biggest channels (FB/Instagram, Google UAC) and their efforts and energy there.
Hopper still leverages Pinterest/Twitter/Snapchat that all together bring an interesting amount of volume. It allows to diversify/spread when seasonality happens (to avoid peaks) and gives some leverage with the different channels (even the big guys).
For the automation of all their "flight deals" ads the (really technical) marketing team created "Hopper Ads Engine" where they can "fetch the top 500 flight deals for existing customers and generate creatives, campaigns and ad sets for X Y Z channel". The data and these creatives are then sent to the networks directly or to Smartly.io which generates the creatives (even dynamic video ads).
With Smartly.io Hopper can generate thousands of videos with just a few clicks that have a very high relevance because of the targeting and deals.
Try to understand where Facebook and Google are going and do work in the opposite direction: try to build on what they are lacking. Because FB/Google take over the optimization part of the business, Hopper focused on creatives instead of an optimization engine.
Hopper tries to build tools that reallocate budget between channels, not necessarily within a channel. Because FB/Google are never going to do that, it is another thing that is valuable for the long term.
Notes for this resource are currently being transferred and will be available soon.
Retention has already been really important at Hopper because it is built into the product.
Hopper predicts the future price of flights: you look for flights/hotels, and it tells you if the price is going to go down. If that's the case, you subscribe to push notifications to be alerted on when to book. This 1:1 communication channel allows to have great retention.
Most travel companies really focus on conversion: they acquire low funnel and expensive traffic from Google and have to convert them right away.
Hopper is mobile only so they have to pay more to acquire users but don't have to convert them right away. They need to focus on retention.
Some people look at D7, D14, etc. Hopper does not optimize for session or retention at DX: most people download the app, book a flight/hotel and won't open the app until they need a new booking.
Hopper actually optimizes to get people to a sale with the minimum amount of time spent in the app.
Realized older cohorts (3-4 years ago) generated more revenue in year 3 than year 1. Half of these Hopper users stay with them "forever". This is hard to model vs. SaaS or mobile in general.
[💎@06:50] When you have a long sales cycle like it's the case for Hopper, it's important to have a couple of early metrics to tell if your users are qualified and shift your budget the right way.
At Hopper they use the "watch rate" (someone watching the price of a flight). 70-75% of new users watch a flight in their first session (depending on the channel).
[💎@07:30] Hopper uses a fixed payback period of 12 months. They look back on cohorts and measure the average revenue per new user install during these 12 months in order to define a CPI target/limit. They then spend as much as possible within that limit.
(They use to spend a fix marketing budget for month and try to get the best watch rate and ROI).
North start metric for campaigns: optimize for early engagement that predicts future behavior (→ watch rate).
As a team they try to grow sales as fast within a 12 months net payback period.
Hopper is customer-obsessed and constantly communicates with users via focus groups, typeforms, etc. The #1 feedback is that they do not send enough notifications (people get worried).
Between the moment a user watches a flight and books it there are now about 40 notifications.
They have now created a retention group and started to send notifications that are not related to watching a flight. Example: after someone comes back from a trip. Also added email communications in order to stay top of mind.
Challenge: mixing price updates and marketing notifications that end up being annoying then users might deactivate notifications and lose the experience. So they are now thinking about giving more control to users by letting them decide (separated notification preferences).
[💎@13:10] Some things can work well from an initial engagement perspective but backfire on internet or social media. Example at Hopper: when they tried "fake Hawaii flight notifications" (backslash on Reddit).
Hopper experimented by sending on notifications to users that were not watching any flights (example: "Your flight to Hawaii went down - false alarm, but this is the kind of notifications you would get"). This seem to work really well from an engagement perspective. But people complained on Reddit, which was their signal that you can't go to far.
Hopper is still flying under the radar so this kind of Reddit backlash die pretty quickly. But they took the feedback very seriously and shut down the feature.
In this case, one engineer decided to reply and explain why they did it.
Hopper really focuses on paid social media.
That said a lot of people turn to social media for travel advice so Hopper is organically mentioned as a tool or travel hack but it is hard to increase/accelerate these conversations.
8 out of 10 user has recommended Hopper. If each of them convinced 1 or 2 users to download the app, they would have viral organic growth but that is actually not the case. But somewhere in that conversation something gets lost so there might be a way to help that conversation: gift, financial incentive, etc.
You need to talk to your users to better understand organic traffic.
When started, Simon wanted the biggest amount of channels and opened Quora/Reddit and other smaller channels. But it's not set it and forget it.
[💎@21:38] Additional smaller channels can become a distraction: tracking, talking to reps, etc. Hopper now focuses on the biggest channels (FB/Instagram, Google UAC) and their efforts and energy there.
[💎@22:00] Hopper still leverages Pinterest/Twitter/Snapchat that all together bring an interesting amount of volume. It allows to diversify/spread when seasonality happens (to avoid peaks) and gives some leverage with the different channels (even the big guys).
Advertising Hopper in a generic fashion (talking about features and the app) doesn't work as well as "flight deals", which gives them the best engagement.
Flight deals require automation because prices change all the time, are relevant to users only if it's from your own airport (requires geo-targeting). And you need 100s of creatives on all the different platforms, which sometimes have different creative specs/ratios.
[💎@24:45] For the automation of all their "flight deals" ads the (really technical) marketing team created "Hopper Ads Engine" where they can say "fetch the top 500 flight deals for existing customers and generate creatives, campaigns and ad sets for X Y Z channel". The data and these creatives are then sent to the networks directly or to Smartly.io which generates the creatives (even dynamic video ads).
[💎@26:42] With Smartly.io Hopper can generate thousands of videos with just a few clicks that have a very high relevance because of the targeting and deals.
Try to understand where Facebook and Google are going and try to do work in the opposite direction.
[💎@27:38] Try to understand where Facebook and Google are going and do work in the opposite direction. Try to build on what they are lacking. Because FB/Google take over the optimization part of the business, Hopper focused on creatives instead of an optimization engine.
[💎@28:44] Hopper tries to build tools that reallocate budget between channels, not necessarily within a channel. Because FB/Google are never going to do that, it is another thing that is valuable for the long term.
Retention has already been really important at Hopper because it is built into the product.
Hopper predicts the future price of flights: you look for flights/hotels, and it tells you if the price is going to go down. If that's the case, you subscribe to push notifications to be alerted on when to book. This 1:1 communication channel allows to have great retention.
Most travel companies really focus on conversion: they acquire low funnel and expensive traffic from Google and have to convert them right away.
Hopper is mobile only so they have to pay more to acquire users but don't have to convert them right away. They need to focus on retention.
Some people look at D7, D14, etc. Hopper does not optimize for session or retention at DX: most people download the app, book a flight/hotel and won't open the app until they need a new booking.
Hopper actually optimizes to get people to a sale with the minimum amount of time spent in the app.
Realized older cohorts (3-4 years ago) generated more revenue in year 3 than year 1. Half of these Hopper users stay with them "forever". This is hard to model vs. SaaS or mobile in general.
[💎@06:50] When you have a long sales cycle like it's the case for Hopper, it's important to have a couple of early metrics to tell if your users are qualified and shift your budget the right way.
At Hopper they use the "watch rate" (someone watching the price of a flight). 70-75% of new users watch a flight in their first session (depending on the channel).
[💎@07:30] Hopper uses a fixed payback period of 12 months. They look back on cohorts and measure the average revenue per new user install during these 12 months in order to define a CPI target/limit. They then spend as much as possible within that limit.
(They use to spend a fix marketing budget for month and try to get the best watch rate and ROI).
North start metric for campaigns: optimize for early engagement that predicts future behavior (→ watch rate).
As a team they try to grow sales as fast within a 12 months net payback period.
Hopper is customer-obsessed and constantly communicates with users via focus groups, typeforms, etc. The #1 feedback is that they do not send enough notifications (people get worried).
Between the moment a user watches a flight and books it there are now about 40 notifications.
They have now created a retention group and started to send notifications that are not related to watching a flight. Example: after someone comes back from a trip. Also added email communications in order to stay top of mind.
Challenge: mixing price updates and marketing notifications that end up being annoying then users might deactivate notifications and lose the experience. So they are now thinking about giving more control to users by letting them decide (separated notification preferences).
[💎@13:10] Some things can work well from an initial engagement perspective but backfire on internet or social media. Example at Hopper: when they tried "fake Hawaii flight notifications" (backslash on Reddit).
Hopper experimented by sending on notifications to users that were not watching any flights (example: "Your flight to Hawaii went down - false alarm, but this is the kind of notifications you would get"). This seem to work really well from an engagement perspective. But people complained on Reddit, which was their signal that you can't go to far.
Hopper is still flying under the radar so this kind of Reddit backlash die pretty quickly. But they took the feedback very seriously and shut down the feature.
In this case, one engineer decided to reply and explain why they did it.
Hopper really focuses on paid social media.
That said a lot of people turn to social media for travel advice so Hopper is organically mentioned as a tool or travel hack but it is hard to increase/accelerate these conversations.
8 out of 10 user has recommended Hopper. If each of them convinced 1 or 2 users to download the app, they would have viral organic growth but that is actually not the case. But somewhere in that conversation something gets lost so there might be a way to help that conversation: gift, financial incentive, etc.
You need to talk to your users to better understand organic traffic.
When started, Simon wanted the biggest amount of channels and opened Quora/Reddit and other smaller channels. But it's not set it and forget it.
[💎@21:38] Additional smaller channels can become a distraction: tracking, talking to reps, etc. Hopper now focuses on the biggest channels (FB/Instagram, Google UAC) and their efforts and energy there.
[💎@22:00] Hopper still leverages Pinterest/Twitter/Snapchat that all together bring an interesting amount of volume. It allows to diversify/spread when seasonality happens (to avoid peaks) and gives some leverage with the different channels (even the big guys).
Advertising Hopper in a generic fashion (talking about features and the app) doesn't work as well as "flight deals", which gives them the best engagement.
Flight deals require automation because prices change all the time, are relevant to users only if it's from your own airport (requires geo-targeting). And you need 100s of creatives on all the different platforms, which sometimes have different creative specs/ratios.
[💎@24:45] For the automation of all their "flight deals" ads the (really technical) marketing team created "Hopper Ads Engine" where they can say "fetch the top 500 flight deals for existing customers and generate creatives, campaigns and ad sets for X Y Z channel". The data and these creatives are then sent to the networks directly or to Smartly.io which generates the creatives (even dynamic video ads).
[💎@26:42] With Smartly.io Hopper can generate thousands of videos with just a few clicks that have a very high relevance because of the targeting and deals.
Try to understand where Facebook and Google are going and try to do work in the opposite direction.
[💎@27:38] Try to understand where Facebook and Google are going and do work in the opposite direction. Try to build on what they are lacking. Because FB/Google take over the optimization part of the business, Hopper focused on creatives instead of an optimization engine.
[💎@28:44] Hopper tries to build tools that reallocate budget between channels, not necessarily within a channel. Because FB/Google are never going to do that, it is another thing that is valuable for the long term.
Retention has already been really important at Hopper because it is built into the product.
Hopper predicts the future price of flights: you look for flights/hotels, and it tells you if the price is going to go down. If that's the case, you subscribe to push notifications to be alerted on when to book. This 1:1 communication channel allows to have great retention.
Most travel companies really focus on conversion: they acquire low funnel and expensive traffic from Google and have to convert them right away.
Hopper is mobile only so they have to pay more to acquire users but don't have to convert them right away. They need to focus on retention.
Some people look at D7, D14, etc. Hopper does not optimize for session or retention at DX: most people download the app, book a flight/hotel and won't open the app until they need a new booking.
Hopper actually optimizes to get people to a sale with the minimum amount of time spent in the app.
Realized older cohorts (3-4 years ago) generated more revenue in year 3 than year 1. Half of these Hopper users stay with them "forever". This is hard to model vs. SaaS or mobile in general.
[💎@06:50] When you have a long sales cycle like it's the case for Hopper, it's important to have a couple of early metrics to tell if your users are qualified and shift your budget the right way.
At Hopper they use the "watch rate" (someone watching the price of a flight). 70-75% of new users watch a flight in their first session (depending on the channel).
[💎@07:30] Hopper uses a fixed payback period of 12 months. They look back on cohorts and measure the average revenue per new user install during these 12 months in order to define a CPI target/limit. They then spend as much as possible within that limit.
(They use to spend a fix marketing budget for month and try to get the best watch rate and ROI).
North start metric for campaigns: optimize for early engagement that predicts future behavior (→ watch rate).
As a team they try to grow sales as fast within a 12 months net payback period.
Hopper is customer-obsessed and constantly communicates with users via focus groups, typeforms, etc. The #1 feedback is that they do not send enough notifications (people get worried).
Between the moment a user watches a flight and books it there are now about 40 notifications.
They have now created a retention group and started to send notifications that are not related to watching a flight. Example: after someone comes back from a trip. Also added email communications in order to stay top of mind.
Challenge: mixing price updates and marketing notifications that end up being annoying then users might deactivate notifications and lose the experience. So they are now thinking about giving more control to users by letting them decide (separated notification preferences).
[💎@13:10] Some things can work well from an initial engagement perspective but backfire on internet or social media. Example at Hopper: when they tried "fake Hawaii flight notifications" (backslash on Reddit).
Hopper experimented by sending on notifications to users that were not watching any flights (example: "Your flight to Hawaii went down - false alarm, but this is the kind of notifications you would get"). This seem to work really well from an engagement perspective. But people complained on Reddit, which was their signal that you can't go to far.
Hopper is still flying under the radar so this kind of Reddit backlash die pretty quickly. But they took the feedback very seriously and shut down the feature.
In this case, one engineer decided to reply and explain why they did it.
Hopper really focuses on paid social media.
That said a lot of people turn to social media for travel advice so Hopper is organically mentioned as a tool or travel hack but it is hard to increase/accelerate these conversations.
8 out of 10 user has recommended Hopper. If each of them convinced 1 or 2 users to download the app, they would have viral organic growth but that is actually not the case. But somewhere in that conversation something gets lost so there might be a way to help that conversation: gift, financial incentive, etc.
You need to talk to your users to better understand organic traffic.
When started, Simon wanted the biggest amount of channels and opened Quora/Reddit and other smaller channels. But it's not set it and forget it.
[💎@21:38] Additional smaller channels can become a distraction: tracking, talking to reps, etc. Hopper now focuses on the biggest channels (FB/Instagram, Google UAC) and their efforts and energy there.
[💎@22:00] Hopper still leverages Pinterest/Twitter/Snapchat that all together bring an interesting amount of volume. It allows to diversify/spread when seasonality happens (to avoid peaks) and gives some leverage with the different channels (even the big guys).
Advertising Hopper in a generic fashion (talking about features and the app) doesn't work as well as "flight deals", which gives them the best engagement.
Flight deals require automation because prices change all the time, are relevant to users only if it's from your own airport (requires geo-targeting). And you need 100s of creatives on all the different platforms, which sometimes have different creative specs/ratios.
[💎@24:45] For the automation of all their "flight deals" ads the (really technical) marketing team created "Hopper Ads Engine" where they can say "fetch the top 500 flight deals for existing customers and generate creatives, campaigns and ad sets for X Y Z channel". The data and these creatives are then sent to the networks directly or to Smartly.io which generates the creatives (even dynamic video ads).
[💎@26:42] With Smartly.io Hopper can generate thousands of videos with just a few clicks that have a very high relevance because of the targeting and deals.
Try to understand where Facebook and Google are going and try to do work in the opposite direction.
[💎@27:38] Try to understand where Facebook and Google are going and do work in the opposite direction. Try to build on what they are lacking. Because FB/Google take over the optimization part of the business, Hopper focused on creatives instead of an optimization engine.
[💎@28:44] Hopper tries to build tools that reallocate budget between channels, not necessarily within a channel. Because FB/Google are never going to do that, it is another thing that is valuable for the long term.
