Teardown: Marrowlight's Pricing Page

Marrowlight is a log-analytics tool I made up for this post, with a pricing page I've seen a hundred real ones imitate. Three columns, a highlighted middle, a "talk to us" enterprise link. Teardowns are a good way to learn pricing because the page is a small, public argument, and arguments have weak points.

Here is the page as I imagine it:

Do the per-seat maths first

Before any opinion, divide. It takes three lines and it's the fastest way to see what a page is really saying:

plans = {"Solo": (12, 1), "Team": (49, 5), "Studio": (149, 20)}
for name, (price, seats) in plans.items():
    print(f"{name:7} ${price:>3}/mo  {seats:>2} seats  ${price / seats:5.2f} per seat")

Solo    $ 12/mo   1 seats  $12.00 per seat
Team    $ 49/mo   5 seats  $ 9.80 per seat
Studio  $149/mo  20 seats  $ 7.45 per seat

Per seat, the price falls from $12.00 to $9.80 to $7.45. That's a volume discount, and it's a reasonable one. The page, though, never shows it. It shows monthly totals, so a buyer has to do the division I just did. Most won't.

The middle plan is the anchor

The highlighted Team plan isn't there to be the cheapest or the best. It's there to make the other two look like decisions. Solo at $12 looks tiny beside it. Studio at $149 looks expensive until you notice it has four times the seats for three times the price.

If the goal is to sell Team, which is where most small companies will land, the page is doing its job. If the goal is revenue per customer, the weak point is the gap between $49 and $149. A team of six needs the expensive plan for one extra seat, and I'd bet that's the most common reason for churn on a page like this: not that the product disappoints, but that the nearest step is steep.

Three things I'd test first

1. Show the per-seat price under each total. It costs nothing and it turns an implied discount into a visible one.

2. Add a seat add-on to Team. Let the six-person team pay $49 plus a few dollars for the sixth seat. Compare plan-upgrade rate before and after.

3. Move retention above seats in the table. For a logging tool, how long you keep data is the real limit, and it's the line a buyer checks first.

What I can't tell from a page

A pricing page is a hypothesis. I can tell where the arithmetic is lopsided, but I can't tell which customers actually churn at six seats. That takes a dataset. If you run something like this, pull the last twelve months of cancellations and sort them by seat count before changing a single number. Anything you read here, test against that.