Nobody's Steering, a Compass Left Compass Point

Technology doesn't fail because the software is bad

It fails because nobody is holding the compass.

A software company lives inside its tools, so it staffs people to watch them. A company with real moving parts, trucks, warehouses, crews, production lines, bolts tools onto a physical business. Technology is a supporting actor, so it often gets managed like the phone bill. Then, one reasonable purchase at a time, it becomes the thing your margin and your key people depend on. Nobody decided that, and nobody was assigned to notice.

The damage shows up in four recognizable ways. We map them to the four points of a compass.

North · DirectionNobody owns where technology is taking the business. You are here
East · ExecutionRollouts don't land where you aim.
South · Sinking MoneyMoney leaves quietly, and it never looks like waste.
West · WorkaroundsThe team fights the tools instead of using them.

None of the four is a technology problem. All four are ownership problems, and every one is fixable without buying anything.

Compass Points is a series of four, one for each direction. Part 1, South, followed the money. This part asks where all those purchases were taking you. East and West follow. All four come out of the same instrument, the Compass Score, a twelve-statement self-assessment that takes three minutes and gives you a bearing out of 100.

This Compass Point is about direction, and about taking control of it.

What follows is the pattern, one composite company where it played out, and the four habits that put someone back at the wheel.

What changed

Growth chose your technology

You can date most of your systems by what was happening the year you bought them. Companies your size usually arrive here along one of two roads. Some bought one system to run everything, chosen the year the business finally needed it, and it fit the company perfectly at that size. The business has grown past it since, and every year adds another workaround to keep it useful. Others never bought the one system. They added a piece at a time, each one solving something real on the day it arrived. Most companies are on both roads at once, a core system with a decade of additions built up around it. Different roads, same destination. Whether that is one system or six, it is your ERP, the system or systems that run your business from lead to cash. The technology you run today was chosen by a smaller version of your company, and no one chose the sum.

What changed between those two sizes was not the quality of the decisions. It was who was in them.

You knew every system, because you approved every one.

Systems arrive faster than you can learn their names.

Decisions were slow enough that you were in them.

Decisions move at the speed of the department that needs them.

Growth was the goal, and technology followed it.

Growth is still the goal, and technology is chosen by whoever is chasing it.

Every company with moving parts does three things. It wins the work, it delivers the work, and it settles up for the work. When you were small, all three happened within earshot of you. Scaling separates them, and each stage grows its own staff, its own pressure, and, without anyone deciding it, its own software.

Chasing revenue

Sales wins the work. It buys what helps it quote faster and keep the pipeline visible.

Keeping promises

Operations delivers what sales sold, at a volume it did not set. It buys what helps the same crew do more, and it inherits what a large customer required before awarding the work.

Squaring accounts

Finance collects, pays vendors, and closes the month. It buys what helps it see what already happened.

Each stage bought well for itself. No one bought anything for the space between them, and that space is where your business actually runs.

None of this argues for the old way. When operations picks its own scheduling tool, it usually picks well. The side effect is the problem. Every department gained the power to buy, and no one was given the job of steering the whole.

Both columns are technology. Only one of them has someone accountable for it.

The column without a name is the one that decides where you end up, and it falls to the person who already owns everything else.

Ask your IT person or your provider what the technology plan is beyond the next renewal. If what comes back is a hardware refresh schedule, you have your answer.

The Pattern

No one drew the map

No one decided to stop steering. As the business scaled, each stage solved its own problem faster than anyone could connect them.

What each stage can see

Stage	Sees clearly	Cannot see
Chasing revenue	The pipeline, and what the customer will accept.	What the promise will require downstream.
Keeping promises	This week, the jobs already sold, the crews out today.	Anything past ninety days, or where the company is heading.
Squaring accounts	The WIP schedule, and every invoice already sent.	Whether what sits in work in progress can actually be billed.

Each stage is looking hard at something real. No one is looking at the whole, and the parts they cannot see are the parts that meet.

And between them, the seams

A quote becomes a job. A job becomes an invoice. The place where two stages meet is a seam, and the pass itself is a handoff. Both moments cross a line on your org chart, and at both of them the information is retyped, interpreted, or carried across by someone who knows how it really works. Every tool in the building was bought to work inside a stage. Nothing was bought to cross one. Ask who owns the moment a quote becomes a job, and you will hear about the handoff, the process, the two systems involved. You will not hear a name. Ask the same question in your own language, who owns the WIP schedule, not who prepares it, who owns it, and the room goes quiet the same way.

How you find out

The direction question

Ask your leadership team where technology is taking the business. You may hear five different answers, or silence. Nobody is expecting a forecast. The finding is that no two people are pointed the same way.

The initiative that stalls

The system upgrade, the customer portal, the AI pilot. It starts with energy and then slows, because no one can say how it fits the whole, so every meeting starts the scope argument over.

The budget question

A banker, a board member, or a new CFO asks what the technology budget is. The number that comes back is the provider's invoice plus whatever hardware is due, because the rest is spread across departments buying on their own cards. That is not a budget, it is a receipt.

The roadmap request

A buyer, a lender, or a new advisor asks to see the technology roadmap. There isn't one, and the room goes quiet. What they conclude is that a company this size depends on systems no one is steering.

Every decision behind this was reasonable on the day it was made. Together they add up to another decision, about where the business is headed, and no one made that one. If you bought a single system, you are running a larger company on a choice made when it was smaller. If you bought a piece at a time, you are running it on choices no one has counted. Either way, the technology is set to the size you used to be.

Which is how you became the CTO

Because the stages own their tools and no one owns the seams, every question that crosses a seam arrives at your desk. You approve the purchases, settle the arguments, and sign the renewals, with no time for it and no framework to decide against. No one assigned you the job. It fell to you at the exact moment the company became too large for you to hold in your head. If you are not the owner, you already know whose desk this lands on.

From the corner office

The job nobody could follow

The company is invented. The problems are not.

A $52M commercial mechanical contractor. 100 employees, second generation owner, thirty years in business, install work and a service department, strong reputation, and a book of business that has grown every year for the last seven. The company was doing $11M when it bought its core system, which handles accounting and service dispatch and handles both well. Since then, sales added a CRM, estimating moved into its own tool, the field crews received a mobile app for work orders, because the core system dispatches a job but was never built to travel with the crew, a hospital client required a portal, and finance built the reporting that leadership actually reads in spreadsheets. Nobody would call this company disorganized. Every one of those additions was a good decision.

A chiller replacement at a regional hospital, quoted at $480,000. To win the work, sales agreed to an install window inside the hospital's shutdown weekend, and absorbed two scope items the customer asked for late. Operations held the weekend by pulling two crews off service calls and paying overtime. The late scope items surfaced on site, on a Friday, and the crew did the work, because the alternative is telling a hospital no. That Friday went into work in progress like everything else, and without a change order behind it, it could not be billed.

Then follow the job across the business.

The work sits in WIP the whole way across. Four months later the controller can prove the margin, and cannot say where it went, because no system in the company holds the whole job.

Sales believes it sold a good job. Operations believes it saved one. Finance can only prove how it ended. All three are correct, and no two of them are looking at the same job.

Here is the part that matters more than the job. When the owner asked what happened, the answer from every direction was a request to buy something. Sales wanted the estimating tool connected to the CRM. Operations wanted a better scheduling module. Finance wanted the field app replaced. Three reasonable requests, each one aimed at a single stage, none of them aimed at the seam where the job actually broke. Meanwhile the customer portal the hospital required, live for two years, had been scheduled for expansion to other accounts for eight months, and had not moved, because no one could say what it was supposed to become.

Notice what nobody proposed that week. Every instinct in the building was to buy the missing piece, because buying is fast and deciding is uncomfortable, and because there was no page anywhere in that company for a purchase to be checked against.

What good looks like

Four habits, none of them technical

You do not need new software to fix this. The contractor in the last section had plenty of software, and every request that followed the bad job was a request to buy more. What the company lacked was a person, a page, and one door. Together they move you from reacting to what arrives to choosing what comes next.

Name one owner of the direction

Most companies your size can name a leader for every function. This is the name above all of them, accountable for where technology is taking the business, not for whether it is running today. This person installs nothing. They hold the destination, they say no to purchases that do not serve it, and they can answer the direction question in one sentence. At your size the role rarely justifies a full time executive, and that is not a reason to leave it empty. Fractional, part time, or someone inside the business with a set percentage of their time dedicated to it, the structure matters far less than the name existing. If it is not on their calendar and in their goals, it is not dedicated.
Put the destination on one page

Where the business is heading, stated plainly enough to steer by. Not a forecast, a heading. Written in business language, not system names. How many locations, how much volume, which customers, and what has to be true about the way you quote, deliver, and bill to support that. Every system you own should be answerable to it. If your leadership team cannot agree on that page, technology was never the problem.

A bearing is not a forecast. Next year you can plan. Three years you can only aim.
Agree on how the work actually flows, then name every seam

Put the three stages in one room, sales, operations, finance, and walk one real job from the first call to the cash. Not how the process is supposed to work, how it worked on that job. Most of the time will go to disagreements about what happens after each handoff, and that is the point, because the disagreements are the seams. Then put a name on each crossing, not the department, the person.
One door, and a bearing check twice a year

The room you used the first time does not dissolve when the map is done. It becomes the door. Any new system, any new subscription, and any change to how work moves between the three stages goes to that group before it is bought or switched on, whenever it comes up, not at the end of a quarter. It is a door, not a committee. The group asks one question: does this carry us toward the destination, or does it only solve this week? Solving this week is sometimes the right answer, it simply has to be said out loud. With that door in place, the formal review only has to happen twice a year. Keep it quarterly in the first year, while the habit is setting.

A company that does these four things will still have technology problems. The difference is that it will choose them, instead of inheriting them, and it will hear about them while there is still time to decide.

✦ Where to start, this month

Walk one job

Two hours, three leaders, one wall, from the first call to the cash.
Mark the breaks

Every point where work is retyped, waits on someone, or belongs to no one.
Name the worst three

The three that cost you the most, each with a name and a date.

None of that requires a purchase, a vendor, or a system. It requires a calendar invite.

What to expect from the Compass Score

Someone in our company owns technology outcomes, not just uptime.
Our tech spending plan comes from next year's goals, not last year's budget.

Twelve statements like these, rated 1 to 5. Three minutes gets you a score out of 100 and the direction costing you the most. The link is on the next page.

The Compass Score

Your bearing

Your ratings convert to a 0–100 score, and one benchmark we can defend: a well-run company scores 85 or better. Rating yourself all 4s lands you at 75, still Drifting. The most useful output is not the total, it is your weakest direction. If yours is North, you already know where to start, because the four habits are on the page before this one.

Book your Compass Score debrief

calendly.com/joe-infante/compass-score-debrief

Rather do this with your leadership team in the room? The Navigator Briefing is our live working session, thirty minutes, no pitch. The North session builds the Bearing Page, one page that sets your destination over the map of how your business runs today. Dates at compassleft.com.

About

Compass Left

Compass Left provides outsourced technology leadership to growing companies with physical operations. We don't sell software, resell licenses, or take referral fees, so the only thing we have to gain from your assessment is an accurate one. Compass Points is a series of four. Part 1 was South, Sinking Money.