Enterprise AI advice can arrive with an invisible team attached: technology specialists, an operations manager and spare capacity for implementation. A small service business may have none of them.
I run SolvStream with a small team and a lean operating setup. If you are a solo management consultant or a service business owner with a small team, advice written for a larger organisation may solve a problem you do not have with resources you do not possess. Clients work directly with me, Shaun, on delivery.
What the small-business source says
According to the U.S. Small Business Administration, 82% of the smallest businesses (under 5 employees) say AI "isn't applicable" to them. That's not a technology problem. That's an advice problem. The guidance they're hearing was never written for how they actually work.
What to do at small scale
One process at a time
Choose one process with recorded friction. Proposals, follow-up and onboarding are possible candidates, not the answer for every consultant.
Structure before AI
Map the current steps, inputs, decisions and handovers. Separate reusable work from the parts that need human judgement. Add AI only where the output can be checked.
The smallest suitable tool set
Use the tools the job needs and the business can maintain. Check the subscriptions, feature limits and operating costs of the stack you choose, even when the entry price is zero.
A measured target
Define the result before the build starts. A universal split, typical time cost or same-day outcome tells you nothing about this business. The before-state and after-state decide whether the change helped.
How do you know which process to fix first?
Start with work that repeats and leaves visible friction. Look through recent examples for searching, rebuilding, delayed handovers or repeated decisions. Choose from that record, not from a table of invented typical times.
Possible categories include proposal creation, lead follow-up, onboarding and reporting. They are examples, not a ranked list.
What about AI agents?
Judge any agent or automation against the process it is meant to run. Define the inputs, permissions, decision limits, checks and fallback before it touches live work.
Do not promise what future systems will run on or how today's build will position the business later.
Common questions
Do I need to understand AI to benefit from it?
You do not need to become a developer. You do need to explain the process, identify the judgement points and recognise a correct result. SolvStream clients work directly with Shaun on the technical and delivery decisions.
What if I have already tried ChatGPT and it did not help?
Inspect the input, the task and the standard used to judge the answer. A weak result can have more than one cause, so test before blaming the tool or the input.
How long does it take to see results?
There is no general result timeframe. The free AI Workflow Map helps name the first process. AI Workflow Architecture is a plan delivered within 72 hours of receiving full context and costs €450. A One Week Ops Reset rebuilds one agreed priority process in a week, from €2,500, with a typical scope of €2,500 to €4,500. The €450 Architecture fee is credited in full towards an Ops Reset booked within 30 days.
Advice should fit the business
Start with one process, keep judgement visible and use only the result your own record supports.
