The 10-20-70 Rule for AI, Explained for Small Business Owners
Short answer: the 10-20-70 rule, popularized by Boston Consulting Group, says that in a successful AI project about 10% of the effort goes to the algorithms, 20% to the technology and data, and 70% to people and process: changing how work gets done and getting the team to actually use it. For a small business it means the win comes from fixing the workflow and training the people, not from picking the fanciest model.
Key takeaways
- Most AI projects that fail don't fail on the technology. They fail on the 70%: nobody owned the system, nobody changed the process, nobody was trained.
- For a small business the "10%" is usually already solved; you're buying a proven tool, not inventing one.
- The "20%" is your data: if leads live in five places, that's the first job.
- The "70%" is your week: who answers, who follows up, what happens when you're busy. That's where a consultant should spend most of their time.
Where the 10-20-70 rule comes from
Boston Consulting Group's AI practice has described the split for years as a way to explain why so many corporate AI investments don't pay off: companies over-invest in models and under-invest in the messy human work of changing how a business operates. The exact percentages are a rule of thumb, not a law. The point is the ratio: the people-and-process part is the biggest by far.
What each part means in a small business
The 10%: algorithms
You are not going to train a model. You're going to use one that already exists inside a tool: an AI voice agent, a chat assistant, a drafting assistant in your CRM. This part is mostly done for you. Don't let anyone bill you like it isn't.
The 20%: technology and data
This is your CRM, your phone system, your website, and whether they talk to each other. It's also your data hygiene: are your leads in one place, tagged, with a phone number and a source? If not, no AI can follow up on what it can't see. In our audits this is the most common first finding, and it's unglamorous.
The 70%: people and process
Who answers the phone after 5 PM? What happens to a web form on Saturday? When a customer says "call me next week," where does that live? The 70% is deciding the new process, rolling it out in stages, and training the team so the system survives the week you're too busy to look at it.
How to use the rule when you hire help
- Ask where the hours go. If a proposal is mostly "build" and barely any "train and hand off," it's upside down.
- Insist on a process map before a tool list. Attract, capture, respond, qualify, follow up, book, serve, review. Put the AI where the leak is.
- Roll out in stages. Manual first, then AI-drafted, then supervised, then automated. That's the 70% done safely.
- Name the number. Response time, booking rate, follow-up rate, missed calls. If nothing moves, the process didn't change.
A real example
A service business we work with had "AI" already: a chat widget nobody read. The fix was almost all 70%. We decided that every missed call triggers a text within seconds, that every lead lands in one pipeline, and that the owner reviews a morning summary instead of a voicemail box. Same tools, mostly. Different process. That's the rule in practice.
Frequently asked questions
Is the 10-20-70 rule about budget or about effort?
Effort and attention more than dollars. In a small business the dollars often skew even harder toward the 70%, because the tools are subscriptions and the process change is the real work.
Does the rule apply if I just want ChatGPT for writing?
Yes, in miniature. The tool is the 10%. Deciding what it drafts, who reviews it, and how it fits your week is the 70%. Skip that and it becomes a toy.
Who should own the 70% in a small business?
One named person, not "the team." Usually the owner or an operations lead, with the consultant coaching them through the first 30 days.
Source: Boston Consulting Group's 10-20-70 principle for AI transformation (BCG AI practice).
About the author. Brandon C. Young is the founder and CEO of Productive Focus LLC, a certified AI consultant with 16 years in IT and B2B who runs three Minnesota businesses and builds the same systems he sells. More about Brandon · Book a free AI audit
