For twenty years, the BI industry has sold small business a smaller version of something that was never designed for it.
Enterprise-grade dashboards, priced down. Enterprise-grade consultants, on "SME packages." Enterprise-grade software, with "starter tiers." The packaging changed. The product didn't.
And the pattern is predictable: the tool gets bought, a dashboard gets built, someone spends three months trying to make the numbers match what the P&L already says, and eventually it gets quietly abandoned. A spreadsheet comes back. The owner returns to running the business on gut feel and the monthly bookkeeper email.
This isn't a tooling failure. It's a design failure.
What enterprise BI is actually designed for
Enterprise BI was built for organisations with three things a small business doesn't have:
- A data team. At least one analyst, usually more. Their job is to keep dashboards accurate, answer one-off questions, and maintain the connection between source systems and reporting layers.
- Governance infrastructure. Data dictionaries, version control, approval processes. These aren't bureaucratic overhead — they're what stops three teams from each having their own definition of "revenue."
- Change tolerance. Enterprise implementations assume a 6–12 month deployment window with external consultants, staff training, and iteration. The business can absorb that cost because the productivity gain from a working BI function pays back over years.
Small businesses have none of this. No analyst. No governance layer. No capacity to absorb a year-long implementation before seeing any value.
So when "enterprise-lite" BI gets sold into a small business, every single assumption the product was built on is wrong. The dashboards still need maintaining — by someone who doesn't exist. The definitions still need governing — by a process that doesn't exist. The deployment still takes months — that the business can't afford.
It's not that the tools are bad. It's that the tools are answering the wrong question.
The right question for small business BI
Small business owners don't need more data. They need a clearer answer to one question: what's actually happening in my business, and what do I do about it?
That question has three practical parts:
- What are the 5–10 numbers that actually matter for my business?
- Which of those are reliable, and which are guesswork dressed up as data?
- What does each one tell me to do — if anything — this week?
Notice what's not on that list: a real-time dashboard. Predictive analytics. AI-powered insights. Self-service reporting. These are answers to enterprise questions — not small business questions.
The small business BI job is smaller, sharper, and almost entirely about trust.
You need 5–10 numbers you can trust. You need a clear line from each number to a decision. And you need the whole thing to survive the fact that no one in the business has "BI" in their job title.
What actually works
Over the last two decades inside SME and enterprise BI teams, the pattern that works for small business is almost boringly consistent:
1. Start with definitions, not dashboards.
Ninety percent of SME BI failures trace back to this: five people in the business each have a slightly different definition of the same metric. Revenue with VAT or without. Customers including trial accounts or not. Margin before or after returns. Until the definitions are agreed and written down, every dashboard built on top is arguing with itself.
2. Audit what you already have.
Most SMEs have more data than they think. Accounting system, CRM, POS, stock system, ad platforms — each one stores something. The first job isn't to bring in more; it's to see what's already there and decide what's worth trusting.
3. Build for one person to maintain.
If the BI system needs a specialist to keep running, it's wrong. The system a small business can sustain is one where the owner, a bookkeeper, or an operations person can walk in after six months and still understand it. That constraint is a feature, not a limitation.
4. Answer-oriented, not report-oriented.
Every report should answer a specific question. "How are we doing on cash?" "Which customers are profitable?" "Where is margin leaking?" A report that doesn't answer a specific question is just wallpaper.
5. Decouple the BI layer from the operational systems.
The single biggest source of SME BI pain is building dashboards that break every time the accountant changes something in Xero. The BI layer should be clearly separate — a read-only view that survives operational change.
None of this is new. None of it requires enterprise tooling. It requires someone to sit with the specific business, understand what matters for it, and design a BI system around those facts.
What this looks like in practice
For a £1.5M turnover service business, a working small-business BI function might be:
- A one-page monthly scorecard with 7 metrics: revenue, gross margin, utilisation, cash runway, top-5 customer concentration, win rate, average project margin
- Definitions for each one written down and agreed with the bookkeeper
- Automated where possible (pulled from Xero + project tool), manually updated where not (rare — once a month the owner checks two things)
- Reviewed on the same Monday every month, with the same four questions asked: what moved, why, what do we do, what do we watch next month
Total time to build: two to four weeks. Total time to run: thirty minutes a month once it's running. Total cost: nothing that compares to a $50K BI consulting project.
This is the answer the industry has failed to deliver for twenty years. Not because it's hard, but because it's less lucrative than selling $10K/month software licenses and 6-month implementations.
Where The BI Playbook fits
Everything published here is built on this thesis. The BI Baseline Score is a two-minute assessment to see where a business currently sits. The BIP Method™ is the four-stage sequence for going from "we don't know what we don't know" to "we have a working scorecard and we trust it." The Manifesto is the longer version of the argument this post makes.
The industry hasn't built BI for small business. So we're building it here — in public, one piece at a time, for people who run businesses not data teams.