The first engagement is three weeks. It ends with a written readout and two scoped builds, the highest-leverage automation candidates for your business, each priced as a standalone project. It is the guided way to start, and the discovery itself does not put you on a retainer.
What the three weeks cover
Week one maps how the operation actually runs, not the org chart, but the workflows: where work waits, who owns each step, what it costs when the step is slow. Week two scores the automation candidates that come out of that mapping. Week three is the readout: a written document and a live session that ends with two builds scoped and ready to decide on.
Most services businesses come in with a vague sense that something in quoting, operations, or outreach is slower than it should be. Three weeks is enough to name the specific workflows, rank them, and know what fixing them would actually require.
Week one: map the operation
We start with workflows, not tools. Every services business has three or four places where work regularly stacks up. Quotes wait on the owner and hours wait on the coordinator, while a follow-up sits untouched until someone digs up last month's email thread.
Week one is a structured discovery across those areas: a workflow map, the time-cost of each delay, a data audit (what exists, where it lives, how clean it is), and a first cut of automation candidates. We work from whatever you already have: your CRM, your spreadsheets, your timesheet system. We do not ask you to buy or switch anything before the readout.
By the end of week one, the map is on paper and the candidate list is in front of us.
Week two: score and prioritize
A typical discovery surfaces 20 to 40 candidate ideas. Most are not worth building. Either the data is not there, the workflow does not have a clear owner, or the frequency is too low to justify the maintenance cost.
We score each candidate on six criteria: frequency, commercial impact (revenue or gross margin touched), data readiness, workflow ownership, risk, and time to a first useful version. Frequency and data readiness together eliminate most of the list.
The two winners are rarely the most exciting ideas. They are the ones that will actually run and that someone in the business owns. Each one can change a real operating number inside a quarter. A rep can quote in the field. A coordinator reviews exceptions instead of rebuilding the record, and a BD lead opens a draft instead of starting from a blank document.
Week three: readout and scoped builds
The readout is a written document and a live session. It covers what we heard in the discovery, the market context that makes the opportunity real, the full prioritized use-case list, and two builds scoped in detail: workflow description, success metric, data requirements, integration points, and time estimate.
The scoped builds are not estimates on a napkin. Each one specifies what changes in the workflow, what the operator sees, what the system does, and how you measure whether it worked. The written readout travels. We have had founders share it with their operations manager and their board in the same week.
You leave the readout with a clear decision: start one or both, defer, or stop. Nothing requires a next step. The scoped builds are priced individually after the readout. You are not locked into anything.
What comes after the readout
If you start a build from the discovery, that project is scoped and contracted on its own. You decide which builds to start and in what order, one project at a time. Colby hosts and maintains what it builds for you.
The handoff includes the workflow map, owner list, admin guide, exception-handling rules, and the specific metrics that tell you whether the system is working. Before handoff, the internal owner runs the workflow without us. If they cannot recover from a bad input or a missing integration, the build is not done.
The goal is a system the team trusts and can operate without anyone from Colby in the room.
Who this is for
Services businesses with $1M to $50M in annual revenue and at least one operational bottleneck that recurs. The primary vertical is commercial janitorial and building services. Staffing, field services, and facilities management are the next tier. The workflow problems are similar enough that the same discovery process applies.
The engagement works when the bottleneck is real and the team is willing to be honest about how work actually runs, not how the process diagram says it should run. It does not work when the goal is a demo for an investor deck or a vendor evaluation for a platform that is already selected.
US businesses only. The bottleneck should be costing you something measurable: time, margin, deals that do not close, invoices that go out wrong.
Common questions
Do I have to replace the software I already use?
No. Most of what you run works fine, and the readout usually recommends keeping it. Colby builds the one or two workflows your current tools do not cover and leaves the rest alone.
What do I actually get at the end of the three weeks?
A written readout that scores your bottlenecks and names the specific builds worth doing first, with the reasoning behind each. You decide what to build from there, one project at a time.
Who runs the software after you build it?
Colby hosts and maintains what it builds, under a separate agreement. You do not need an IT team or a developer on staff to keep it running.
We are not a technical company. Is that a problem?
No. The point is software shaped around how your team already works, in plain language, on a phone if that is where the work happens. The discovery interview is a conversation about how the operation runs.
How do we know which problem to fix first?
That is what week two does. Every bottleneck gets scored on impact and effort, so the first build is the one with the most return for the least lift.
More on choosing what to build: when off-the-shelf software stops fitting, and the use cases that show what these builds look like in practice.