Re-plan from today

A reforecast is a new forecast made from where the project stands today. It replaces the question "how far are we from the plan?" with a more useful one: "from here, when will it be done?"

The recipe has two inputs: the work actually left and the pace the team has really shown. Divide the days spent by the units finished to get the real pace, then multiply by what remains.

At Klasna the plan gave each practice topic two working days: 12 topics in 24 days. On 7 January, 21 days in, 7 topics are done. The plan says three days are left, but that is the plan's arithmetic, not the work's. The real pace is 21 ÷ 7 = 3 days a topic; 12 − 7 = 5 topics are left; so the work needs 5 × 3 = 15 working days.

The planned pace isn't a forecast once the team has shown a different one, and the days already spent don't repeat: what matters is the work left at the pace you have.

One forecast number hides the spread, so a reforecast is given as a range: fast, likely and slow, each from the pace the team has really seen. Then each end is compared with the working days left to see the buffer or the miss.

At Klasna some practice topics took two days and the electricity topic took four. For the 5 remaining topics that gives 10 to 20 working days, most likely 15. From 7 to 29 January, today included, there are 17 working days. The fast case leaves 7 days of buffer, the likely case 2, and the slow case misses by 20 − 17 = 3 days.

Read the range for what it tells you. "Most likely 27 January" sounds safe; "likely 2 days to spare, possibly 3 days late" is the honest picture, and it is the one a decision needs. Remember also when the work is really needed: tutors need the last sets before opening to prepare, not on the 29th, so even the likely case is tight.

A reforecast looks forward only. It is built from four things: the work actually left, the pace the team has really shown, who is really available in the coming weeks, and a range rather than one date.

Availability is the input most often forgotten. Calendars and org charts show people; the forecast needs their real days. At Klasna, Vira teaches the January intensive until the 20th, so her real days on the practice sets are fewer than the calendar suggests, and the forecast has to say so.

Two things stay out. The days already spent don't change what lies ahead; adding them only makes an old total look tidy. And the original estimate for the remaining tasks is exactly the number reality has already corrected: reusing it brings back the error the reforecast is meant to fix.

The result is a forecast that may look worse than the plan. That's its value: it shows the problem while there are still options.

Sunk effort is money or time already spent that you can't get back whatever you choose next. Because it is the same under every option, it can't tell the options apart, so it must not decide. Compare only what lies ahead: future money, future days and the dates each option gives.

At Klasna the illustrator has been paid 18,000 UAH for the diagrams of topics 1 to 8. For topics 9 to 12 he asks another 9,000 UAH and can't deliver before 25 January, which leaves tutors almost no time. The alternative: Vira draws simple diagrams in two days now. Danylo argues that switching would waste the 18,000 UAH, but that money is gone under both options.

What does belong in the comparison: the extra 9,000 UAH, the late date, and the cost of Vira's two days in a month when she also teaches. Those differ between options. The pull of sunk effort is natural, which is why it helps to name it out loud: "that's spent; what does each option cost from today?"

A re-plan offers the sponsor options, each with its cost in days and money and its effect on the forecast range. An option without a cost is a wish; an option without an effect isn't worth listing.

For the practice sets at Klasna there are three real options from today. A helper: a vetted tutor who already solved the sample tasks for three topics knows the format, so she can write topics 11 and 12 while Vira reviews. It narrows the range and costs her paid hours plus Vira's review time, and it avoids the start-up cost of someone new. Thinner scope: topics 11 and 12 open with ten tasks each instead of twenty and the rest come in February. It cuts scope below the must line at no extra money. Keep the plan and hope: no cost today, but the full risk of the slow case stays.

Adding a person who doesn't know the work rarely helps late in a project: learning the format eats the time saved. That is why the helper here is someone who already knows it.

When a forecast looks bad, the most tempting line in a re-plan is "the team will push harder". It isn't an option, because it has neither of the two things an option needs: a cost you can name in days or money and an effect you can check.

Effort that isn't in anyone's calendar changes nothing on paper. It moves the risk to the last week, when people are most tired and there is no time left to choose anything else. It also quietly turns a planning problem into a personal one: if the date slips, the story becomes that people didn't try hard enough.

If extra hours are genuinely available, write them as a real option: whose hours, how many, what they cost, which tasks they shorten and what they push out. At Klasna, "a vetted tutor writes topics 11 and 12 for paid hours" is such an option. "Everyone works evenings in the last fortnight" is not, and it belongs out of the note to Leo.

A re-plan ends in a short note to the sponsor, not a new plan document and not an apology. Its order is fixed so the reader gets to the decision quickly:

1. The new forecast as a range, from today.
2. Why it changed, in one line of facts.
3. The options with their cost in days and money.
4. Your recommendation.
5. What you need from the sponsor, and by when.

Leading with the forecast tells the sponsor where the project really is before any explanation. The options and the recommendation turn the problem into a choice. The last line turns the note into action: without a request and a deadline, a note is just news.

Who decides follows the change rule from earlier in this module: if the recommendation stays inside the lead's agreed threshold, the lead decides and the sponsor reads it in the status; if it touches scope in the brief, the sponsor decides in writing. At Klasna, Leo asks for the note today and promises an answer by Monday.

The "why it changed" line of a re-plan note is one line of facts about the work: the planned pace, the real pace and any outlier. At Klasna: "The real pace is 3 days a topic, not the planned 2; electricity took 4."

Three other versions are common and all fail. Blame ("Vira was slower than she promised") turns a fact about the work into a verdict on a person; it teaches the team to hide bad news next time and gives the sponsor nothing to decide with. Drama ("we are badly behind, things look serious") raises the temperature without saying how far behind or what to choose. Vagueness ("some topics are taking a bit longer") is too soft to act on.

A factual line respects both the sponsor and the team. It says the estimate was wrong, which is normal and correctable, rather than that someone failed.

How it fits together: forecast from today with a range, drop sunk effort, list real options with costs, and send a short note with one factual reason and a clear ask. The case applies all of it when the checkout contractor slips two weeks.