Sieve — a free scan that finds where your website leaks customers
Sieve
Where is your website losing customers?
Free. No signup. About a minute.
What does it actually check?
Whether the page asks the visitor to do anything and whether that ask is findable; whether it works on a phone; whether anything on the site suggests other people have bought; whether a price exists anywhere; how long the page takes to arrive; and whether search engines and shared links have anything to work with. Around thirty-five checks, plus a written review of the copy.
Will it tell me how much money I am losing?
No, and be suspicious of anything that does. We cannot see your traffic or your sales, so any euro figure would be invented. You get the problems, how sure we are about each one, and what to change.
Do I need to install anything?
No. Type the address and press scan. There is no account, no tag to add and nothing to verify — everything is read from the public version of your site, the same way a visitor sees it.
How long does it take?
About a minute. It reads your homepage and up to four pages a customer would go looking for — pricing, booking, shop, contact — one at a time, slowly enough not to bother your server.
What is in the paid report?
Every finding with the reason it costs customers and the specific change to make, plus the five worth doing first, ordered by what they are worth against how long they take. €19, once, for that one report. No subscription.
Is this just Lighthouse with extra words?
No. Speed is one of six areas and the smallest one. The question here is why somebody would visit and not become a customer; performance is only ever evidence towards that, and most sites lose people for reasons a performance tool has no opinion about.
What this cannot do
It does not look at your site. Not the way you do. It reads the markup, the headers and the timings — there is no browser here drawing the page — so it can tell you that a button says “Learn more” and cannot tell you that the button is beige on beige.
It does not know your numbers. No traffic, no conversion rate, no revenue. Every finding is a probable leak with a stated confidence, never a counted one, and nothing in the report is expressed in money.
It is polite about it. The scan reads five pages, one at a time, obeys robots.txt, and identifies itself as SieveBot. A tool that floods a small business’s server to tell them their server is slow has caused the problem it is selling.