Every plan has every feature

There is no tier that withholds the part that makes Scout worth having. Plans differ by how many visitors a month Scout is watching, because that is what decides how much work it is doing for you — and it means that if the pixel ever stops reporting, your bill goes down rather than up.

14 days free on every plan. Billing runs through Shopify, so it lands on your Shopify invoice and cancels from the same place as every other app.

In every plan, including the first one

Questions

How is this different from Shopify Analytics?
Shopify Analytics answers 'what happened'. It is good at it, and Scout does not try to replace it. Scout answers the two questions after that — why did it probably happen, and what should I do about it. If all you need is yesterday's revenue, use Analytics; it is already open.
What am I actually paying for?
Visitors a month, counted by Scout's own pixel. It is the denominator of every figure Scout reports, so it is the one number that decides how much work Scout is doing for you — and it means that if the pixel ever stops reporting, the bill goes down rather than up. Orders would charge a high-traffic store that converts badly the least, which is exactly the store with the most for Scout to find. Every plan has every feature; the only thing that changes is the ceiling.
Is there a trial, and how does billing work?
14 days, and the length is chosen rather than copied. Scout needs about a week to learn what a normal Tuesday looks like for your store before it can tell a real drop from a weekday — so a seven-day trial would end at the moment it first had something worth saying. Billing runs through Shopify, so it appears on your Shopify invoice and cancelling is done from the same place as every other app.
What happens if I go over the visitor limit?
Nothing stops being watched and nothing is deleted. Scout keeps reporting and tells you the band you are in, because a store that grew past a tier is a store the product is working for. Sustained growth is a conversation rather than an automatic invoice — a single busy month, which is usually a sale, is not a new plan.
What if the diagnosis is wrong?
It tells you which it is. Every investigation separates what Scout measured from what it inferred, and the inference is labelled as one with a confidence level beside it. Scout cannot see your rendered page, so when it says the add-to-cart button is below the fold, that is a conclusion from the data rather than something it looked at — and it says so in those words.
Does it change anything on my store?
Nothing you can see, and nothing you own. Scout never touches a product, an order, a theme, a discount or a customer. The one thing it does create is its own analytics pixel, because measuring sessions and checkout steps is impossible without one. That permission cannot reach anything else, and the code refuses any other write at two separate points rather than relying on us to behave.
How much work is it to set up?
One click. Scout installs its own pixel during setup — there is no tag to paste into a theme, no developer, no data warehouse and nothing to configure before it starts. Connecting Google Analytics is optional and takes a further click; every page works without it.
What data does it read?
Orders, sessions, products, checkout steps, and the theme and app events Shopify records. No customer names, addresses or payment details — none of it would make a diagnosis better, and all of it would make a breach worse.
How long before it is useful?
It needs about a week to learn what a normal Tuesday looks like for your store. Before that it can still catch large absolute moves, but it will not know that your orders always dip on a Friday — which means some of what it flags in the first week is seasonality rather than a fault. It says so rather than leaving you to discover it.
How does it decide what to do first?
By impact × confidence ÷ effort, and all three are shown on every row so you can disagree. Effort is why a $2,600 fix you can make from a settings page outranks an $8,200 one needing a developer — that is the right order for someone with one afternoon, and the wrong order for someone with a quarter. Sort by money if you have the quarter.
What happens when a recommendation fails?
It stays in the ledger with the reason it did not work. A bundle recommendation that lifted average order value 7% and lost money because conversion fell further belongs on the Impact page as a loss. A tool that quietly drops its failures is a tool whose successes you cannot evaluate.