The Rescue Job

Sustaining Turnaround Gains Into Year Two

Most turnarounds decay in month fourteen. Here is what the operator has to install by month twelve to prevent it, and what quietly rots first when nobody is watching.

Kiyong (Eric) Lee·Operations Executive·August 2026·10 min read

The short version

Year-one gains are held in place by the operator's attention. Year-two work is about transferring that hold to the organization.
The classic decay point is month fourteen. Two months after the operator's presence starts to fade.
Institutionalize the cadence, promote a real second-in-command, tighten the dashboards, add one strategic bet.
Labor decays first. Waste and comps decay next. Food cost decays last.
Five strategic bets in year two is how you dilute the gains. One bet is the right cadence.

The first year of a turnaround gets all the attention. The board deck. The recovery number. The story. Somewhere around month eleven the operator looks around and thinks the hard part is done. The P&L looks healthy. The team knows the drill. The units are running.

Month fourteen is when the trouble starts. The labor line drifts up half a point. Then another. The waste log gets filled in less often. A comp policy that everyone knew a quarter ago is now interpreted three different ways at three different units. The dashboard the general manager used to open every morning is open some mornings. By month sixteen the gains are quietly leaking, and by month eighteen the group is having a hard conversation about why the recovery did not hold.

This is not a story about the wrong operator or the wrong team. This is a story about the difference between year-one work and year-two work, and the fact that operators trained in year-one work often do not know that year-two is different.

Why year one is fragile

Year-one turnaround gains are load-bearing on the operator. The operator sees the number drift, the operator has the conversation, the operator rebuilds the schedule, the operator sits in on the weekly review. This works while the operator is present at that intensity. It does not survive that intensity ending.

In year one, the operating rhythm you installed was scaffolding. The weekly P&L review, the monthly regional operating review, the dashboard. Those things worked because you were there to enforce them. Year two is when the scaffolding either becomes a real building or comes down with you.

Fig. 1 · Same year-one recovery. Two completely different year-two outcomes.

The four pieces of year-two work

There are four things that have to be true by month twelve if you want the gains to hold through year two. All four are different in kind from year-one work. All four take longer to install than you think. Start them at month nine at the latest.

1. Institutionalize the operating cadence

By month twelve, every recurring meeting, dashboard, and review has to live on the organization's calendar, not the operator's. This sounds administrative and is actually cultural. If the weekly P&L review is scheduled because the operator sends the invite every Monday morning, the review will stop happening within eight weeks of the operator stepping back. If the review is a standing organizational meeting with a documented agenda, a named owner other than the operator, and a location on the shared calendar that a new hire could find in five seconds, it survives.

Practical test: could a new area director, hired in month fifteen, run the exact same operating rhythm without any coaching, just by reading the documented cadence? If yes, the cadence is institutionalized. If no, it is still living inside a person.

2. Promote a real second-in-command

The turnaround operator is often the only person who has held the full picture across all units. If that person leaves and there is no successor who has held the full picture, the gains cannot survive.

The second-in-command has to be in place, running the operation with real decision rights, by month twelve at the latest. That means the promotion happens by month six, the overlap runs from month six to month twelve, and by month twelve the second-in-command is running the standing meetings, owning the dashboards, and making the operating calls. The turnaround operator is a coach in that window, not a driver.

This is a promotion, not a hire, in almost every case. The internal candidate has context that no outside hire can replicate in year one. Bring in an outside second only if the internal bench is empty, and even then, expect the outside hire to be operationally productive around month nine of their tenure, not month one.

The second-in-command is not the operator's replacement. The second-in-command is the operator's proof that the operating system exists outside the operator's head.

3. Tighten the dashboards from diagnostic to operational

Year-one dashboards are long. They have to be, because the operator does not yet know which symptoms matter most. You are surfacing many numbers so the picture can form. That dashboard was right for its job.

Year-two dashboards are short. Once the operator knows the operation, the dashboard should collapse to the four numbers that predict everything else at the unit level, updated daily. For most restaurants those are labor as a percent of sales, food cost as a percent of sales, comps and voids, and one unit-specific leading indicator (usually cover count, average check, or a customer-satisfaction proxy).

Everything else moves to weekly or monthly reporting. If the daily dashboard is long, the general manager stops opening it. If the daily dashboard is short and the four numbers are the right four, the general manager opens it every morning without being asked.

4. Add one strategic bet, not five

Year two is the earliest you should introduce real growth work. A new channel. A new market. A new format. A meaningful investment in technology or brand.

The instinct is to add several. The group looks healthy, the team looks capable, and there are five strategic ideas that have been queued up all through year one waiting for the operation to be ready. Do not add five. Add one.

The bandwidth to run a newly stabilized operation and manage one strategic bet is realistic. The bandwidth to manage three or five strategic bets on top of an operation that is still consolidating is not. Overloading year two with strategic work is the single most common way I have seen year-one gains dilute and disappear.

On the Bay Area group, year-two strategy was expansion from three to five units. That was the one bet. The catering channel expansion into Stanford, Google, Apple, Meta, LinkedIn, Salesforce, Cisco, Adobe, and Nvidia had already started in the back half of year one. New menu development, new tech stack changes, and a marketing rebuild all waited for year three. That sequencing was not an accident. It was what made the two expansions work.

What decays first if you are not watching

The gains do not all decay at the same rate. There is a predictable order, and knowing the order tells you where to look first when you audit at month twelve.

Fig. 2 · Labor is always the canary. If it drifts, the rest is coming.

Labor decays first

The labor schedule is the hardest gain to defend because it requires weekly rework against the demand curve. Every week, someone has to look at the last four weeks of hourly sales, adjust the shift shapes, and publish. If that weekly discipline slips, labor variance drifts within four to six weeks. Labor is the canary. When the labor number starts moving in the wrong direction, everything else is coming.

Waste and comp discipline decays next

Closing rituals slip on a slightly longer timeline. The waste log gets filled in less faithfully. Comps get processed with less discipline. This shows up on the P&L about two months after labor starts drifting, and it tells you that the closing-shift culture has softened.

R&M and food cost decay last

Root-cause equipment maintenance and full inventory discipline are usually the last to go, because they were the last to install and the team's ownership of them tends to be strongest. When these decay you know the whole system is in trouble, not just one habit.

The month-twelve audit

To catch decay before it costs you, schedule a month-twelve audit at the start of year two. This is not a general operations review. It is a specific check against the top three or four turnaround metrics.

The audit answers four questions:

Are the labor, food, and comp numbers still inside the year-one recovered range, at every unit? If any unit has drifted more than a point on any of them, dig into that unit specifically.
Is the operating cadence happening on the documented schedule, without the operator prompting? Sit in on a weekly review you did not initiate. Watch.
Is the second-in-command actually running the operation, or are the general managers still routing decisions to the turnaround operator? Ask the general managers, not the second-in-command.
Is the daily dashboard being opened every morning by every general manager? Check the usage logs, not the reports.

If any of those four is soft, you have between four and eight weeks to reinforce before the number moves. If all four are strong, year two will hold.

Who should run the audit

The turnaround operator should not run the month-twelve audit alone. If they do, the audit becomes another form of the operator holding the operation together, which is the exact pattern year two is supposed to break. Run it with the second-in-command holding the pen and the operator listening. If the second-in-command finds the drift on their own, the transfer of ownership is real. If the operator is still surfacing every issue, more year-two work remains.

What to do when a unit is already drifting

Occasionally the audit surfaces a unit that has slipped further than the four-to-eight-week reinforcement window can catch. Do not try to fix it from the road. Send the second-in-command to spend a working week onsite, at that unit specifically, running the standing meetings in person and rebuilding the general manager's daily habits alongside them. A concentrated week of presence corrects a quarter of drift more reliably than a month of remote nudging. The presence, not the report, is what resets the habit.

The year-two operator identity shift

There is a psychological shift most turnaround operators do not prepare for. In year one you are the person who fixes things. In year two, if you are doing the job right, you are the person who trains the person who fixes things. That is a different identity, and operators who cannot make the shift end up starting new fires so they can put them out again, which is how gains get burned twice.

Watch yourself for this pattern. If you find yourself, in month fifteen, jumping into a labor variance issue at one unit instead of coaching the second-in-command through it, you are re-enacting year one instead of executing year two. The fix is uncomfortable and simple: hand the issue back, watch the second-in-command handle it, and resist the urge to touch it.

What sustains, in one sentence

Year-two work is not more of year-one work. It is the deliberate transfer of ownership from the operator to the organization: the cadence to the calendar, the picture to the second-in-command, the diagnostic to the operational dashboard, and the growth energy to exactly one strategic bet.

Turnaround operators who fail at year two are almost always people who could not stop being the operator. The gains hold when the operator has done the harder work of building someone else who can hold them.

The point

Most turnaround stories end at month twelve because month twelve is where the recovery number gets reported. The real story of whether a turnaround worked runs through month twenty-four, and it is decided by the four things installed in the second half of year one.

Institutionalize the cadence so it lives on the calendar and not in your head. Promote a second-in-command who is running the operation with you visible only as a coach by month twelve. Tighten the daily dashboard from a diagnostic instrument to an operational one, four numbers per unit. Pick one strategic bet, not five, and let the rest wait for year three.

Then, at month twelve, audit against decay. Look at labor first, because labor is always where decay shows up first. If the four things are in place and the audit is clean, year two will hold, and year three becomes possible on top of what you built.

The turnaround does not end when the numbers land. It ends when the numbers hold themselves.

Sitting on a year-one recovery, worried about the decay?

I take on a small number of multi-unit turnaround engagements each year, mostly restaurant and hospitality groups in the $10M to $60M revenue range. If you are approaching month twelve and want a second read on whether the gains will hold, I am happy to trade notes.

Get in touch

Questions people ask

Why do turnaround gains decay in year two?

Because year one was held together by the operator's attention and personal cadence. When that operator steps back or moves on, the cadence goes with them if it was never institutionalized. The dashboards stop being read, the standing meetings drift, and the discipline that produced the gains erodes. The classic decay point is month fourteen, roughly two months after the operator's presence starts to fade.

What is the biggest year-two mistake in a restaurant turnaround?

Adding too many strategic bets too early. Year two feels like the moment to grow, and operators often stack a new channel, a new market, and a new format all at once. The bandwidth required to run the newly stabilized operation cannot also support three strategic projects. One strategic bet per year is the right cadence in year two. Five is how you burn the gains.

How do you institutionalize an operating cadence?

Move every recurring meeting, dashboard, and review from the operator's calendar to the organization's calendar. Write the standing agendas down. Assign owners other than the turnaround operator. Publish the schedule. When someone new joins, they should be able to look up when the weekly P&L review happens and who runs it, without asking anyone. If the cadence only lives in the operator's head, it leaves when the operator leaves.

When should you promote a second-in-command after a turnaround?

By month twelve at the latest. The second-in-command needs at least six months of running the operation with the turnaround operator still present before the operator steps back. Promoting in month eighteen and stepping back in month twenty is how second-in-commands fail publicly. The overlap matters.

What decays first in a turnaround if you are not watching?

The labor schedule. In almost every case. It is the hardest gain to defend because it requires weekly rework against the demand curve. If the standing weekly scheduling exercise slips, labor variance drifts back within four to six weeks. Waste logs and comp discipline decay next, on a slightly longer timeline. R&M and food cost discipline decay last.

How should the year-two dashboard differ from year one?

Year-one dashboards are diagnostic and long. They have to surface many symptoms because the operator does not yet know which ones matter most. Year-two dashboards are operational and short. Four numbers per unit, updated daily: labor as a percent of sales, food cost as a percent of sales, comps and voids, and one leading indicator specific to that unit. Everything else moves to weekly or monthly reporting.

Can you sustain turnaround gains without the turnaround operator?

Yes, but only if the year-two work has been done. The operating cadence has to be institutionalized, the second-in-command has to be in place and running the operation with real ownership, and the dashboards have to be operational rather than diagnostic. If those three are in place by month twelve, gains hold through year two and beyond. If any one is missing, decay is almost certain.