SOXOA
Book a callGet free assessmentFree assessment

Navigate Soxoa

Services, methods, research, and the free assessment.

The symptom

Nobody can tell me the status of a job without asking three people

Triage table included4 cited sourcesLast reviewed September 1, 2026

The short answer

Status is a byproduct that nobody in your business is responsible for producing. The scheduler knows when it was booked, the crew knows what happened on site, the office knows what was invoiced, and no system holds all three. So the only way to know where a job stands is to reassemble it from three people, which is exactly what you are doing.

The fix is not a dashboard. It is agreeing a small set of states a job can be in, and making each state change get recorded as a side effect of work that is already happening. Scheduled when it lands on the calendar. On site when arrival is logged. Complete when the completion note is filed. Invoiced when the invoice is raised. Four states, each set by something someone already does, and the status question answers itself.

The parts to be careful with are the human ones. Assembling today's status from existing systems is mechanical and should be automated. Writing the customer update is drafting, and a person should approve it. Deciding whether to tell a customer their job will be late is judgment about a relationship, and it should never be automated at all.

Why dashboards usually fail here

A dashboard displays state. If nothing is recording state, the dashboard is an empty frame, so the business responds by asking people to update it — which is new manual work layered on top of the old manual work. Within two months it is stale, and everyone reverts to phoning the three people, now with the added cost of a subscription.

The order that works is the reverse: instrument the state changes first, using events that already occur, and only then display them. If a state change requires someone to remember to record it, it will not survive a busy week.

The status question is usually a customer question

In most businesses with this symptom, the three phone calls are triggered by a customer asking. That is worth noticing, because it means the internal visibility problem and the customer communication problem are the same problem, and fixing one fixes the other.

It also suggests the cheapest first move: a proactive weekly update to customers with active jobs. Assembled automatically, reviewed by a person, sent on a schedule. It removes a large share of the inbound status calls and it is a far smaller build than anything involving a live portal.


Triage table — yours to use, nothing to buy

Triage: what to automate, assist, and leave alone

Six tasks that make up status reporting, scored with prioritize_automation_workflows, Soxoa's public scoring tool, on inputs describing weekly hours, predictability, judgment, error cost, data sensitivity and systems touched. Nothing here scored above 70, which is itself the finding: status work is assembly plus judgment, and the judgment half is larger than people expect.

Automate the assembly

Mechanical collection with low judgment. This is where the hours are.

Collecting today's status from field and office systemsscore 66

The top-ranked item and the recommended first workflow. Five hours a week across three systems, and no judgment involved.

Flagging jobs with no activity for five daysscore 62

A single rule over data you already hold. Often the highest value per hour of build time on the whole list.

Assist — assemble it, let a person send it

Predictable inputs, but it goes to a customer, so someone reads it first.

Building the weekly customer status updatescore 55, human review required

The draft is assembly; the decision about what to say when a job has slipped is not.

Leave alone for now

Medium-to-low predictability or genuine judgment. Investigate, do not build.

Answering ad-hoc where-is-my-job callsscore 43

Fix the cause instead — most of these calls disappear once the proactive update above exists.

Reconciling the schedule against what actually happenedscore 39

Four systems and real ambiguity. Worth doing eventually, not first.

Deciding whether to tell a customer a job will be latescore 7

The lowest score on the list. This is a relationship decision and it stays with a person.

These scores are a discovery aid, not a promise of savings or feasibility. The tool returns instructions to confirm data access, exception paths, security requirements and ownership before building, and to pilot on a reversible workflow with a measurable baseline. Your weekly hours and judgment ratings will differ.

Terminology bridge

What consultants call this

The words below will get you to the right software category and away from the wrong one.

Single source of truth

One place where a job's state lives. What you are missing, and what a dashboard is not.

Status model / workflow states

The agreed short list of states a job can be in. Four to six is right; twelve is a project nobody finishes.

Event-driven updates

State changes triggered by things that already happen, rather than by someone remembering to update a field.

Stale-job alert

The no-activity-for-N-days rule. Cheap, and usually the first thing worth building.

Proactive status communication

Telling customers before they ask. The measurable outcome is fewer inbound calls.

If a vendor leads with the dashboard rather than with how state gets recorded, ask them the second question first.

Common questions

Fair questions

Do we need one system for everything?

No, and consolidating everything is usually the more expensive answer. What you need is one agreed status model and a way to read state changes out of the systems you already run. That is a much smaller piece of work than a platform migration.

Can customers see the status themselves?

They can, eventually, and a portal is a reasonable second phase. It is a poor first phase, because a customer-facing view of unreliable data damages trust faster than not having one. Get the states recorded and trustworthy first.

How many states should a job have?

Four to six. Every additional state is another thing someone has to be right about, and the marginal value drops quickly. If a proposed state does not change what anyone does, it does not need to exist.

Sources

Where this comes from

  1. CRM API: deals

    HubSpot developer documentation

    Example of a CRM exposing pipeline stage and last-activity date as readable fields — the raw material for a stale-deal alert.

  2. Overview of cloud flows

    Microsoft Power Automate documentation

    Reference for what an off-the-shelf flow tool can and cannot do on its own.

  3. AI Risk Management Framework (AI RMF 1.0)

    National Institute of Standards and Technology

    The public reference for deciding where a human checkpoint belongs in an automated process.

  4. Annual Business Survey

    U.S. Census Bureau

    Free public data on business technology use and characteristics, if you want a baseline outside vendor marketing.

Last reviewed: September 1, 2026 · every source link checked on that date

Where to go next

Three useful next reads

Free · No strings

See where your job status is actually going missing

Share your company, website, and email. We research your operations, show the sources and assumptions, and send a private assessment link when it is ready — no call required to read it.

Draft protection

Draft saving starts when you type; submitting the request remains available even if draft storage is unavailable.
Add optional context files

Only attach material you are authorized to share. We use it to prepare your assessment; see our privacy policy.

We email you when your page is ready. Usually within 24 hours. No spam. No card. No call required.

Prefer to talk first?Book 30 minutes(916) 857-9148

Other things owners tell us

More operational symptoms

See all ten →