Odbierze AURA — ten sam asystent, z którym rozmawiasz na czacie. Opowiedz o firmie własnymi słowami.
Zadzwoń terazPołączenie płatne wg cennika Twojego operatora. Rozmowę prowadzi asystent AI.
AURA może działać przez własny interfejs, aplikację oraz kanały, z których zespół korzysta już dziś. Każdy człowiek i każde narzędzie pracują na tym samym stanie biznesu.
Siedem kroków, które AURA wykonuje bez przerwy — od pojedynczego zdarzenia do wniosku, który zmienia następną decyzję.
…i z powrotem do Observe — cykl nie ma końca.
Działanie bez śladu w danych jest dla nas niedokończone. Ślad ma zawsze te same cztery ogniwa:
Poniższe liczby są ilustracją samego mechanizmu, a nie rezultatem, który osiągnęliśmy u klienta. Pokazują, jak wyglądałby ślad w danych, gdyby taka sytuacja wydarzyła się w Waszym lokalu.
AURA nie przypisuje sobie wyniku bez śladu w danych.
We look for reasons why visitors don't leave an inquiry and remove them one by one — instead of rebuilding the whole site by feel.
Companies with traffic but too few contacts.
Traffic is there, requests are not. The most common reaction is changing the button color or a new photo at the top — basically guessing. Meanwhile the reason can be trivial: a form with seven fields, phone number as an image, payment step that doesn't fit on a phone screen. Without session recordings and funnel nobody will see this because the owner knows their site by heart and can't walk through it like a stranger anymore.
First diagnosis, then fixes — and only those that can be measured.
An audit is a finished document; optimisation is continuous work. You can start with the diagnosis alone and stop there for a while.
The audit itself takes 6 hours – 2 days. Separately we count time needed to gather a meaningful traffic sample: below roughly a thousand sessions conclusions become guesswork, so with low traffic we collect data longer and say so directly. Optimization is monthly work: one or two changes at a time, because ten at once means it's unclear which one worked.
The cost depends on two things: how much traffic has to be reviewed for the conclusions to be conclusions rather than guesses, and how deep the fixes reach — moving a button is not the same as rebuilding the whole path to an order. If the list of fixes turns out to be a rebuild of the site, we say so before we start, not halfway through.
The same number of visits starts generating more inquiries, so every euro spent on ads and SEO works harder. You also get a list of things not to do — and that usually saves more than the fixes themselves.
Before you pay for a change, its worth can be estimated. The formula: (visits to that step per month x the lift in the share who move on) x your share of deals closed x margin per deal = the monthly value of the fix. Visits to the step is not traffic to the whole site, only the people who reached the form or the basket; that number comes from the funnel. The lift in the share who move on is the one figure nobody knows in advance - you enter your own assumption there, then after the test you put in the result and compare it with the assumption. Margin is what is left, not turnover. If the value of a fix is lower than its build cost spread over a year, the fix waits and we do a different one.
There are three situations where we honestly talk you out of it. First: too little traffic. Below roughly a thousand sessions, conclusions from recordings and the funnel are a story about a handful of people, not knowledge about your customers - so traffic first, audit second. Second: a site that is going to be rebuilt in the coming weeks anyway. A list of fixes for a site that will not exist in a month is work for the bin; better to enter the new build and guard the path there. Third: the problem sits outside the site - the price is off the market, the offer is unclear, or nobody answers enquiries for two days. No button change repairs that, and we do not pretend it does.
These are two different rhythms of the same work, with different effects. The choice depends on how much traffic you have and who on your side is going to implement the changes. Three options we meet most often in practice.
What we do not recommend: monthly experiments on traffic where the difference between two weekdays is larger than the effect of the test. Then you are paying for charts, not for knowledge.
The site address alone is not enough for an audit. We need access to a measurement tool with history, and if there is none - permission to install measurement plus a few weeks to gather data; we say so immediately, so nobody waits for a report that has nothing to grow from. We also need your definition of an enquiry: does a submitted form count, a phone call, a booking, or only a confirmed job. Without that we count different things and argue about the result. Plus access to the mailbox or system where enquiries land, so we can compare forms sent with forms actually received - that comparison alone uncovers part of the faults.
A checkable result is not 'it looks better', it is four things at once. First: we change one thing at a time, because with three at once nobody can say which helped and which got in the way. Second: we compare the same slice of the week with the same one - Mondays against Mondays, not a working week against a long weekend. Third: we look at the step we changed and at the end of the funnel together; a fix after which more people enter the form and fewer send it is a step backwards, not a success. Fourth: the result has to hold through a second identical period. One good week is weather, not change.
The most common mistake is switching a test off on the day the result looks good. Somebody looks in on Wednesday, sees the new version ahead and declares victory - while the lead has simply not settled yet. It happens because a test costs patience and the answer is wanted today. The second mistake of the same family: changing several things at once to speed things up. Afterwards the site does behave differently, only nobody knows why, so the next decision is guesswork again. That is why we agree in advance on one change and on the period after which we look - rather than looking daily and changing our minds halfway.
It does not increase traffic. We work with the people who already came in, and that is the whole point — if there are too few visits, what you need is advertising, maps or content, not an audit. It does not change your price or your offer: we can show where people drop off after seeing the price list, but the pricing decision is yours. We promise no specific percentage of growth and quote no typical industry figures, because such numbers sound good and mean nothing for your site. Nor do we rebuild the whole site along the way: if a fix turns out to be a rebuild, we say so before we start, not halfway — because that is different work and a different timeline.
Session recordings, click maps, measurement accounts and every report are set up under your accounts, not ours. That means when the engagement ends you keep the history, not just a memory of it - the next person who sits down with this site has something to compare against. You get the audit document in a form that can be handed on: a list of fixes with a note on what each one covers and how to tell it worked. Access is returned on request at any moment, with no transition period. Recordings that touch personal data are configured with field masking - that is not a courtesy but the condition for collecting them at all.
If your problem sounds different, neighbouring areas of the same system stand right next to it — AURA connects them to each other rather than selling them separately:
Tell us how this process looks at your company today: how many enquiries come in, who answers them and where they get lost. We will tell you what can be taken off a person, what is not worth touching, and how this area fits into the rest of the system.
Talk to Aura →Next step
We do not take everyone: first we look at your processes, sales and current systems and tell you honestly whether it makes sense for us to come in. A few questions, about five minutes.
Take the assessment →