Step 1. Know the business you're in

Before you can build great analytics, you need a clear and honest picture of the business you're working in. Not how leadership describes it in a strategy deck, but what it actually does: who it sells to, how it makes money, what it depends on, and where it spends.

The tool I use for this is the Business Model Canvas (BMC). It's a single page, nine sections, and when filled out correctly it surfaces what is actually important about a business in a way that a hundred-page strategy document rarely does.

How to fill out the BMC

I follow a specific order when filling out the BMC:

Customer segments. Who do you sell to?
Value proposition. For each of your customer segments, why do they choose you?
Channels. How do you reach your customers?
Customer relationships. What is the nature of the relationship you have with each of your customers? Do you serve them directly, or does a partner do it for you?
Revenue streams. How do you generate revenue from the value you provide?
Key activities. What do you need to do in order to fulfil the value proposition?
Key resources. What do you need to have in order to fulfil the value proposition?
Key partners. Who helps you to fulfil the value proposition?
Cost structure. What are the key costs associated with fulfilling the value proposition?

Mistakes to avoid

There are some classic mistakes with the BMC that are easy to avoid if you know about them:

Using too much jargon. If you want the BMC to be useful, write in plain language.
Not being specific. For example, "Everyone" is not your customer segment.
Confusing product/service features with the value proposition. Your customers don't care about your product or service. They care about what it does for them. This is your value proposition.
Listing what they want to see vs what is today. There will be time later to understand the goals of your business. This is about understanding and documenting what is today.

Examples

To make this guide concrete, I'll use two running examples throughout: a mid-market B2B SaaS company and a non-bank residential mortgage lender. We'll follow both businesses through each step of the process so you can see how the output differs depending on the kind of business you're working in.

Example 1 — B2B SaaS

A mid-market software company selling project management and workflow tools to businesses with 50 to 500 employees.

Customer segments. Team leads and operations managers who buy the tool for their teams, and the finance or leadership sign-off that often sits above them. Primarily in professional services (consulting, agencies) and technology companies.
Value proposition. One place to plan, track, and report on work across teams. Easier to set up than large enterprise tools, more structured than spreadsheets, and it saves team leads hours of manual reporting each week.
Channels. A free trial that anyone can sign up for, paid marketing (search and social) to drive signups, a sales team that reaches out to larger accounts, and consultancies that recommend and set up the tool for their clients.
Customer relationships. Smaller accounts set themselves up with minimal help. Larger accounts get a dedicated person to support them, particularly around renewals.
Revenue streams. Monthly or annual subscription fees charged per user. Larger accounts pay for setup and onboarding support.
Key activities. Building and improving the product, marketing to attract signups, getting new customers set up successfully, and keeping existing customers happy enough to renew.
Key resources. The software itself, the team that builds and supports it, and the data on how customers use the product.
Key partners. The cloud providers that host the software, and the other tools it connects to (email, accounting, CRM) that make it more useful.
Cost structure. Engineering and product salaries (the largest cost), hosting and infrastructure, sales and marketing, and the team that supports customers.

B2B SaaS Business Model Canvas

Notice the shape of this business: revenue is predictable month to month, but growth depends on getting people to sign up, convert from free to paid, and stay. The analytics questions that matter most will be about the signup funnel, which customers churn and why, and which parts of the product keep people coming back.

Example 2 — Non-bank lender

A mid-sized non-bank residential mortgage lender operating in Australia. These businesses are a useful case study because they serve multiple customer types, depend heavily on external funding, operate in a regulated environment, and have a complex chain of partners between them and the end customer.

Customer segments. Home buyers (both straightforward and more complex credit profiles), property investors, self-employed borrowers who don't fit the major banks' standard criteria, and mortgage brokers who act as the main route to market.
Value proposition. Faster decisions than the major banks, more flexibility for borrowers who don't fit standard lending criteria, and a smooth experience for brokers that makes it easy to recommend them to clients.
Channels. Mortgage brokers (the dominant channel, typically 80-90% of volume), a direct online application portal for borrowers who come without a broker, and the aggregator platforms that brokers use to submit deals.
Customer relationships. The initial transaction is usually handled through a broker, but once a loan is settled the lender manages the ongoing relationship directly through loan servicing.
Revenue streams. The margin between what it costs to borrow money and what they charge customers (net interest margin), upfront fees charged at settlement, and ongoing fees paid by brokers who refer business.
Key activities. Assessing and approving loan applications, settling and funding loans, managing the loan book, engaging and supporting brokers, and staying compliant with lending regulations.
Key resources. The loan origination and servicing systems, the credit assessment process, access to wholesale funding, and the team of credit, operations and broker-facing staff.
Key partners. The banks and institutions that provide funding, credit bureaus that supply borrower credit data, broker networks and aggregators, property valuers, and legal and settlement service providers.
Cost structure. The cost of borrowing money to on-lend (the largest cost by far), broker commissions, staff, technology, credit checks and compliance.

Non-bank lender Business Model Canvas

Notice how specific this is. We're not saying "people who need loans" — we're identifying the actual segments and why they come to a non-bank rather than a major bank. This specificity is what makes the BMC useful for analytics. If you know that self-employed borrowers are a key segment, you can start thinking about what questions matter for that segment specifically.