Most small business owners I talk to have the same problem with reporting. It isn't a lack of data. It's too much of it, spread across too many logins, with nobody deciding what matters.
Google Analytics alone has hundreds of reports. Add Search Console, Google Ads, Facebook, your Business Profile and your CRM, and you could build a dashboard with 80 charts on it. You'd look at it twice.
A good dashboard for a small business is short. It answers a few questions you actually ask, and it answers them the same way every week. Here's what I put on one, and what I leave off.
Start with the questions, not the charts
Before anything gets built, write down the questions you want answered. For most service businesses, the list looks like this:
- How many leads did we get this week, and is that up or down?
- Where did they come from?
- What did each lead cost us?
- Is anything broken or dropping off?
That's it. If a chart doesn't help answer one of those, it doesn't go on the main screen. It can live on a second tab for the day you need it.
The numbers that earn a spot
Leads, counted the same way every time. Form fills, phone calls and booked appointments, added up. This is the number at the top of the screen. If your calls aren't tracked, that's the first fix, because for a lot of local businesses calls are most of the leads.
Leads by source. Google Ads, your Google Business Profile, organic search, Facebook, referrals. This is where a dashboard pays for itself, because it shows you which channel is doing the work and which one is just spending money.
Cost per lead, by paid channel. Ad spend divided by leads from that channel. Not clicks, not impressions. Cost per lead is the number you can compare across Google and Facebook and against what a customer is worth to you.
Website visits from search. Organic traffic from Search Console or Analytics, as a trend line. You don't need the daily number. You need to know if the line is going up, flat or down over a few months.
Business Profile activity. Calls, direction requests and website clicks from your Google listing. For a local business, the map listing often sends more calls than the website does, and most reports skip it entirely.
Something from the sales side. If you have a CRM, pull in how many leads turned into booked jobs or sales. Even a rough close rate changes how you read everything above it.
The numbers to leave off the main screen
Impressions, follower counts, page likes, average time on page and bounce rate all have their uses. None of them tell you whether marketing is bringing in business this week. On a small business dashboard, they mostly create noise and false comfort.
The same goes for anything you can't act on. If a number goes down and you wouldn't do anything differently, it doesn't belong on the first screen.
Show the trend, not just the total
A single number with no context is hard to read. Was 48 leads good? Compared to what? Every number on the main screen should show a comparison: last week, the same week last month, or the last 12 weeks as a line. That's how you spot a slow decline before it becomes a bad quarter.
Make it refresh on its own
A dashboard that someone has to update by hand stops getting updated. It's not a discipline problem. Everyone gets busy. The numbers should refresh daily or weekly without anybody touching them, and something should flag it when a connection breaks so you're not reading stale data without knowing it.
Check the numbers before you trust them
The fastest way to lose faith in a dashboard is to catch it being wrong. Before you rely on any number, it should be matched against the tool it came from. If Analytics says 30 form fills and your inbox has 22, somebody needs to find out why before that number goes on a screen.
This is the part AI agents can't do on their own yet. They're good at connecting sources and building charts quickly. They are not good at noticing that a number looks off for your business. That still takes a person who knows what normal looks like.
Add a short written summary
The best reporting I've seen pairs the dashboard with a few sentences each week: what moved, the likely reason, and what to do about it. The chart shows the what. The note covers the why. Together they take about two minutes to read.
Start small and add views as you go
Your first version should fit on one screen. Once you're actually using it, you'll start asking new questions, like leads by service area or calls by day of the week. Add those as new views when you need them. A dashboard that grows with your questions gets used. One that tries to answer everything on day one gets ignored.
