49. Analytics a Founder Can Check in Ten Minutes

Analytics tools can swallow a morning. A founder does not have that morning. You need a ten-minute check that answers whether the site is helping the business, not whether you have become a person who collects charts. The check should fit on a note you can repeat every week or every fortnight. If it needs a training course, it is the wrong check. Pick a few figures that relate to enquiries, calls, and the pages you actually care about, and ignore the rest until you have a specific question.

This is not a claim that numbers run the company. A quiet week in a shop can have a loud website, and the reverse. The point of the ten minutes is to notice breaks and obvious shifts. The contact form that stopped sending. The campaign page nobody opened. The article that suddenly brings the right kind of buyer. You are looking for surprises you can act on, not for a score.

<>

Decide the questions before you open the chart

Write three questions and keep them. Did people reach the page that asks them to act, and did they act? Which pages did new visitors see first? Did anything important break, such as a sudden drop to zero on a form that usually moves? Those questions are enough. They stop you from wandering into charts about average session theatre that you will not change this month. If you run ads or a seasonal push, add one question. Did the people from that push land on the right page?

Define an action in your own words. A form sent, a tap on the phone number, a booking started. If your tool cannot see the phone tap, you can still count calls with a notebook at the counter and compare weeks. A tally is allowed. The goal is a founder check, not a perfect instrument. Do not pay for a complicated setup before you have used a simple one long enough to know what you wish it included.

Look at the same small set each time

Open the same report, for the same comparison, such as this week against the previous week. Look at the number of visits to the homepage, the main service page, and the contact or booking page. Look at the count of the action you defined. Glance at the list of pages that brought people in. If one page collapsed or spiked, click it and ask why. A spike from a single odd page may be a mislinked advertisement or a robot. A collapse on the booking page may be a broken button. Ten minutes is enough to spot those shapes.

• Visits to the homepage and the main offer page.
• Visits to contact or booking.
• The action you care about, even if it is a manual tally.
• Any page that suddenly fell to nothing.
• One sentence on what you will do, or why you will do nothing.

Ignore vanity that you cannot tie to a conversation. A large number of brief visits from another country, a pretty map, or a ranking you cannot act on this week can wait. If you sell only in Penang, a spike elsewhere is a curiosity unless it is costing you money in ads. Curiosity is for a longer session you schedule on purpose. The ten-minute check is for the shop you run.

<>

Write one sentence after you look

The sentence is the product of the ten minutes. Booking page steady, two forms, phone tally up, no change. Or service page dropped after we renamed it, ask the maintainer to redirect. Keep these sentences in a note. After two months you will see patterns a single dashboard glance hides, such as a form that always dies after a plugin update. The note also helps a staff member cover for you. They can repeat the check without inheriting your instincts.

Ten minutes is useful only if it ends with a sentence you could say to your maintainer or to yourself.

Do not change the website during the check. The check is for noticing. Changes belong in a short list you handle afterwards, or hand to the person on the retainer. Mixing the two is how a quick look becomes an afternoon of tinkering that also ruins the numbers you were trying to read. Look, write, close the tab. Then decide if anything deserves work.

<>

Revisit the questions twice a year

The right questions change when the business changes. A new branch, a shop section, or a campaign habit may earn a fourth line. Remove questions that you never act on. If you have not used a chart in three months, stop pretending it is part of the ritual. A short ritual that survives is better than a sophisticated one you avoid. Teach the replacement founder or the spouse who covers leave. Analytics that live only in one person's head are not a system.

Be honest about what the numbers cannot tell you. They cannot tell you whether a caller trusted you, whether the price was fair, or whether the photo matched the dish. Those answers are still in the shop. The ten-minute check keeps the website from failing in silence. It does not replace talking to customers. Used that modestly, it is worth putting on the calendar like any other small operational task, next to checking that the door was locked and the inbox was read.