Six months ago you set technology goals for the year. If you're honest, you have not looked at that list since. A mid year technology review fixes that in two hours, before the second half runs on momentum instead of decisions.
This is not a strategy offsite. It is a working session with a clear output: a short kill list, a short keep list, and one initiative you commit to doubling down on for the rest of the year. Most owners skip this because it feels like admin. It is actually the cheapest risk-reduction exercise available to you all year, because subscriptions and half-finished projects bleed money quietly, month after month, without ever showing up as a single alarming number.
I run a version of this with clients every July. It consistently surfaces the same three things: software nobody opens anymore, a project with no clear owner that everyone assumes someone else is driving, and one initiative from January that turned out to matter more than expected and deserves more investment, not less.
The two-hour agenda
Block two hours, no more. If it takes longer, you're debating instead of deciding, which defeats the point of a mid year technology review.
- Pull the list (20 minutes). Every software subscription, every active project, every January technology goal, in one document. Include costs where you have them.
- Score usage, not intent (30 minutes). For each subscription, ask: who logged in this month? Not "who is supposed to use it," who actually did.
- Score projects by owner, not by plan (30 minutes). For each active project, ask: who owns this, and when did they last report on it? No named owner is itself a finding.
- Build the kill list (20 minutes). Anything with no usage and no owner goes here. Cancel or shelve without ceremony.
- Pick the one to double down on (20 minutes). Of everything still standing, which single initiative, if you gave it more budget or attention right now, would move the business most in the next six months?
That's the whole agenda. The output should fit on one page.
The kill list: what actually qualifies
Owners hesitate to kill things because cancelling feels like admitting a January decision was wrong. It usually wasn't wrong, it was reasonable given what you knew then. Killing it now is not an admission of failure, it's normal portfolio management. Three categories almost always qualify:
- Subscriptions nobody opens. If login data shows zero or near-zero usage for two consecutive months, the tool has already failed silently. Cancel it before the renewal, not after.
- Projects with no owner. A project everyone agrees is "important" but nobody can name as their responsibility is not a project, it's a line item slowly consuming budget with no one accountable for finishing it.
- Initiatives solving a problem that already changed. Sometimes the January plan targeted a problem that's since been solved another way, or deprioritized by a bigger shift in the business. Don't keep funding a solution to a problem you no longer have.
If any of this points to a deeper pattern rather than a one-off, it's worth reading Technical Debt Explained for Business Owners, since unkilled software projects are one of the most common ways technical debt accumulates without anyone noticing.
The one thing worth doubling down on
Every review surfaces a kill list easily. The harder, more valuable part is naming the single initiative worth more investment. Owners tend to spread attention evenly across everything still standing, which guarantees nothing gets the push it needs to actually finish.
Ask this question directly in the room: "If we could only fund one more thing properly for the rest of this year, what would it be?" Force a single answer, not a top three. The discipline of choosing one is what makes the exercise useful. This is the same logic behind Set 3 Technology Goals for the New Year (Not 10): fewer, better-resourced bets beat a long list of half-funded ones.
A simple scoring table to bring into the room
| Item | Monthly cost | Last real usage | Named owner | Verdict |
|---|---|---|---|---|
| CRM add-on | Rp 1.2M | 3 months ago | None | Kill |
| Inventory dashboard project | Rp 8M/mo dev time | Weekly | Ops manager | Keep, evaluate for double-down |
| Old scheduling tool | Rp 450K | Never logged in since Feb | None | Kill |
| Customer chatbot pilot | Rp 3M/mo | Daily, growing | Product lead | Double down candidate |
Fill this in honestly with your own numbers. The pattern that emerges is usually obvious once it's written down. It rarely is obvious while it's still living in six different people's heads.
The takeaway
A mid year technology review costs two hours and saves months of quiet waste. Pull everything into one list, score by actual usage and named ownership rather than intent, kill what's dead without ceremony, and pick exactly one initiative to fund harder for the second half. Do this every July and January stops being the only month you make deliberate technology decisions. If you want an outside eye on the list before you finalize the kill decisions, that's a conversation I'm glad to have as a technology partner.