Growth Marketer

Marketing was understood for a long time as the work of making messages and sending them out: decide what to say and where to run it, buy the placement, then report impressions and clicks. What the report left out was whatever happened after the click. Whether the visitor reached signup, and whether the new account was ever used, were not marketing's numbers.

So campaign reviews ended the same way every time. Someone reports two million impressions and forty thousand clicks, the CEO asks how many signups that produced, and nobody in the room owns that figure. If forty thousand arrivals turned into four hundred accounts, the problem isn't the ad but the screen that came after it — and that screen belonged to the product team, not to marketing. However good the ad is, money drains out when the part behind it leaks.

The growth marketer role exists to remove that boundary. Creating arrivals and pushing arrivals to the next step become one person's job. That puts the work across more than the ad account: the copy on a landing page, the number of fields in the signup flow, and the email that goes out right after someone signs up.

The first thing on screen each morning is where yesterday's arrivals came from and how far they got, split by acquisition channel. How many came from search, from ads, from referrals, and what share of each reached signup. A thousand people from ads and a thousand from search are not the same thousand. When one converts at 3 percent and the other at 9 percent, where the budget moves is settled on the spot.

Most of the remaining hours go into testing.

1. Picking the target — choose the single biggest leak. If forty thousand ad viewers produced four thousand site visits and four hundred accounts, the gap worth working on is between the site and signup, not between the ad and the site.
2. Building and shipping — write an alternate version of the copy or the screen and run it as an A/B test, which splits arrivals at random and shows the new version to only one side so the results can be compared. Ad copy and other things a marketer controls directly go live that afternoon; product screens get built with an engineer.
3. Waiting — a good-looking number on day two doesn't end the test. Daily swings of that size are normal, so it runs until the agreed number of people have gone through.
4. Calling it and writing it down — the winner opens to everyone and the loser gets deleted. Losing tests get recorded too, with what was tried and how it lost. Without that record the same headline resurfaces two quarters later as a new idea, and the team spends a second week learning what it already knew.

The number behind all of these judgments is customer acquisition cost (CAC), the money spent to bring in one customer. At $0.60 a click, a hundred clicks that produce three signups put roughly $20 behind each account. When that figure runs above what the customer will pay over their life, running the ads better only loses money faster.

Inside marketing, the work touches the performance marketer and the content marketer. Watching the paid accounts day to day belongs to performance; the growth marketer follows whether those arrivals reach signup. What the next screen should be for someone who arrived through an article is decided by the two of them together. At a small company one person holds all three jobs.

Engineers and designers come into every test. A single landing page a marketer can build alone, but cutting fields out of the signup screen means changing product code. The pace of this role is usually set by how fast that round trip runs, which is why many teams reserve a share of one engineer's week for it.

With the product manager the boundaries get agreed in advance, because one side often wants to add a feature to a screen while the other wants to strip it down. With sales the conversation is about lead quality. Plenty of people leaving contact details means little if none of them turn into contracts, so which company sizes actually closed gets fed back into narrowing the ad targeting.

"Clicks are up and signups are flat. Let's look at the landing page before the ads."

Rather than buying more arrivals, fix the place where the arrivals leak out.

"This week's test is one line of copy on the signup screen. Half and half, we read it Thursday."

One small change, split at random, judged on a set date.

"I turned that channel off. It was running $90 per signup."

The cost of an arrival exceeded what that customer would be worth, so the budget came out.

At Pagedeck this is Ravi's job. His quarter is about raising the share of teams that move from the free plan to a paid one, so each morning he checks how many teams signed up yesterday and how many of them created three or more documents.

Last week he tested two versions of the screen that appears right after signup. One opened an empty document; the other showed a meeting-notes template first. The template side produced more first documents, so he and Theo opened it to everyone. The ad budget went untouched through all of it, because the same money bringing in the same number of people still ends up ahead when more of them stay.

The most common assumption is that this is the person who runs the ads well. The title says marketing and the job does touch the ad account, so it reads that way. But more of the work sits behind the ad than in it. Doubling arrivals means doubling the budget, while lifting the share that reaches signup from 3 percent to 6 percent gets the same result on the money already being spent. Hire the role as an ads role and growth stops the moment the budget does.

The other frequent misreading is that the job is finding hidden tricks for free traffic, since the stories that circulate are the ones where a single clever move changed everything. The actual work looks more like running tests that fail seven or eight times out of ten. A company that creates the role expecting one big trick tends to conclude there's no return after five losing tests in the first quarter — which is usually right before the returns start.