A lot of what gets written about analytics assumes you have a data team, a warehouse and a budget for software licences. Most businesses I speak to don't. They have a shop or a service, an accounting package, perhaps a website, some orders arriving on a marketplace or WhatsApp, and a feeling that they should be doing more with all of it.
Good news: you can get a lot of value without any of the heavy machinery. Here is how I'd go about it.
Start with three questions
Before touching any tool, write down the three questions you most wish you could answer. Not twenty. Three. For a small business they are usually things like:
- Which customers or products actually make us money, after costs and returns?
- Where do new customers come from, and what does each one cost to win?
- Which customers come back, and what makes them do it?
Everything you do next is aimed at answering these. If a piece of data doesn't help with one of them, leave it for now.
Find where your data already lives
Most businesses are sitting on more than they realise.
Sales and invoices. Your accounting software or billing system. Tally, Zoho Books and similar tools can export sales by customer, product and date.
Marketplaces. Sellers on large marketplaces can usually download order, return and settlement reports from their seller dashboard.
Website and ads. Google Analytics for the website, plus the reporting inside Google or Meta ads.
Payments. Your payment gateway's reports, which often have details that your own system doesn't.
Conversations. Orders and enquiries that come through WhatsApp, calls or email. These are the hardest to count, so decide early on a simple way to log them, even if it's a shared sheet where someone types a line per order.
One place, one sheet
You don't need a database yet. Start with a spreadsheet, or a set of them, where each row is an order and each column is something you know about it: date, customer, product, price, channel, city, whether it was returned. Keep it in one consistent format.
This single, tidy table will answer most of your first questions with pivot tables. It's also the thing you'd later load into a proper database or BI tool, so the work isn't wasted.
The commonest mistake is to build lots of different sheets for different purposes with different formats. Resist that. One table, kept clean, beats ten clever ones.
Free and cheap tools that go a long way
Google Analytics will tell you where website visitors come from and what they do. Google Search Console shows what people search for before they land on your site. A free dashboard tool such as Looker Studio can turn a spreadsheet into something your team can look at without opening Excel. Power BI Desktop is also free to use for building reports on your own machine.
Be careful about adopting expensive software before you have the habit of looking at the numbers. Tools don't make people use data. Routines do.
Build a weekly habit
Pick five numbers. Orders, revenue, average order value, new customers, repeat customers, for example. Look at them every week at the same time, with the same people, and ask one question: what changed, and why?
It sounds small, and it is. But a business that looks at five numbers every week will spot a problem months before a business that looks at fifty numbers once a quarter.
Mistakes I see
Counting everything and measuring nothing. A dashboard with forty charts and no decision attached.
Trusting numbers without checking them. Do a spot check. Pick ten orders and trace them from the sale to the payment. If they don't match, fix that before building anything else.
Ignoring returns and cancellations. A product that sells well and comes back often may be losing you money.
Not respecting customer data. India's Digital Personal Data Protection Act, 2023 sets out duties for how businesses handle personal information. Keep only what you need, be clear about what you'll use it for, and keep it safe. It's good practice even before it becomes a legal worry.
When to get help
You'll know it's time when the questions outgrow the spreadsheet. Maybe the file takes minutes to open, or two people keep getting different totals, or you want to forecast demand or predict who is likely to stop buying. That's when it's worth talking to someone who can set up proper tracking, a small database and a few dashboards.
Until then, start small, keep it tidy, and look at your numbers every week.