Can Brownsmith Help? Scope the Problem Before the Software

Start With the Work

A useful first question is not “Where can we use AI?” It is “Which job keeps taking more time than it should, and what makes it hard?” Choose one route through the business: an enquiry becoming a quote, a new client being onboarded, a weekly report being assembled, or a customer question being answered. A narrow example gives everyone something concrete to inspect.

Describe what starts the work, the steps people take, the information they need, and where it waits or gets repeated. Include exceptions and the decisions that should remain with a person. Do not begin with an ideal future process. The awkward steps and workarounds are often where the explanation lives.

For example, a hypothetical firm might say: “When a quote request arrives by email, we copy details into a spreadsheet, look for a similar past job, ask an engineer about exceptions, then prepare a reply. Requests sometimes sit when nobody knows who owns the next step.” That is enough to start discussing the process. It is not evidence of a measured saving or a promise that the same answer fits every firm.

Make the Decision Legible

Brownsmith’s Advisor is an MCP server that exposes six read-only tools to a connected AI client: diagnose_business, analyse_workflow, audit_customer_acquisition, create_implementation_brief, find_brownsmith_service, and get_brownsmith_info. The server supplies Brownsmith’s structured frameworks and results; the connected model interprets the business description. It does not run a second language model of its own.

The workflow analysis is designed to consider what can be removed, what can be automated with predictable rules, where AI may help with interpretation or drafting, and which decisions need human judgement. Customer acquisition can be considered through Signal, Place, Presence, Action, and Learning (SPPAL). A brief can then organise a small phased next step once the problem is understood. A service match checks the described need against Brownsmith’s configured service information; it is not a verified quotation or commitment to deliver.

A practical prompt might be: “Here is our current quote-request process, including the steps, delays, exceptions, tools, and decisions that need an engineer. What can we remove first? What should stay human? Where might simple rules or AI help, and what information is missing?” Use a made-up or suitably anonymised example while exploring. The supplied server does not store diagnostic histories, but the connected client and hosting infrastructure may process or log information separately. Keep private business information out of an exploratory example.

What the Preview Can and Cannot Establish

The tools work from the information supplied in the conversation. They do not visit a website, fetch an analytics account, check a CRM, or verify a claim because a URL was pasted. If a page matters, provide the relevant text as context and make clear what is observed, assumed, or still unknown. The output is structured input for a model to interpret, not independently verified business facts or a quantified return on investment.

That limit can be useful when treated honestly. An advisor should help expose unanswered questions rather than paper over them. Ask which figures would need to be measured, who owns a decision, what an exception looks like, and how a small trial could show whether the change helped. Remove a needless step before automating it; use AI when its ability to interpret or generate adds something specific. Keep consequential decisions with the people accountable for them.

The Business Advisor is being presented as an Advisor preview at mcp.brownsmithdynamics.com. Access may be required, and public tool availability has not yet been confirmed. This is a first version built around structured business questions. Its purpose is to make an initial conversation more useful, with the delivery scope confirmed directly afterwards.

From Analysis to Action

Sometimes the answer is a clearer ownership rule or a simpler form. Sometimes it is ordinary automation, a focused AI assistant, a system integration, or custom software. Sometimes the business has not yet agreed on the process well enough to build anything. The Advisor can help frame that choice; it cannot inspect the real operation or make the decision on the owner’s behalf.

Brownsmith’s approach is to reduce busy work and introduce technology only where it improves the work. That means a possible service fit is a prompt for a conversation, not a reason to force the business into a predetermined solution. Our business efficiency service and Why Brownsmith Dynamics page explain the work we take on. If the configured service information does not fit, the right answer is to say so or ask for more detail.

Before a project grows, agree on one observable sign of improvement: fewer handoffs, less time spent locating information, fewer missed follow-ups, or a response that reaches the customer sooner. Record the current state if you can. Run a small change, review exceptions with the people doing the work, and expand only if the evidence supports it. No invented percentage is needed to make the next decision clearer.