Most CEOs preparing for an exit think their job is to find the right banker.
The banker's job is to find the right buyer.
The CFO's job is the one in the middle. And it's the one that quietly decides whether your multiple holds, slips by a turn, or implodes in week six of diligence.
I learned this from both sides. As CFO of Medicx Health, I sat next to the banker through the process that ended in OptimizeRx acquiring us for $95M in October 2023. Before that, I had been on the other side of enough banker-led processes to know what makes them go and what makes them stall. The bankers I worked with over the years had strong opinions about what they wanted in a seller's CFO. Most CEOs never hear those opinions until after they have already hired one.
This is what they actually want.
01. Numbers that survive QofE without rework
Bankers have one big fear in the early weeks of any sell-side process. Not that the buyer will walk. That the QofE team will walk in, dig through the books, and produce an EBITDA number that's materially lower than the management number in the teaser.
When that gap shows up in week 4, three things happen. The buyer's offer comes in below the teaser range. The banker has to defend a credibility hit they had no part in creating. And the CEO learns mid-process that the multiple they were working toward was theoretical.
The CFO who survives this moment is the one who built the management EBITDA anticipating what QofE would find. Not after-the-fact reconciled. Built that way from the beginning. Reasonable add-backs documented with policy memos. Owner expenses identified and stripped before they become a discovery moment. Out-of-period items quantified and shown net. Revenue recognition tied to GAAP, not to the cash hitting the bank.
Bankers want the CFO who can walk through every adjustment with a paper trail, not one who can only narrate the highlights. If your QofE comes back with a meaningful negative adjustment, your offer comes back lower. That is not a banker preference. That is how the market clears.
02. A CFO who owns the data room personally
The fastest way to telegraph that the seller is not ready is for the data room to be managed by an analyst.
Bankers see this constantly. The data room goes live. The buyer sends a diligence request. The request gets passed to a controller, who passes it to an analyst, who pulls a report, sends it back up the chain, and three days later a banker gets a partial answer that triggers a follow-up question. The cycle repeats. By week 2, the buyer's diligence team is frustrated, the banker is losing momentum, and the CEO is wondering why everything feels heavier than it should.
The CFO who owns the data room personally short-circuits this. They know where every file lives. They have anticipated the questions the buyer will ask before the buyer asks them. They can answer 70% of diligence requests within 24 hours and pre-empt half of them. They keep the banker on offense.
That does not mean the CFO is doing all the work. It means the CFO is the single accountable owner. Analysts and controllers do the underlying pulls. The CFO reviews, sequences, and delivers. The banker has one phone call to make when something is missing, not three.
This is why bankers will privately tell CEOs that the CFO is the most important hire before launch. Not after.
03. A management presentation that holds up under pressure
The management presentation is the moment buyers stop reading the deck and start studying the people in the room.
A polished deck does not survive sophisticated diligence. A polished CFO does.
What bankers want is a CFO who has been pressure-tested on every page of the model. Customer concentration question? The CFO has the answer in one breath, with context, and pivots to why the concentration is durable. Margin compression question? Already modeled in three scenarios. Working capital growth assumption? Tied to revenue with a stated days-on-hand standard that survived two prior reviews.
Bankers will rehearse the management presentation with the CFO three or four times before the first real session. The good CFOs want the rehearsals. The ones who push back are the ones who underperform when the room is full.
The CFO is the most visible signal of management quality during the process. Calm, prepared, and direct under questioning is the signature buyers and bankers want to see.
04. Working capital normalization done before the data room opens
One of the most expensive moments in any sell-side process is the working capital true-up debate at the closing table.
Buyers want the working capital peg to reflect "a normal level required to operate the business." Sellers want it set as low as possible. The QofE team writes a memo with their view. The banker negotiates. The lawyers paper it. And somewhere in the middle, a million dollars or more of purchase price moves in one direction or the other.
The CFOs who win this negotiation are the ones who did the work before the data room opened. They have a documented working capital normalization methodology. They have a trailing 12-month average, a seasonal adjustment, and a deferred revenue treatment that's been thought through. They know what a "normal" AR aging looks like in their business and can defend any deviation. They have a view on inventory build for new product launches and on prepaid expenses that recur quarterly.
The buyer's team will arrive with their own view. The CFO who already has theirs documented in a memo and reflected in the management projections is the one who controls the negotiation. The CFO who is reacting in real time to the buyer's working capital schedule is the one who loses purchase price.
This is one of the hardest workstreams to retrofit in the middle of a process. Bankers wish more CEOs understood that the working capital work should start 6 to 9 months before launch, not 6 to 9 weeks.
05. A forward model with assumptions you would defend in deposition
The forward model is where banker frustration with CEO-CFO dynamics peaks.
The CEO wants the forecast to be aspirational. Wider TAM, faster ramp, expanding margin. That is how multiples expand. The CFO wants the forecast to be defensible. Tied to documented unit economics, tested against history, sensitized for downside cases. That is how multiples survive diligence.
Both views are right, in tension, and need a single voice that reconciles them.
The bankers I respected most all said the same thing in different words. They wanted a forward model where every assumption had a defense. Not a hope. A defense. New customer ramp tied to documented pipeline conversion rates. Margin expansion tied to a specific operational lever, not "scale." Cash flow conversion modeled honestly with working capital build at higher revenue.
The CFO's job is to give the CEO a model both can stand behind. If the CEO is the only voice on the forward picture, the model loses credibility the moment a serious diligence team gets a look at it. If the CFO is the only voice, the model loses ambition and the multiple compresses.
What this means for CEOs preparing for exit
The hardest part of preparing for exit is not strategic. It is not even financial. It is the realization that most of the work needs to happen 12 to 18 months before the banker is mandated, not after.
The seller's CFO either is, or is not, the person who has done that work. Bankers know within the first hour of meeting your CFO whether the process is going to be a clean run or a managed mess.
If you are 12 to 18 months out from a transaction and you are not certain your CFO is the right person to sit next to the banker, that is the question to answer now. Not in week 4 of diligence.
A few specific signals worth checking:
- Has your CFO sat through a QofE process before, on either side?
- Can your CFO produce a defendable normalized EBITDA bridge today, without two weeks of prep?
- Does your CFO own the data room when you run a refinancing or a board reporting cycle, or is it always delegated?
- Does your CFO push back on aspirational assumptions in your operating plan, or sign off without challenge?
The answers tell you whether the seat is filled.