May 26, 2026

The gold data no one is using...silence

START HERE

The 90-second version

Author’s note

This is a piece about a kind of silence no monitor will ever page you about — the business quietly routing around IT. It is not a throw back to the days when IT owned every server; those days are not coming back, and they were not better. It is not a tooling pitch. It is about reading the silence you already have, before it hardens into a seat at the table you no longer hold.

Your dashboards are green. Your seat at the table is smaller. You are invited to fewer tables.

The business stopped needing IT to own its technology between 2006 and 2010. The migration is complete — by the last full count (Gartner, 2021), 74% of technology purchases were funded at least partially outside IT, and 41% of the workforce did technology work that never reported to IT.
When the systems left, the strategic conversations left with them. IT cheered for the handover at the time. The relief was real. The cost was not visible yet.
The cost is visible now as silence. Business units that file zero tickets are not fine — they have gone quiet for a reason, and the reason has shape.
The shape is green, amber, or red. Your dashboard cannot tell them apart.
- Green — genuinely fine. Watch for the look-alike: a BU that says we’re good because they built their own world and stopped expecting anything from IT. Different conversation, not this mission.
- Amber — running on tools IT is not in the loop on, for an outcome IT could help with. The tell is the pause before they answer.
- Red — They stopped trying. They have outcomes they can’t hit because of IT friction, and they decided the cost of asking is higher than the cost of failing quietly. When you ask what’s hard, they’re not angry. They’re tired.
The first move is three to five 20-minute conversations with BU leads who have gone silent — three at a smaller company, five if you have a sprawling org with many BUs. Not a program. Not a tool. A tracer bullet to find out which kind of silence you are looking at.

⚠ Warning

This does not apply when the service IT runs is the product — a SaaS company where engineering and operations are the business. There, green dashboards and zero business-unit tickets mean roughly the opposite of what this piece describes. This is written for IT inside a company whose product is something other than the technology IT operates.

THE HANDOVER

What IT cheered for while it was happening

Before the silence, the part nobody talks about.

Between 2006 and 2010, three things gave the business an exit it never had before. AWS S3 went live in March 2006. AWS EC2 followed in August. Workday started taking enterprise share. Salesforce was already proving that outcome-shaped SaaS worked. The corporate card became a real procurement channel inside the same window.

2006 → 2010 — The exit opens

The migration was rational. The teams that own the outcomes should own the systems that serve them — Sales should own CRM, Finance should own ERP. One function at a time, the systems moved out of IT:

System	Moved from IT to
CRM	Sales Operations
Marketing technology	The CMO
HRIS (as Workday rolled out)	HR Operations
Analytics	The Chief Data Officer
Procurement technology	A function Procurement built itself

The migration was correct.

The numbers point one direction. Gartner measured this head-on only once, in its 2021 survey. 74% of technology purchases were funded, at least partially, by business units outside IT. 41% of the average enterprise workforce sat in the business technologist category — doing technology work whose reporting line lives outside IT. The trend has not aged out. In the survey Gartner published in October 2024, business and corporate areas outside IT still dedicated, on average, 26% of their staff to building, implementing, or managing technology themselves. Gartner’s later cycles turned to GenAI and never re-counted the split, so these remain its most recent direct measurements of it.

When each category left IT, the response inside IT was relief. The queues were overwhelming. The compliance audits were expensive. The escalations ate half the team’s calendar. When a BU said we will take it from here, the response was thank you. We meant it. We held all-hands meetings about being lean. We retired the support contracts. One less thing to manage. That sentence was said in operating reviews across the industry, for twenty years, with relief on every face.

What we did not say out loud is that every category that left IT was a category of business-outcome conversation IT was now no longer in the room for. Sales took CRM, and the pipeline acceleration conversation went with it. Marketing took the campaign stack, and brand performance went with it. HR took HRIS, and workforce planning went with it. Analytics took data strategy. Procurement took spend. Finance took the close.

By 2026, IT supports the same systems with the same headcount it had in 2010, with one critical difference. Those six conversations — among the most strategic in the company — now happen in rooms IT does not enter. You signed the transfer order yourself. You celebrated it at the time. You were not wrong to be relieved. You were wrong about what the relief meant.

The asymmetry — Same systems, same headcount, six fewer rooms. The cost never showed up on the IT P&L. It showed up on the seat at the table.

THE WALL

The watermelon — green outside, red inside

Your service dashboard is the wrong instrument for the question you are trying to answer.

Outcomes rescued rise fast with the first service improvements. Then the curve bends. Past the bend, tightening service targets is adding decimal places to a test you already passed. You pay real money. You get near-zero movement on the thing the user is upset about.

Where I’d be wrong — Tighten your SLAs, leave the service-only definition of success in place, and watch where users go. If they route to IT more — fewer colleague hand-offs, fewer shadow spreadsheets, fewer quietly abandoned tasks — then the curve above is wrong and so is this piece. I have not seen it happen. If you have, bring the data.

A ticket closes green within all targets. The user could not achieve the outcome. The CSAT can come back 5/5 and the reopen rate zero — a satisfied interaction sitting on top of an unmet outcome, because a CSAT grades the exchange, not the result. The silence becomes proof nothing is wrong. You conclude the business is being unreasonable. They are not. You are judging them on a measurement they could not care less about while they are losing.

This is the watermelon. Green on the outside, where the dashboard lives. Red on the inside, where the trust erodes.

Watermelon — A status report whose surface color contradicts the state underneath. Term of art borrowed from project-management folklore — the dashboard is green, the delivery is red, and the gap between the two is the question worth asking.

Somewhere this week, a person in your business hit a problem and did not report it. They priced IT against a clock — this thing is worth X, losing it costs Y, can IT fix it before Y lands. Sometimes IT wins the bid. When IT loses, the task does not become a ticket. A colleague solves it. A spreadsheet absorbs it. The work gets dropped. Your dashboards never see it. That loss has a name — bleed — and three shapes: it lands late, it lands through a workaround, or it does not land at all.

The dashboard is not lying. It is answering a different question than the one the business is asking.

FIG 1.2 — Two questions, two instruments

IT asks	Business asks
Is the service working?	Is the sale successful?
Did we answer within SLA?	Did we meet the deadline?
Did the queue breach?	Did we hit revenue targets?

IT’s dashboard answers the left column. The business is graded on the right. Both can be green at the same moment.

FIG 1.3 — One incident, scored in both ledgers

One incident. Blocked Tuesday mid-morning; the outcome was due Wednesday by lunch.

IT’s ledger	The business’s ledger
Response: 24 hours	Outcome due: Wednesday, lunch
Resolution: within SLA	Resolved by: a peer, +3 hours
Ticket: closed Wed 15:00	IT’s fix: arrived too late
CSAT: 5/5	In the system: yes
Closed · green	Red — the work got done, but not without bleed

Same incident, two ledgers. Every cell on the left is true and green. The outcome on the right still bled.

This is not a fringe case. Gartner’s 2025 CIO and Technology Executive Survey, released in October 2024, covered more than 3,100 CIOs and technology executives. Only 48% of digital initiatives met or exceeded their business-outcome targets. More than half fell short on the right-hand column while the service metrics on the left stayed green.

The silence in your dashboard is not the absence of a problem. It is the absence of an instrument that can see one.

GO LISTEN

What to do Monday morning

The goal is not a program. It is a conversation that earns the next conversation.

Forget transforming anything. A program has a roadmap, a steering committee, a tool evaluation, an RFP. By the time it ships, motivation is dead and the watermelon rind is thicker.

Your first move is a tracer bullet — three to five conversations. Not with IT people. With business unit leads whose teams have gone silent. The point is not to fix anything. It is to find out what kind of silence you are looking at.

Step 1 — Find the silence

Ticket volume by business unit for the last 90 days is the obvious starting point, and it is correct as far as it goes. It is also one signal among four. Any single signal can mislead. The strongest first-call lists triangulate across two or three.

FIG 2.0 Step 1 — The four signals

Volume anomalies

Engagement decay

Shadow footprint expanding

Narrative absence

Two signals stacking is louder than any one alone. Volume and engagement dropped together — likeliest red. Quiet queue plus growing shadow footprint — likeliest green-disguise. Normal engagement but conspicuous narrative absence — possible federation: they run their own IT off the books, and no one has said so out loud. Pick three to five BUs — three at a smaller company, five if you have a sprawling org. Include one or two wildcards whose silence doesn’t fit a clean pattern; the wildcards are where the model itself gets tested.

Entire business units do not run at zero technology friction. When the ticket count says they do, the friction did not vanish — it moved somewhere your dashboard cannot see.

The silence is data. You don’t yet know what kind.

Step 2 — Go ask

Get 20 minutes on the lead’s calendar. Walk in with no slides, no dashboards, no proposal. Do not announce a program. That is the kind of move that earned the silence.

The conversation has shape — an opener, four probes, a close. The shapes are below as ideas, not scripts. The actual language is yours.

This is not a blame instrument. You are not collecting grievances or handing them out — nobody is on the hook for what they don’t control. You are there to find the bleed, so the next conversation can be about closing it.

Open with a cost paid up front. Three shapes work. Use only the ones that map to what you actually have. The point is asymmetric — pay something only you can pay before the BU has spoken.

FIG 2.0b Open with a cost — pick one shape

The confession

What

A specific failure named out loud from IT’s own logs, tied to a moment in the BU’s recent past.

Use when There is a real dropped ball in the data.

Say

I pulled our ticket aging this morning. During your platform migration last quarter, four of your requests sat past ten days before anyone touched them. We never told you we’d dropped them.

Your access requests during the reorg averaged nine business days. The SLA is two. I didn’t know that number until last week, which is its own problem.

Trap

Softening with context — “of course the team was juggling…” — retracts the cost just paid.

The found-money gift

What

IT-side telemetry the BU can’t see on their own, translated into something useful to them, handed over with no agenda attached.

Use when There’s IT-only visibility worth giving away.

Say

Twenty-two of your thirty-eight licenses on the analytics platform haven’t been touched in 90 days. Roughly forty thousand a year if you want to reclaim it. Your budget, your call.

Your lifecycle email pipeline is calling our directory forty thousand times a month. Not a governance thing. If you’re sized to it, ignore me; if not, that’s a number worth knowing.

Trap

Turning the gift into the pitch — “…and so we should consolidate this” — undoes the move in one breath.

The admission of ignorance

What

A self-aware admission that IT has not done its homework on the BU’s actual business.

Use when

The relationship is cold — no recent failure, no useful intel.

Say

I’m going to embarrass myself — I don’t know what’s going to keep someone on your team up at night in October. Not the OKR-slide version. I should know that and I don’t.

I couldn’t tell you the last three pieces of work that mattered most to your team last quarter. I could tell you our ticket queue, but that’s clearly not the same question.

Trap

Listing what you do know to cushion the admission empties it.

Of those, pick the one with the highest cost to you. The opener is a trust deposit and only counts if the depositor felt it. Then a flat follow-up that hands the floor over:

No slides, no program. A few questions and one thing I’m hoping to walk away with.

Then probe. Four shapes, asked in this order. Each surfaces a different kind of silence.

FIG 2.1 The conversation — four probes, then close

PROBE 1 The last time they didn't ask

PROBE 2 The vocabulary mirror

PROBE 3 The Chesterton invitation

PROBE 4 Beside the point

CLOSE One commitment, on a date

FIG 2.1b — The 20-minute shape

#	Move	Time	Job
1	Opener (A, B, or C) + intent line	60–90s	Pay the upfront cost; flip the dynamic
2	Probe 1 — last time they didn’t	4–5 min	Surface red as story
3	Probe 2 — vocabulary mirror	2–3 min	Read the ambient register
4	Probe 3 — Chesterton invitation	3–4 min	Surface walls and waste
5	Probe 4 — beside the point	3 min	Map federation
6	Closing commitment	60s	Replace warmth with a calendar item

Six moves, ordered. Time budgets are floors, not ceilings — if a probe is landing, stay in it.

Through all of it — listen more than you ask. Take notes. Don’t defend. Don’t propose a fix. Don’t explain the support model. You are not there to solve. You are there to learn what the dashboard can’t tell you.

Step 3 — Sort the silence

After your three to five conversations, you’ll be able to sort each unit into one of three buckets.

FIG 2.2 Step 3 — Sort the silence

Green silence

The look-alike

Amber silence

Red silence

* * *

Step 4 — Believe what came back

THE BET PENDING

The piece rests on this — the silence in your dashboard has shape. Some of it is green, some amber, some red. If three to five BU leads, picked from the silent column, all sit down with you and say honestly, things are fine, the claim does not survive contact. Run the tracer to find out which kind of silence yours is.

OPEN

If all of them land in genuine green — specific about what works, specific about what they would lose — your dashboard was telling the truth for those units. A handful of units is not the whole organization, so repeat the tracer somewhere else before you generalize. If any unit lands in green disguise, amber, or red, the silence has shape. The model holds. The silence is data, and the dashboard was missing it.

The temptation now is to find a comfortable story — that unit is an outlier, those workarounds are user preference, the spreadsheet is fine. Resist. The whole reason you ran the tracer was to find out which story is true. Believe what came back.

Belief has to cut both ways, or the model is rigged. A lead who names two or three specific things that work, and what they would lose if those stopped, has passed — write green and move on. Re-reading that as a disguise is the same comfortable story running in reverse. And if you find yourself recoding most of your conversations as green-disguise, stop: the instrument under suspicion is the model, not the business.

After three to five conversations you can walk into the next leadership meeting and say:

Of our [N] silent business units, [a] are genuinely fine, [b] have built their own world and we should leave them to it, [c] have workarounds we can fold back in or unblock, and [d] have stopped trying. Here’s what I learned. Here’s what I’m doing about [c] and [d].

That statement earns you the next meeting. The strategic planning one. The budget one. Not a program.

A seat at more tables, earned by walking in with a map of silences instead of a sea of green dashboards.

WORKSHOP

The next step

Three to five conversations will tell you which kind of silence you are looking at. They will not tell you what to do once you have sorted it.

What you are building toward has a name, even if you do not need it to make the first call: an experience-level agreement — an XLA. An SLA grades whether the service met its target. An XLA grades whether the person met the outcome the work existed for. Not a product you buy; a second number you agree to be graded on, set next to the green one.

Building that number is the workshop. We teach the instrument that reads the silence once you have found it — how IT shows up at a table it does not own, what you say at the operating review when the win belongs to the BU, and what kills the practice at each stage of maturity. It runs as a two-day intensive built around your org, not around a stack or a vendor.

If the tracer bullet tells you the silence has shape, come. If you ran the tracer and everyone said fine, save your money — you do not have the problem this workshop solves.

Continue

If this changed how you see the problem

The diagnosis names the gap. These workshops are where you close it — start with the one built for the call you actually make.

Start here — The Experience Outcome Layer1 day live / 3 self-paced segments · Orchestrator

Or take it from another angle