← Back to blog

Hidden Turnaround Costs Nobody Forecasts

Six line items that never make the operating plan and always show up in month three. Real dollar ranges, from a $30M multi-unit restaurant turnaround that ended clean.

Kiyong (Eric) Lee·Operations Executive·August 2026·9 min read

The short version

Every multi-unit turnaround plan I have seen underbudgets restructuring costs by 4 to 6 percent of annual revenue.
The six line items that always show up: outside legal, severance, retention bonuses, deferred R&M catch-up, tech migration overlap, and temp coverage during restructure.
On a $30M group with 3 underperforming units, the hidden bucket ran roughly $1.4M over 11 months. It was not slack. It was the actual cost of doing the work.
Retention will cost more than severance. The people you need to keep are worth more than the people you are letting go.
Itemize the reserve up front. Boards approve 5 percent when it is defended. They reject it when it shows up as a variance in month five.

Every turnaround plan I have ever written or read has the same structural flaw. It models the recovery in the operating P&L and forgets to model the cost of doing the recovery. The operating gains show up clean in the spreadsheet. The cost of getting to them shows up as variances, and the board reads variances as bad news.

The frustrating part is that the missing costs are the same six or seven line items every time. They are not exotic. They are not one-off. They are structural to how turnarounds actually work. If you have run one, you know them. If you are about to run one, this is the list I wish someone had handed me the first time.

The numbers below are from a Fractional Head of Operations engagement I ran at Zareen's, a $30M Michelin-recognized Bay Area restaurant group with a workforce of 215 and an enterprise catering channel serving Stanford, Google, Apple, Meta, LinkedIn, Salesforce, Cisco, Adobe, and Nvidia. Three of the five locations were losing money. Over 11 months we recovered roughly $4.9M in operating profit. The hidden cost bucket, the thing that never made the original plan, ran about $1.4M in the same window. That is real money. It is also the money that made the $4.9M possible.

Why the plan always misses these lines

Turnaround plans get built by the same operators who are going to execute them, usually under time pressure, usually to defend a specific outcome to a specific audience. That audience wants to see recovery, not restructuring cost. So the plan gets built as an operating recovery model, with a small "restructuring reserve" line at the bottom that is almost always half of what it should be.

The other reason is more human. Nobody wants to write down, in month one, that they are going to spend $180K on separation agreements and $220K on retention. It looks defeatist. It is not defeatist. It is arithmetic. You cannot rebuild an operating system without paying to move people and to keep people, and those two costs sit on opposite sides of the same restructuring.

The recovery is the visible number. The cost of the recovery is the invisible one. If you do not name the invisible one out loud in month one, it will name itself as a variance in month five.

The six line items that always show up

Below are the six buckets that missed the original plan on my last three engagements. Ranges are for a three to five unit restaurant group in the $20M to $40M revenue band. Yours will vary. The pattern will not.

Fig. 1 · Typical hidden buckets, three to five unit restaurant group, 12 months.

1. Outside legal fees

Even the cleanest operational turnaround pulls outside counsel. Separation agreements need drafting. Lender covenants need reviewing before you send anything in writing. Vendor contracts get renegotiated and someone needs to red-line the master service agreements. Lease amendments show up if you are consolidating or subleasing space. On our engagement, outside legal ran about $95K across 11 months and I still consider that light. Budget $40K to $120K for a three to five unit group on a normal turnaround. If there is any active litigation or wage-and-hour exposure, add a zero to the top of the range and do not argue about it.

2. Severance and separation

You will let people go. Not on day one, and not many, if the sequencing is right. But by month four you will have identified two to five people whose seat is more valuable empty than filled, and the cost of that decision is not zero. Separation agreements, unused PTO payouts, COBRA subsidy for a defined window, and in California specifically, final-check timing that has to be right or you compound the cost. On our engagement severance ran about $185K for six separations across two locations. That is roughly $30K per separation for mid-level ops leadership. A general manager separation with tenure will run higher.

3. Retention bonuses

This is the line most operators refuse to write down until they lose someone. The retention line is bigger than the severance line in every turnaround I have run. The math is simple: the people you cannot afford to lose in the middle of a restructure are worth paying to keep. On our engagement I identified nine people across the three locations and the corporate team whose departure would set the turnaround back a full quarter. We put stay agreements in place, mostly 15 to 25 percent of annual comp payable at month twelve, contingent on staying through the transition. Total retention pool came to $220K. Every one of them stayed. I would spend the money again tomorrow.

4. Deferred repair and maintenance catching up

This is the one operators tend to underestimate the most because it is invisible until you start looking. Underperforming units defer R&M to protect the current-month P&L, which is a rational short-term move that becomes an expensive long-term problem. When a new operator arrives and installs a real R&M log, the backlog surfaces all at once.

Compressor rebuilds. Hood cleanings that got skipped for two quarters. Floor drains that have been slow all year. HVAC compressors limping. Walk-in door gaskets that stopped sealing eight months ago and drove the utility bill up 12 percent per month while nobody noticed. On our engagement, catch-up R&M ran about $260K across the three underperforming locations in year one, roughly $85K per unit. Some of that was capital-preserving work that should have been done in the two years prior. All of it hit the current period.

5. Tech migration overlaps

If your turnaround involves replacing the POS, scheduling system, back-office platform, or catering CRM, you will run old and new in parallel for 60 to 120 days. Both vendors bill. Both need administration. Both need training time. Nobody plans for this. It always happens.

On our engagement we replaced the scheduling system, added a real BI layer over Toast data, and migrated the catering CRM. Duplicate subscription cost during overlap ran about $58K, integration and data migration work another $45K, and internal labor time for training and parallel-running easily another $30K in fully-loaded cost. Call it $130K. It looks like nothing until you add it up.

6. Temp coverage during restructure

Every time you move a general manager, promote a sous chef into a chef role, or backfill a corporate seat, you create a temporary coverage gap that has to be paid for. Interim managers, cross-market travel for area directors covering an extra unit, contract chefs for six weeks between the old chef leaving and the new chef starting. It adds up faster than anyone expects.

Across our 11 months, temp coverage ran about $215K. That included two interim general manager stints of six weeks each, one contract executive chef bridge for 90 days, and roughly $40K of travel and per-diem for area leaders covering out of market. This is the invisible tax on organizational change, and it is not optional.

What these numbers look like at different scales

The ranges above are calibrated to a three to five unit restaurant group in the $20M to $40M revenue band. The pattern scales in both directions, but not linearly. Two things worth knowing.

At the small end, a two-unit or single-brand group in the $6M to $12M range, the six buckets still all appear. What changes is that the fixed cost of each one lands heavier as a percent of revenue. Outside legal has a floor that does not compress much below $30K in a turnaround year regardless of unit count. Retention has a floor because the number of critical seats does not shrink proportionally to unit count. A $10M group running a turnaround will often see the hidden bucket land at 6 to 9 percent of revenue rather than 4 to 6, and the plan should reflect that.

At the larger end, seven or more units and $60M plus, the buckets get bigger in absolute dollars but usually shrink as a percent of revenue, because the fixed pieces amortize across more units. The exception is tech migration overlap, which scales roughly linearly with unit count because every location has its own POS terminal count, its own training curve, and its own parallel-run window. A ten-unit tech migration is not twice a five-unit migration, it is closer to two-and-a-half times, once you account for the coordination cost.

The bucket that used to bite me the hardest

Early in my career I underbudgeted retention every single time. I thought severance would be the big number because it was the more visible one. I was wrong every time. The people who leave voluntarily during a turnaround are almost never the people you wanted to leave. They are the strong ones who read the room, decide it is going to be a hard year, and go take the calmer job at your competitor.

The second-worst was deferred R&M. I would inherit a set of P&Ls that looked artificially clean in the months just before the engagement started, and I would build the operating plan against those numbers. Six weeks in, the hood cleaning invoice would land. Then the compressor. Then the HVAC. By month three the R&M line would be running triple the original budget and I would be defending it in the board pack.

Retention costs more than severance. Deferred R&M costs more than either. Plan for both in month one, or defend both as variances in month five.

How to itemize the reserve so it actually gets approved

Boards do not reject 5 percent of revenue in restructuring reserve. They reject vague restructuring reserves. The difference is entirely in how the ask is presented. When I bring a turnaround plan to a board or a lender now, I put the six buckets on a single page with:

A dollar range for each bucket, benchmarked to a comparable prior engagement.
The specific decision each bucket enables. Retention buys 9 named people through month 12. Legal enables 6 clean separations. Tech overlap enables the POS migration that unlocks 2 points of margin visibility.
A trigger threshold. If bucket X exceeds range by 25 percent, we surface it in the next weekly and re-plan.

Presented that way, a $1.4M reserve on a $30M group reads as diligence, not defeatism. Presented as a single "restructuring costs" line with a round-number placeholder, it reads as a blank check and gets cut in half. Same money. Different framing. Different outcome.

The point

The operating recovery is the visible half of a turnaround. The restructuring cost is the invisible half, and the invisible half is what pays for the visible half to be real. Legal, severance, retention, deferred R&M catch-up, tech overlap, temp coverage. Six line items. Roughly 4 to 6 percent of annual revenue on a normal engagement. Higher if the restructure is deeper.

Write them down in month one. Defend them by naming the decisions they enable. Then run the recovery in the open, so that the board reads the reserve as the price of the outcome, not as a surprise in the middle of the story.

Turnarounds are not free. Pretending they are is what turns clean engagements into ugly ones.

Building a turnaround plan right now?

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 working through the reserve line and want a second pair of eyes on the ranges, I am happy to trade notes.

Get in touch

Questions people ask

What costs are usually missing from a turnaround budget?

Six line items show up in almost every multi-unit turnaround budget after the plan is already signed: outside legal fees, severance and separation costs, retention bonuses for the people you need to keep, deferred repair and maintenance catching up, tech migration overlaps where old and new systems both bill, and temp coverage during restructure. Together these usually add 4 to 8 percent of annual revenue to the plan operators originally wrote down.

How much should a turnaround plan reserve for hidden costs?

For a multi-unit restaurant group in the $10M to $50M revenue range, reserve 4 to 6 percent of annual revenue for costs that are not in the operating plan. On a $30M group that is roughly $1.2M to $1.8M. If you are also doing a tech migration or a significant separation, budget closer to 7 or 8 percent. The reserve is not slack. It is what pays for the actual work of changing the operating system.

Why do retention bonuses matter more than severance in a turnaround?

Severance is the price of the people you are letting go. Retention is the price of the people you cannot afford to lose while you rebuild. In every turnaround I have run, the retention line ended up larger than the severance line. Losing your best area manager or your best executive chef in month three costs more than paying them to stay through month twelve.

How do deferred R&M costs show up during a turnaround?

Underperforming operations defer maintenance to protect the monthly P&L. When a new operator arrives and starts logging every equipment issue, the backlog surfaces all at once. Compressor rebuilds, hood cleanings that were skipped, floor drains, HVAC that has been limping. Plan on $8K to $25K per location in year one of catch-up R&M, and treat it as capital preservation, not a surprise.

What is a tech migration overlap and why does it cost so much?

During any turnaround where you replace the POS, scheduling system, or back-office platform, you will run the old and new systems in parallel for 60 to 120 days. Both vendors bill. Both need administration. Both need training. On a five-unit group this overlap is usually $15K to $40K in duplicate subscription and integration cost, plus internal labor. It is real and it is unavoidable.

Do outside legal fees always show up in a restaurant turnaround?

Yes. Even a clean operational turnaround pulls outside counsel for separation agreements, lender covenant review, vendor contract renegotiation, and lease amendments. Budget $40K to $120K for a three to five unit group over a 12 month turnaround. If there is any litigation exposure, add a zero to the top of the range.

How do you defend a bigger turnaround budget to the board?

Show the six hidden line items with dollar ranges, benchmarked to prior engagements, and label them what they are: the cost of actually executing the plan the board just approved. Boards accept 4 to 6 percent of revenue in restructuring reserve when it is itemized and defended. They reject it when it shows up as a surprise variance in month five.