Analyze your website traffic

Kite reads traffic from the source that actually measures your site, because the right numbers depend on where the visits are recorded. A team can have more than one website, so Kite confirms which site it's reading before it reports anything.

On a website Kite builds and hosts, Kite's built-in analytics record visitors, sessions, page views, bounce rate, and session duration, along with daily traffic. Depending on your plan, they also show your most visited pages, the sites that referred visitors, and visitors' countries and devices. These numbers live only in Kite. Connecting Google Analytics to a Kite-hosted site wouldn't add anything, because the visits are recorded by Kite's own analytics.

For a website measured outside Kite, connect the analytics product you already use. With Google Analytics, Kite runs reports on your Google Analytics 4 property for sessions, active users, views, and engagement, grouped by page, channel, source, country, or device. With PostHog, it reads visitors, page views, sessions, session duration, and bounce rate, broken down by page, campaign, device, and country.

To see how this month compares with last month, ask Kite for exactly that, and it reads both months side by side. It compares the same number of days in each period, so a month that isn't finished yet isn't measured against a full one.

Here's an example. You ask on the 15th, "How many visitors did we get this month compared with last month?" Kite reports the visitors, sessions, and page views for the first half of this month next to the same days of last month, along with the change in each. If your site had 312 visitors against 268, Kite says that's up 16 percent and tells you which pages and sources the extra visitors came from.

Small numbers get careful treatment. When a site has fewer than a couple of hundred visitors in a period, a percentage can look dramatic while meaning very little. Kite leads with the actual counts in that case, such as "31 visitors, up from 22," and tells you the base is small, so a handful of extra visits doesn't read like a breakthrough.

Your traffic is growing when visitors, sessions, and page views rise over a full period compared with the previous period of the same length. A single good day doesn't count, so Kite always answers for a stated window rather than one day.

Each of those numbers tells you something slightly different:

• Visitors are the number of different people who came to your site. This is the clearest measure of how many people you're reaching.
• Sessions are individual visits. One person who comes back three times counts as one visitor and three sessions, so sessions rising faster than visitors usually means people are returning.
• Page views count every page people opened. More page views per session means visitors are reading further into your site.
• Session duration shows how long a visit lasts, which helps separate people who read from people who glance and leave.

Kite reads these alongside where visitors came from, because growth means more when you know its source. Kite groups referrers into channels. Visits from search engines mean people found you through search, visits from social sites mean your posts are being shared or clicked, and direct visits usually come from people who already know your name. On a site Kite hosts, it also recognizes visits that arrive from ChatGPT, Perplexity, Claude, and Gemini and groups them as visits from AI assistants. Kite tells you which channel moved, not just which website address appeared in the list.

Bounce rate is the share of visits where someone viewed one page and left without opening another. Whether yours is too high depends on what that page is supposed to do, so Kite reads bounce rate page by page rather than as one number for the whole site.

A high bounce rate is normal on a page that answers its question in one go. Someone reads your blog post, gets what they came for, and leaves. Someone finds your phone number on the contact page and calls. Those visits did their job. A high bounce rate on your homepage or pricing page is different, because those pages exist to move people somewhere else, such as a signup form or a booking page. That's where Kite looks first.

It also helps to know that tools define bounce rate differently. Kite's built-in analytics count a bounce as a visit with only one page view. Google Analytics 4 counts a bounce as a visit that wasn't "engaged," meaning it lasted under ten seconds, viewed only one page, and didn't trigger a key event. The same site can show two different bounce rates in two tools, so Kite always tells you which definition it's using and compares a page with its own history rather than with someone else's number.

Traffic only matters if visitors go on to do something, so Kite reads your top pages against your site's goal. A page that gets plenty of visitors but has no clear path toward signing up, booking, or getting in touch is usually the next thing worth improving, and Kite points it out by name.

On a website Kite hosts, Kite connects traffic to what visitors did next. It reads form submissions, clicks on calls to action, and conversions, and it can build a funnel that follows visitors from one page to the next. On a site measured in Google Analytics, it builds the same kind of funnel from your property and finds the step where the most people leave.

When Kite finds a weak page on a site it hosts, it can propose the fix and make the change once you approve it. Then it reads the same numbers again later to see whether more visitors made it through.

Kite explains a change only when it can point to a cause in the same window. If visitors jumped the week a new page went live, or traffic from LinkedIn doubled after a campaign started, Kite connects the two and tells you so. When nothing in the data lines up with the change, Kite says the cause is unknown rather than guessing.

It separates what it measured from what it infers. "Signups dropped the same week the homepage changed" is a strong lead worth testing, not proof, and Kite presents it that way. That keeps you from undoing good work because of a coincidence.

Kite also checks for the dull explanations first. A real zero, such as a day with no visitors, is reported as zero. A day when the analytics couldn't be read is reported as missing data, never as a drop in traffic. On a site Kite hosts, it can also check its error reports, so a broken page that's turning visitors away shows up as a named problem with the pages it affected, instead of an unexplained dip.

When the change is in search traffic and you've connected Google Search Console, Kite can look one step earlier. It reads the searches your pages appeared for, how often they were shown, how often people clicked, and where they ranked. A page that slipped down the results or a query that lost clicks often explains a fall in visitors before anything on your site does.

The simplest way to get a summary is to ask Kite a question in plain words, such as "How did the website do last week?" Kite answers that question first, with the window and source, then adds a short breakdown of the numbers you didn't ask about. If the read is too big for a chat message, Kite can publish it as a findings page or a dashboard at a stable link your team can open.

If you want the summary to keep coming, ask Kite to send it every Monday or on the first of each month. Kite sets it up as a recurring workflow in your timezone. Each run reads the latest period, compares it with the one before, and delivers the summary to the Kite web app, to a Slack channel, or to your inbox.

Add Kite to Slack and it can also keep watch on your traffic between summaries. It checks your site's daily visitors and tells you in Slack when they fall well below their recent level, so you hear about a sudden drop the day after it happens instead of at the end of the month.