Getting started with data and analytics

Most businesses already have plenty of data: sales in the shop or accounts system, customers in a database, ad results in Google and Facebook, and a dozen spreadsheets. The problem usually isn’t a lack of data. It’s that the numbers live in different places, disagree with each other, and take too long to pull together.

This guide covers where to start.

Step 1: start with questions, not tools

Write down the five to ten questions you’d most like answered every week or month. For example:

• How many new customers did we win, and where did they come from?
• What does it cost us to win a customer through each channel?
• Which products or services make the most profit?
• How many customers come back, and how much are they worth over time?
• Where are we losing people: enquiries that don’t convert, carts that are abandoned?

Every report and tool you set up should help answer one of these. If it doesn’t, you probably don’t need it yet.

Step 2: list where your data lives

Make a simple list of every system that holds customer, sales or marketing information, and who looks after each one. Typical examples:

• Online shop (such as Shopify or WooCommerce)
• Customer database (such as HubSpot or Salesforce)
• Accounting software
• Ad platforms (Google, Facebook and Instagram, and others)
• Google Analytics
• Email marketing tools
• Spreadsheets kept by individual staff

Most businesses find a system or two they’d forgotten about.

Step 3: agree what the numbers mean

Many arguments about data are really arguments about definitions. Agree, in writing:

• What counts as a sale, a lead or an active customer
• Which system is the final word for each number. For revenue, it’s usually the shop or accounts system, not the ad platforms.
• Whether figures include tax, shipping, refunds and discounts

Step 4: get your systems talking

Exporting spreadsheets by hand is slow and error-prone. Most modern tools can share information automatically, either through built-in integrations or through connector tools. Start with the connection that answers your most important question, for example linking ad results to real sales.

As the business grows, many companies bring all their data into one central store, often called a data warehouse, so reports can combine information from every system without anyone copying and pasting.

Step 5: build one report people trust

• One page, a few numbers. The handful of figures that answer your key questions.
• Updates itself from the source systems.
• Same definitions every time.
• Reviewed regularly, with decisions made from it.

A simple, trusted report used every month beats a complicated dashboard nobody opens.

What about “big data”?

“Big data” usually means volumes of information too large or fast-moving for spreadsheets and everyday tools, such as website clickstreams, sensor readings or millions of transactions. Most small and medium businesses don’t need big data technology to start. The value comes from connecting and trusting the data you already have. When volumes do grow, cloud data warehouses can handle very large amounts without you running your own servers.

Look after the data

• Limit who can see customer information to people who need it.
• Remove access when staff or suppliers leave.
• Know what personal information you hold and why.

Our cyber security fundamentals guide covers the basics.

Common mistakes

• Buying a dashboard tool before agreeing what to measure.
• Trusting each platform’s own numbers instead of checking against real sales.
• Reports that take days to build by hand and are out of date when they arrive.
• Nobody owning the numbers.

Where Blend can help

Take the free Data Maturity Assessment for a score across six areas of your marketing and sales data, or see how we connect systems and build reports.