The Q4 Project Triage: Which Tech Projects Finish the Year

Late September has a familiar inventory: the website refresh at eighty percent, the phone system migration stalled since July, three integrations someone promised a customer, the security tooling rollout, and a half-dozen smaller items that each seemed small in March. Everyone asks the same question with the same desperation: what actually finishes this year?

The wrong answer is 'everything,' because that answer means the year ends with twelve projects at ninety percent and nothing shipped. The right answer comes from a triage that most companies skip because it feels like admitting defeat. It is the opposite: it is how anything finishes at all.

The triage takes one meeting, run well. Here is the meeting.

Key takeaways

• Twelve projects at ninety percent is worth less than five at one hundred.

• Triage by value at completion, not effort already spent - sunk cost lies.

• Finish, freeze, or kill: 'keep slowly working' is not a real category.

• Every surviving project gets a named owner and a December definition of done.

• Communicate the killed list kindly; someone loved each one.

Step 1: Build the honest inventory

List every technology project that touched the year, with three facts each: percent done as the owner honestly estimates it, what remains to ship, and what shipping is worth. The middle column is where the truth lives; 'almost done' has ended more Q4 plans than any budget cut.

Include the invisible projects: the security remediation from the last assessment, the access cleanup, the documentation debt. These compete for the same hours as the visible ones, and leaving them off the list does not exempt them from the year's capacity.

The inventory conversation usually produces the first insight unprompted: there are more projects than anyone thought, and several are the same project wearing different names.

Step 2: Score by value at completion

The trap in triage is honoring sunk cost: the project that consumed the spring feels entitled to the fall. Score instead on what shipping by December 31 is worth - revenue enabled, risk retired, customer promise kept - and what waiting until Q1 would cost.

This fictional example shows one company's scoring of its top six. It is an illustration, not a prescription.

Project	Value if shipped	Cost of slipping	Verdict
Customer portal phase 1	Two renewals depend on it	High	Finish
Phone system migration	Modest savings	Low	Freeze for Q1
Security tooling rollout	Closes audit finding	High	Finish
Website refresh	Nice, not needed	Low	Freeze
ERP reporting module	Team wants it	Medium	Kill - revisit in spring planning

Two finishes, two freezes, one kill. That ratio is typical, and it is what focus looks like in spreadsheet form.

Step 3: Kill cleanly, freeze kindly

The categories matter. Finished means shipped and adopted, not demonstrated. Frozen means a deliberate, dated pause with a named restart condition. Killed means over, with whatever was learned written down in a paragraph so the idea does not zombie back next year.

The category to ban is the fourth one: the project that keeps ten percent of somebody's week forever. Slow-working projects are how portfolios rot; they never finish and never free their hours.

The same logic governs security tooling decisions, and the consolidation question applies here too: sometimes the right triage answer is that two half-projects merge into one real one.

Step 4: Protect the finishes

The triage only works if the survivors get defended. Each finishing project gets a named owner, a December definition of done, and explicit protection from the year's remaining distractions: the conference season requests, the budget fire drills, the executive idea of the week.

Publish the triage list where the leadership team sees it weekly. When a new demand arrives in November, the list is the answer: which of the finishes does this displace?

The year-end review lands differently after a real triage; the discipline in the six-slide year-end review is far easier when the year actually finished something.

The related discussion of consolidation question is worth your time alongside this one. The related discussion of six-slide year-end review is worth your time alongside this one.

Frequently asked questions

How many projects should survive the triage?

Fewer than feels comfortable. For most small and mid-size companies, three to five genuine finishes is a strong Q4. If everything survives, the triage did not happen.

Who decides - IT or the business?

The business, with IT's honest capacity numbers on the table. A triage run by the technology team alone optimizes for technical interest; run by the leadership team, it optimizes for the year.

What about the security work that keeps getting triaged out?

If security items freeze three years running, that is a signal, not a coincidence. Give the risk-retiring projects explicit weight in the scoring, or they will always lose to projects with louder sponsors.

A Defensible Next Step

This week, list everything in flight and ask one question of each: what is shipping by December 31 worth? The answers will sort the list faster than any framework. Book a Decision-Clarity Call if you want the triage facilitated.