When I was a freshman in college, Instagram was all the rage.

Instagram was becoming the primary medium for keeping tabs on friends and influencers alike—perceiving the world through their iPhone lenses, in a way. A simple scroll through an Instagram feed would inform me what my friends ate for dinner, where my ex’s family vacationed, and which Uggs Kim Kardashian wanted me to buy.

It was truly the platform for keeping up—until it made me fall behind.

One day, Instagram released the Stories feature, granting users the ability to post temporary images under their profile. Originally, this was a feature that only Snapchat offered. Because of users’ preexisting familiarity, it was widely adopted almost immediately.

Instagram stories v1

And overnight, what used to be a news feed experience app transformed into a news feed and stories combo. Instagram posts “on the grid” were still comprised of momentous occasions and good photography, but Stories connected the dots and allowed users to capture what they were doing on that specific moment.

And that’s what users did. Except for me.

I didn’t have the Stories feature.

I couldn’t figure out what my friends were doing in order to post stories. I scoured the Share menu, deleted the app, re-downloaded it, toggled my account to a business account and back, updated my operating system, Googled for days on how to “unlock” the feature, and so on, and so on.

Eventually, I threw in the towel and emailed Instagram support, begging for the feature.

The email went something like this, if I'm remembering correctly.

They said no, in case you were wondering.

Meta has been doing rollouts for a while

Missing out on Instagram stories (for a week) was jarring enough, but it reminded me of something from recent memory.

Around February, the same year, I opened the Facebook app and was bombarded with purple flower reacts on every post. Facebook had recently started rolling out a new “Thankful react,” and users were using it to react to literally every single post:

• Just got fired from my job: Thankful 🌸

• Looking for carpenter recommendations: Thankful 🌸

• A compilation of Mark Zuckerberg talking about barbecue sauce: Thankful 🌸

• This thankful react thing needs to stop: Thankful 🌸

Much like with Instagram Stories, I was doomed to sit and watch purple petals cascade down my news feed, as I was apparently in the control group yet again.

Hilariously, the internet was full of flash-in-the-pan SEO-optimized articles and videos instructing users on phony ways to “unlock” the thankful react. Some top contenders:

• React to 10,000 posts

• Tag Mark Zuckerberg in a Facebook post

• Sign up for my random newsletter

Of course, none of these worked (no I didn’t try), but I wonder how many users fell for them. I’d like to think zero, but then I remember that Facebook users sometimes call the police when the app isn’t working, so I’m not really sure.

Eventually the Thankful react went away. I’m not sure if it was always meant to be a temporary feature, or if it simply didn’t increase the metrics that Meta had hoped. Likely the latter, if they were adhering to an experiment review culture.

The rollout standard in tech companies

Meta’s method of gradually releasing features is part of a broader strategy known as staggered rollouts.

This is something that tech companies rely on to ensure the stability of new features, minimize potential risks, and gather valuable data before a full launch.

🧠 Related reading: How to debug your experiments and feature rollouts

Take Instagram’s Stories feature, for instance. Although it was a huge hit from the start, my experience as a user who didn’t have access for an agonizing week highlights how these rollouts aren’t always seamless.

It’s not uncommon for companies to introduce new features to a select group of users, often referred to as control groups or test groups. This way, they can observe how the feature is used, identify bugs, and collect feedback without overwhelming their entire user base.

Take Spotify Wrapped, for example. This ultra-popular Holiday-themed feature started rolling out towards the end of the year, but many users didn’t have it until just before the new year. In jealousy they would have to sit and watch their friends’ cool music stats without even knowing their own. 😞

Google, another tech giant, often does this with its products. You might notice new designs or functionalities in Google Maps, Gmail, or YouTube long before your friends do—or vice versa. This isn’t just about making sure everything works; it’s about understanding user behavior at scale.

Sometimes, a feature that looks great in beta testing fails to generate the expected results when rolled out to a wider audience. In that case, the company might tweak it, hold off on a full release, or even scrap it altogether.

A recent example is Twitter's (now X’s) decision to roll out its "Communities" feature to select users. Like Facebook Groups, Communities was meant to foster niche interactions. But due to lukewarm reception, Twitter quietly scaled back the rollout by removing the prominence of the feature.

Rollout strategies are a must

For me, there was a big “aha” moment when I ultimately found myself working at Statsig, switching from user to provider of feature management tools. Seeing all the different options available for making data-driven product decisions helped me fully understand exactly how extensively big tech companies are managing releases.

Statsig users typically manage feature releases through a combination of:

• Feature flags: Toggle switches for system behavior/features in production that allow for gradual rollouts, A/B testing, kill switches, etc.

• Holdouts: Used to measure the cumulative impact of feature releases and check if wins are sustained over time.

• Dynamic configs: Allow for the switching of hard-coded values in applications with server-defined configuration parameters, giving users dynamic control of app behavior in near real-time.

• Experimentation: Statsig experiments are multivariate tests that enable seamless and consistent user experiences, providing reliable insights for data-driven decision-making in product development.

• And more…

At the end of the day, staggered rollouts, experimentations, and user feedback loops are essential in order to be competitive in the tech world.

The next time you’re missing out on a hot new feature or seeing a strange button no one else has, remember: you're likely part of a much larger experiment.

Request a demo

Statsig's experts are on standby to answer any questions about experimentation at your organization.

Grab a Demo