Life360

Life360 declined a test that doubled subscribers because it hurt free users. Cutting free location history from 2 days to 6 or 12 hours doubled new subscribers from that hook, but Life360 did not ship it because it made the feature less useful and, in some cases, made people less safe.

The company accepted the short-term revenue cost. "We paid for it on short-term bottom line, but we think in the long term it's going to be, you know, the right thing to do for our business." No comparison with the doubled-subscriber version or figure for the forgone revenue is given.

Sub Club▶ 25:36

A large free base is the moat, since the app is useless solo. Parents telling other parents is the main way people find Life360, and a large engaged free base drives invites, word of mouth and density for the Pet finder network and Tile devices. About 3% of its nearly 100 million monthly users subscribe, and each subscription covers about four people.

The free base also supports the ad business, which gives the company an extra reason to keep it large. "Having a large free engage free user base around the world is actually one of the big modes and one of the big competitive advantages for us."

Sub Club▶ 23:32

When free is good enough, add paid value instead of cutting free. The top reason people give for not subscribing is that the free tier is good enough. Life360's answer is to add more value and diversify the subscription offer rather than degrade the free experience.

"Our philosophy is that literally we want to do something about it, but that something is not making the free tier worse." Diversifying means new reasons to pay, not features taken away from free users.

Sub Club▶ 23:32

Gate features with real marginal cost; make software features great for free. Roadside assistance, where Life360 pays for each tow, is premium. For software features, the team builds the best free version first, then a new experience or tier on top worth paying for, not an artificial cap.

Free-experience and paid-experience teams are expected to think about both sides. "If something is very expensive to provide, it will become a premium feature."

Sub Club▶ 28:20

Showing Platinum to likely buyers doubled Platinum uptake without losing Gold. A model using about 900 data points found new users with a higher propensity to buy Platinum and showed them Platinum on the paywall instead of the default Gold. The share of new users subscribing to Platinum doubled, with no Gold subscribers lost.

Most users pick the mid tier because a paywall cannot explain every plan, so the standard offer was missing top-tier buyers. No baseline rates or revenue figures are given.

Sub Club▶ 47:48

One ML layer targets micro audiences instead of a feature for each. Built in the last year at the growth and revenue layer, it orchestrates paywalls, push notifications, in-app messaging, promotions and feature education for small segments.

It could grow into deeper personalization, even of features inside the app. It is also how Life360 shows the right new feature to the right person without splitting into separate apps.

Sub Club▶ 14:15

Life360 turned a growth team into a company-wide experimentation machine. Growth covers subscriptions, ads, Tile and pet GPS device sales, and user growth. Everyone aligns behind top-line revenue and user metrics, and each team owns sub-metrics it can move, with AI tooling lowering the barrier to run experiments.

Finance automated routine work, then built tools to mine data on pricing and tiering; HR applies a growth angle to hiring, performance evaluation and org design. The stated goal is a billion-dollar revenue company in the next couple of years, from revenue near $500 million today.

Sub Club▶ 4:01

Outside product, growth leads self-select as builders. Only product development has an explicit growth mandate. Elsewhere Life360 waits for people with a builder mindset to emerge, then supports them with training and a fellowship where someone swaps their job for 3 months to make their team faster with technology.

"We looking for them to emerge and self-select and and kind of start getting things done."

Sub Club▶ 6:48

Only inconclusive tests count as failures, and Life360 tracks them. Wins get scaled and losses prove or disprove a hypothesis, but an inconclusive test means wasted time, a bad setup or a weak hypothesis. Life360 measures the number of inconclusive tests alongside revenue and works to reduce it.

Teams are still judged on revenue, and each loss is expected to come with learnings and time to act on them.

Sub Club▶ 9:57

Velocity, win rate and average win compound into growth. Life360 manages experiments as a portfolio with three levers and uses AI to raise velocity while keeping win rate and average win roughly stable.

Some tests will lose as long as enough big wins land and the losses yield learning. "We don't think of experimentation as a series of experiments, but we think as a portfolio and a learning portfolio."

Sub Club▶ 9:13

Segment losing tests to find who they worked for. Life360 predicts impact before a test, then segments results afterward. A loss overall can win for a niche, such as suburban users with a dog who have been on the platform a month.

The stack is standard (Databricks, Statsig, some homegrown tools); the edge comes from people who know the product and talk to users reading the data, increasingly helped by automated flagging. New experiments come out of the segments.

Sub Club▶ 10:36

Circles of three to four people convert and retain best. Life360 sees significantly higher conversion and retention in circles of three to four than in smaller or larger ones.

Larger circles tend to be friend groups or extended family, less motivated to pay for safety. Smaller ones are couples or families whose kids left for college, a major churn reason.

Sub Club▶ 34:03

Referral buttons failed because parents, the deciders, are not viral. About 40% of users say they found Life360 through someone else, so the team tested in-app share buttons and gifted Silver subscriptions. They failed across the board.

Kids are viral but are not the ones deciding to use the app. "Parents are not viral."

Sub Club▶ 51:50

Users who join a circle and invite others in week one stay. Creating or joining a circle and inviting people correlates strongly with being active on day seven and a year later, so onboarding and mid-boarding steps push users to the app's value in the first week.

For example, a user who has not used location history by day two gets a prompt. "If you're able to make you understand the value of the app in the first week, we have a long-term users."

Sub Club▶ 18:13

Personalize feature prompts with the user's own circle. Instead of a generic tap-here prompt, Life360 asks something like what John was up to yesterday when John is in the user's circle.

The aim is prompts that feel like coaching tied to what the user is doing, as in the Ladder fitness app, where coaches explain why to set weights or log meals.

Sub Club▶ 16:42

A playful animated dog sold the pet profile where banners failed. On app open, an animated dog asked to be petted, and tapping it opened the pet profile, which lets users alert the neighborhood if a dog runs away. It drove a lot of sign-ups and was called a really huge success.

Users tap through standard banners, emails and push notifications, so novel surfaces cut through the noise.

Sub Club▶ 19:30

Life360 builds for pets because 80% of its families have one. About 80% of families with kids on the platform also have a pet, and many couples have pets too. Pet features give them a reason to subscribe and extend usage after kids leave home.

Pet owners may care less about driving safety and other family features, so pet value fills that gap.

Sub Club▶ 14:15

Tile and pet GPS devices bring users in and lift retention. Life360 treats devices as top of funnel and bundles free Tile offers into subscriptions. Users with Tiles retain better than users without.

It indexes toward reaching more users rather than maximizing device margins, though the devices are not a loss leader.

Sub Club▶ 38:47

Ads and partner offers must first do no harm. Ads started only in the last 2 years of a roughly 15-year-old company, checked for intrusiveness through attention, engagement and user conversations. Useful offers use location and context, like an Uber partnership that grew from a hackathon landing-notification feature.

"The kind of baseline is do no harm, meaning make sure that the ads are not intrusive, not invasive, and they're not unwelcome by the audience."

Sub Club▶ 43:40

New features must solve real, adjacent problems inside one app. For its family super app goal, Life360 asks whether a problem is real, whether it has permission to solve it, and whether the audience overlaps with existing users.

Few companies succeed with a second app, citing Duolingo bringing its separate apps back into the main one, so features go into one app and the ML layer decides who sees what.

Sub Club▶ 40:28