Ask HN: In your organization, what does the growth department do?

I think it's a catch all term for something result driven. They manage bugs, products, do sales and marketing, remove legal obstacles, UI/UX, events, PR. Often just identifying what's preventing further growth and calling for company resources into that direction.

Much of the sales on an app isn't marketing/sales at all - look at something like Fortnite. Part of it is community and UX, but the end goal is more money.

Everything from trying to reduce drop-off during initial sign-up, to increasing likelihood of converting from a trial user to a paying customer, all the way to creating in-app cross-sells to recommend our customers try other products we built.

Almost everything we build is initially an experiment, and we have analysts run the statistics against a control group to see if it worked before productionising.

Top-of-funnel improvements -- a/b testing site changes to increase sign-ups, brand awareness/loyalty, etc.

Managing Growth in a late-stage startup: most depts (hr, mkt, sales, cs) perform growth experiments, I coach/facilitate/educate;

Think like this: most metrics can be improved with expertise; known knowns - - but some metrics need new solutions, and for that you need to analyse, hypothesise, experiment... unknown unknowns. And this is why growth is everybody's business.

Growth process is informed with insights from data science + behavioural science (this is my background)

Experiment design + implementation (randomised trials) + interpretation (bayesian based)

Decision to implement in production or scrub and start again.

quick wins: //> a tweak to the search button yields 15% more sales

//> a change in ad copy lifts conversions 17%

//> behavioural recruiting interviews to be more objective...

//> ...and we do around 200 of these per year

I've seen this kind of position in a lot of startups/tech companies ("Growth Manager", whatever that means). To me, it sounds like a product marketing position.

We don't have that.

A/B testing, basically.

it's their job to figure it out ;)