Written by Chris Walton, JDYou close a $675 million acquisition. Within months you learn that several of the target’s largest customers had cancelled or scaled back their contracts before closing — and no one told you. You put the overpayment at $22 million. You sue for breach of the purchase agreement. You lose.
That is what happened to Zayo Group in Zayo Group, LLC v. Latisys Holdings, LLC, Vice Chancellor Slights’s 2018 Delaware Court of Chancery opinion, which opens with a blunt reminder: “In law, as in life, the well-known principle of caveat emptor (“let the buyer beware”) comes as a rude awakening to many-a-buyer.” The court held that Latisys did not breach the Stock Purchase Agreement, because the parties had bargained away the very protection Zayo needed. And even if there had been a breach, Zayo’s transaction multiple damages theory failed — its damages expert was excluded because her method did not fit the facts. The buyer walked away with nothing.
For M&A counsel, the case is a tutorial in two failures common to post-closing M&A disputes: the drafting concession you should not have made, and the transaction multiple damages theory that does not survive scrutiny.
Zayo — a publicly traded fiber and data-center infrastructure company based in Boulder, Colorado — had grown by acquisition, making forty-one deals since its founding at a clip of three to four a year. In late 2014 it entered a competitive sale process that the seller’s banker, DH Capital, ran under the code name “Project Caribou.” Zayo made a preemptive bid, pressed for exclusivity, and closed in February 2015 at a price of $675 million.
After closing, Zayo found that several of Latisys’s largest customers had let their contracts lapse or pushed for cheaper terms in the run-up to the deal. The churn was large enough that Zayo believed it had overpaid by roughly $22 million. It sued, arguing that Latisys was obliged to disclose the customer changes and had not.
This is the part of the opinion M&A counsel should read closely. Zayo’s claim rested on Section 4.12(b) of the SPA — one of the seller’s representations and warranties, this one about “Material Contracts.” Latisys represented that it had received no written notice that any party to a Material Contract intended to “cancel, terminate, materially modify or refuse to perform” that contract. Schedule 4.12 defined the Material Contracts as the “Top 30 Customer Agreements” — the Master Service Agreements (MSAs) and related Service Order Forms (SOFs) for Latisys’s largest customers.
Latisys’s customers contracted in two tiers. The MSA set the overall relationship; the SOFs fixed the services, pricing, term, and any renewal options, and an average SOF ran about three years. When a customer’s SOFs expired, it went “out of contract” and bought services month to month, terminable on thirty days’ notice. About thirty percent of Latisys’s revenue sat in that month-to-month state — guaranteed for thirty days and no longer. The business ran on short contracts that came up for renewal constantly.
Here is what decided the case. When the SPA was drafted, Zayo proposed adding the words “or refuse to renew” to Section 4.12(b). Latisys struck them, and Zayo accepted the change without objection. The court found Section 4.12(b) ambiguous — “cancel,” “terminate,” and “refuse to perform” overlap — and turned to the drafting history to resolve it. That history was decisive: by proposing “or refuse to renew” and then giving it up, Zayo had shown that everyone understood non-renewal to be something the clause did not already reach. A customer that let a SOF lapse, or negotiated a cheaper new one, had therefore not “cancelled,” “terminated,” or “materially modified” a Material Contract. The SPA, the court noted, contained no representation about Latisys’s churn, churn rates, or revenue at all. Zayo — a serial acquirer with deep M&A experience — could have insisted on those words. It traded them away.
The decisive fact. Zayo proposed “or refuse to renew,” Latisys struck it, and Zayo accepted without objection. A $675 million dispute turned on a phrase the buyer bargained away.
Even had Zayo proven a breach, its damages case carried a separate, fatal flaw: the method. Zayo’s expert, Kyle Anne Midkiff, valued the alleged $22 million overpayment with an EBITDA multiple, taking the recurring revenue tied to the departing customers and multiplying it by the transaction’s implied multiple of 14.1 times.
The court rejected the approach on two grounds. First, the expert had no experience valuing going-concern businesses — by her own testimony she had never valued one — which left her exposed on a method that turns on valuation judgment. Second, and more important, there was no evidence that the churn permanently reduced Latisys’s value. A multiple capitalizes a loss: it assumes a dollar of lost EBITDA cuts enterprise value by that dollar times the multiple, in perpetuity. But churn at a colocation provider is not permanent. Latisys could replace departing customers — and did; Toshiba returned as a Zayo customer in 2017. A transient loss does not support a perpetual-diminution model.
For litigation counsel the lesson is concrete: transaction multiple damages are the right measure only where the breach permanently and irreversibly impairs the acquired business. Where the breach causes a temporary revenue shortfall the business can recover from, damages should track the transient loss, not capitalize it into a permanent cut in enterprise value. The wrong method does not merely produce the wrong number. It gets excluded.
One caveat keeps that lesson current. A later Chancery decision, In re Dura Medic Holdings, Inc., Cons. C.A. No. 2019-0474-JTL (Del. Ch. Feb. 20, 2025), declined to read Zayo as a categorical bar. The sellers there invoked it for the proposition that a multiple cannot apply without permanent impairment; Vice Chancellor Laster found Zayo factually distinguishable and held that it created no test requiring a permanent loss or diminution in value before a multiple may be used. The governing question, he held, is how far the misrepresentation reaches into future earning periods — where the revenue effect recurs across them, a multiple can stand. The practical line for counsel is therefore narrower than a flat rule: permanence is not a litmus test, but the more transient and recoverable the loss, the harder a multiple is to defend. Zayo remains the cautionary example of a one-time, replaceable churn event; Dura Medic is the reminder that the inquiry turns on the durability of the revenue effect, not a single word.
Read together, the two opinions sketch how Delaware now treats transaction multiple damages in post-closing M&A disputes. The agreement controls first: where it speaks to the measure — as the Dura Medic contract did, defining recoverable losses to include multiple-based damages — the parties get what they bargained for. Where it is silent, the court asks an economic question: does the breach depress the future earnings the buyer paid a multiple to acquire, and does that effect persist? A multiple follows recurring, forward-looking harm, not a shortfall the business absorbs and moves past. Which puts the weight back on the contract — say whether a multiple applies, and to what.
The most useful thing in Zayo v. Latisys is not the valuation analysis. It is the list of protections that belonged in the purchase agreement and were not there. For counsel buying a contract-dependent business, the case maps them out.
Reps on churn and renewal, not just on the contracts. Section 4.12(b) covered the Material Contracts but said nothing about the revenue beneath them. If the valuation rides on retention, the representations and warranties should reach churn rates, non-renewal notices, and pending modifications — and should say “refuse to renew” in as many words.
Coverage at the SOF level, not just the MSA. In a two-tier structure the MSA can survive while the revenue-bearing SOFs lapse. The reps and any bring-down should capture changes in the revenue stream, not merely the existence of the governing agreement.
Interim covenants on customer communications. Between signing and closing, require the target to flag material customer departures, non-renewals, and modifications. Reps alone may not reach what surfaces in that window.
A materiality definition that captures concentration. If a handful of customers drive a large share of revenue, the definition of “material” should reflect it. A single non-renewal below the threshold that, taken with others, costs real revenue is exactly the risk the agreement should catch.
The irony is that Zayo was a sophisticated, serial acquirer — forty-one deals — with the resources to negotiate every protection it later wished for. It had even done the diligence. Latisys handed over a customer-by-customer contract waterfall, and Zayo’s own model showed which of the top-thirty contracts were near expiration or already month to month. Zayo knew. It signed anyway, on reps it had agreed to narrow.
So the cheapest protection on a contract-dependent deal is not the lawsuit that follows, and it is not more diligence — Zayo had plenty. It is holding the line on the reps. Insist on language covering renewal rates and pending non-renewals, not just the existence of contracts. Add interim covenants that force disclosure between signing and closing. If the seller resists those terms, the resistance is itself information about what the seller expects those customers to do.
A buyer who wins those points before closing will not need a damages expert after it — and if the reps turn out to be wrong, the buyer will have a claim the agreement actually covers. Zayo did not.
Zayo v. Latisys is a caveat emptor case wearing the costume of a valuation dispute. Zayo thought its problem was price — a $22 million overpayment. Its real problem was the contract: the SPA did not protect against the risk that materialized, because Zayo had agreed to leave that protection out. The breach question came first and ended the matter; the court addressed damages only for completeness, and found the method wanting there too.
For M&A counsel, the agreement is the first line of defense, not the lawsuit that follows. For damages experts, the warning is plain: a transaction multiple is the wrong tool when the loss is not permanent, and a court will throw out a method that does not match the economics of the breach. And for buyers of contract-dependent businesses, the question to ask before closing is not “what are the contracts worth?” It is “what happens to the revenue when these contracts come up for renewal — and does my agreement protect me if the answer is bad?”
Chris Walton, JD, is the President & CEO of Eton Venture Services.
If you are weighing a post-closing damages claim and need to know whether a transaction multiple or another measure fits the facts, I am glad to talk it through; at Eton, the method question often decides whether a claim is viable before anyone runs a number.
Zayo Group, LLC v. Latisys Holdings, LLC, C.A. No. 12874-VCS (Del. Ch. Nov. 26, 2018) (Slights, V.C.).
In re Dura Medic Holdings, Inc., Cons. C.A. No. 2019-0474-JTL (Del. Ch. Feb. 20, 2025) (Laster, V.C.).
Schedule a free consultation meeting to discuss your valuation needs.
