It is June. Whatever technology plans you made in January are now six months old, half-executed, and quietly diverging from reality. This is exactly the right moment for a digital strategy review, because you still have six months to fix what December will only let you regret.
Most businesses skip the mid-year check and run the post-mortem instead. In December, someone tallies up what was spent on the new website, the CRM subscription, the inventory tool, the social media agency, and asks what it all achieved. By then the money is gone and the answer is a shrug. A digital strategy review in June asks the same question while the budget, and the year, can still be redirected.
You do not need a consultant or a two-day offsite. You need a spreadsheet, your numbers, and about two hours of honesty. Here is the structure I use.
The one rule: every initiative must name its number
Before the five questions, establish the rule that makes them work. Every technology initiative, every tool, subscription, project, and retainer, must be attached to exactly one business number it was supposed to move. Revenue, leads, conversion rate, order processing time, no-show rate, stock accuracy, support response time. A real number that existed before the initiative and exists now.
"It makes us look professional" is not a number. "The team likes it" is not a number. Those may be true and even valuable, but a strategy is a set of bets, and bets need scoreboards.
If an initiative cannot name its number even in hindsight, that itself is a finding: you bought a solution without defining a problem. It happens to everyone. The point of the review is to stop paying for it silently.
Question 1: What did we actually spend?
List everything with a technology line attached, for the first half of the year:
- Software subscriptions (the small monthly ones add up: Rp 150 thousand here, Rp 500 thousand there, often 10 to 20 of them)
- Project work: website, app, system builds, integrations
- Retainers: maintenance, agencies, freelancers
- Infrastructure: hosting, domains, devices bought for tools
For a typical Indonesian SME this lands somewhere between Rp 30 million and Rp 300 million for a half year, and the owner is almost always surprised by the total. Subscriptions especially hide well. Pull the actual bank and card statements, not your memory.
Question 2: Which numbers moved?
Now put the initiatives next to their numbers, January versus now.
| Initiative | H1 cost | Number it should move | Jan | Jun | Verdict |
|---|---|---|---|---|---|
| New website | Rp 45M | Inbound inquiries/month | 12 | 31 | Working |
| CRM subscription | Rp 9M | Follow-up rate on leads | unknown | unknown | Nobody uses it |
| Social media agency | Rp 60M | Leads from social | 5 | 7 | Under review |
| Inventory tool | Rp 18M | Stock discrepancy rate | 8% | 2% | Working |
Be strict about attribution but not paranoid. If inquiries doubled and the only change was the website, credit the website. The goal is decision-grade clarity, not academic proof.
Three verdicts are allowed: working, not working, and not measured. That third category is usually the biggest, and it is the most fixable, which brings us to the next question.
Question 3: What are we paying for but not using?
Every business accumulates zombie technology: the CRM with three logins nobody used since March, the project tool that lost to a WhatsApp group, the premium plan bought for a feature no one turned on.
Walk the subscription list and ask, for each item: who used this last week? Not "who is supposed to use it," who actually did. If the answer is nobody, it goes on one of two lists:
- Cut list. Cancel it this month. Zombie tools are pure cost, and worse, they let you believe a problem is handled when it is not.
- Fix-by list. If the tool addresses a real problem and failed for fixable reasons, usually training, ownership, or a missing integration, give it a named owner and a deadline: "Rina owns CRM adoption, we review usage on August 1, if it is still dead we cancel." A fix-by date without a name attached is a wish.
In my experience the cut list alone recovers 10 to 20 percent of the software budget, which conveniently funds whatever the review says you should double down on. Tool failure, by the way, is rarely about the tool. It is almost always adoption and ownership, a pattern I unpack in Technology Is a People Problem Wearing a Software Mask.
Question 4: What changed that the January plan did not know about?
A strategy written in January was written for January's world. Six months in, check the assumptions:
- Did a competitor change the game, new app, new pricing, new channel?
- Did your customers shift behavior, more WhatsApp orders, less foot traffic, different payment preferences?
- Did the business itself change, new product line, new branch, key staff departure?
- Did anything regulatory or platform-level move under you?
The question is not "should we panic," it is "does any H1 assumption no longer hold?" If the plan assumed offline events would drive leads and events still have not recovered, the budget attached to that assumption needs a new home, deliberately, not by drift.
Question 5: What are the one or two bets for H2?
A review that produces a twelve-item action plan produces nothing. Force-rank instead. From everything above, pick:
- One thing to double down on. The initiative with the clearest number movement gets more fuel: more budget, more of your attention, faster iteration.
- One thing to kill or fix. The biggest item on the not-working or not-measured list gets a decision, this month, with a name and a date.
Everything else continues on autopilot or gets parked. Two deliberate moves executed beat ten intentions logged. If you want a tighter, faster version of this exercise to run quarterly rather than twice a year, I outlined one in Q1 Is Over: Audit Your Business Technology in One Hour.
The takeaway
A mid-year digital strategy review is two hours of work with one rule and five questions: total the spend, match every initiative to the number it moved, cancel or fix the zombies, re-test January's assumptions, and choose two bets for the second half. If a tool or project cannot name the number it moved, it goes on the cut list or gets a named owner and a fix-by date. No exceptions, including the projects you personally championed.
December's post-mortem tells you what happened. June's review still gets a vote in what happens. Take the vote.