MALDONADO, Circuit Judge. CLEAResult Consulting, Inc. hired NEXT Payment Solutions, Inc. to develop appointment scheduling software. After a few years of successful partner- ship and use of NEXT’s application, CLEAResult decided to shift to different scheduling software. Before making the tran- sition, and without telling NEXT, CLEAResult modified and refined its new program based on the functionality and

2 No. 24-1377

features of NEXT’s application. CLEAResult then terminated its relationship with NEXT and stopped using its scheduling tool. NEXT sued CLEAResult for misappropriation of trade se- crets under federal law and unjust enrichment under Illinois common law. 1 After years of contentious litigation, the dis- trict court entered summary judgment against NEXT on the trade secrets claim because NEXT failed to define its secrets with enough specificity. The case moved to pretrial proceed- ings on the unjust enrichment claim, and the district court granted a motion in limine limiting the scope of the claim. NEXT then voluntarily dismissed the claim, and the district court entered final judgment in favor of CLEAResult. We af- firm. BACKGROUND I. Factual Background NEXT designs and develops customer service software for businesses. NEXT’s primary software platform is appropri- ately called the “NEXT System.” NEXT offers customizable versions of the NEXT System for its customers’ business needs. CLEAResult provides North American utilities with en- ergy efficiency programs and services for utility customers. As part of its services, CLEAResult offers utility customers the opportunity to schedule in-home appointments for services like home energy-efficiency assessments.

1 NEXT asserted several other common law claims that are not rele- vant to this appeal.

No. 24-1377 3

In 2014, CLEAResult hired NEXT to develop an online- scheduling tool. NEXT worked with CLEAResult to under- stand its needs and created a customized software application for booking appointments called the “FAST Tool.” The FAST Tool contained a public-facing side, where cus- tomers could access their accounts and schedule appoint- ments, and an internal side, where CLEAResult personnel could manage appointments and access customer infor- mation and data. There were also underlying software me- chanics, including rules, algorithms, and source code to which only NEXT had access. CLEAResult used the FAST Tool for several years without issue. Then in early 2017, CLEAResult acquired a separate green technology company that had its own cloud-based ap- plication, known as the DSMTracker. CLEAResult decided to replace the FAST Tool with that new platform. According to CLEAResult, it wanted to analyze how the DSMTracker needed to be refined to meet its functionality needs. So CLEAResult launched “Project Renaissance,” an internal analysis of the gaps in the DSMTracker. As part of that effort, CLEAResult developers looked at the internal-facing side of the FAST Tool—the feature used by customer service person- nel to manage appointments and customer data—to ensure that the DSMTracker covered the same functionality. NEXT presents a different picture of Project Renaissance. According to NEXT, CLEAResult surreptitiously, and with- out NEXT’s approval, stole secret information from the FAST Tool to reverse engineer and replicate its features into the DSMTracker. Notably, however, NEXT concedes that CLEAResult personnel never had access to the underlying processing engine or rules software of the FAST Tool. They

4 No. 24-1377

further agree that there is no evidence CLEAResult ever ac- cessed the FAST Tool’s underlying source code. In November 2017, after CLEAResult finished Project Re- naissance and transitioned several customers to the DSMTracker, it informed NEXT it was ending their relation- ship. II. Procedural History Shortly thereafter, NEXT initiated this lawsuit asserting, among other things, claims for misappropriation of trade se- crets under the Defend Trade Secrets Act, 18 U.S.C. §§ 1836 et seq. (“DTSA”), and unjust enrichment under Illinois common law. NEXT alleged that CLEAResult violated the DTSA by misappropriating trade secret information from the FAST Tool to better the DSMTracker’s functionality. As to unjust enrichment, NEXT pled two separate theories of recovery: that CLEAResult refused to pay certain invoices, and that CLEAResult had “access to NEXT trade secrets and other pro- prietary information” and was unjustly enriched when it “misused NEXT’s trade secrets and proprietary information … to create competitive systems.” Following discovery, CLEAResult twice moved for sum- mary judgment on NEXT’s DTSA claim arguing that NEXT failed to identify its software secrets with sufficient specific- ity. In its first summary judgment ruling, the district court agreed with CLEAResult in part that NEXT’s secrets were too nonspecific and narrowed the DTSA claim to “those parts of the FAST Tool that were not present in the DSMTracker be- fore Defendant transitioned” to its new program. Later, the court ordered NEXT to provide more infor- mation on the features of the FAST Tool that it contended

No. 24-1377 5

were misappropriated and to explain why they were trade se- crets. NEXT produced a spreadsheet identifying a list of thirty-four software “modules,” and five “combinations of modules and features” that it said CLEAResult took from the FAST Tool. The modules—discussed in more detail below— had titles such as “Online Self Scheduling” and “Dashboard Client.” NEXT describes the modules with generic language such as “manages the inventory of appointments,” “pre- sent[s] real-time appointment availability,” and “create[s] open appointment slots, to update appointment availability.” CLEAResult, unsatisfied with NEXT’s updated descrip- tions, again moved for summary judgment. This time, the dis- trict court granted the motion in full. The court explained that NEXT had been given multiple opportunities to identify “spe- cific and concrete” misappropriated secrets in its software but came forward with nothing more than “broad descriptions and jargon-laden terminology” describing what its software did, not how it did it, which made it “hazy what, exactly, NEXT is claiming as its alleged trade secrets.” More than a year later—during pretrial proceedings on NEXT’s remaining claims—a dispute arose over the scope of NEXT’s unjust enrichment claim. In a pretrial status report, NEXT asserted for the first time that it intended to bring to trial an independent claim for unjust enrichment based on the theory that CLEAResult misused “other proprietary infor- mation” separately and apart from misusing NEXT’s trade se- crets. On a motion in limine from CLEAResult, however, the dis- trict court excluded any evidence or argument of misuse of “other proprietary information” as a basis for unjust enrich- ment. The court found that this theory of unjust enrichment

6 No. 24-1377

had never before been presented as a separate theory from misappropriation of trade secrets. The court held such a late change was improper and that NEXT was, in essence, trying to put forth a new theory on the eve of trial as an end-run around the dismissal of its DTSA claim. After the ruling in limine, NEXT voluntarily dismissed its remaining claims with prejudice. It appeals the district court’s dismissal of its DTSA claim and the in limine order on its claim for unjust enrichment. DISCUSSION I. Summary Judgment on NEXT’s DTSA Claim We review de novo the district court’s grant of summary judgment. REXA, Inc. v. Chester, 42 F.4th 652, 662 (7th Cir. 2022). Summary judgment is appropriate when, drawing all reasonable inferences in the nonmovant’s favor, there is no genuine dispute of material fact, so the moving party is enti- tled to judgment as a matter of law. See id. To prevail on a DTSA misappropriation claim, a plaintiff must show that (1) their information was a trade secret; (2) it was misappropriated; and (3) it was used in the defendant’s business. See Learning Curve Toys, Inc. v. PlayWood Toys, Inc., 342 F.3d 714, 721 (7th Cir. 2003) (citations omitted). 2 This case turns on the first element: whether the software modules

2 While this case involves the DTSA, the parties agree that we may

rely on caselaw applying identical state trade secrets statutes (e.g., Illinois, Wisconsin, and California), which, like the DTSA, are based on the Uni- form Trade Secrets Act. See generally Mickey's Linen v. Fischer, No. 17 C 2154, 2017 WL 3970593, at *8 n.1 (N.D. Ill. Sept. 8, 2017) (collecting cases discussing the overlap between the DTSA and various other state statutes based on the UTSA).

No. 24-1377 7

NEXT identified in the FAST Tool constitute protectable trade secrets. Under the DTSA, information qualifies as a “trade se- cret” if (1) the owner has taken “reasonable measures” to keep it secret, and (2) the information derives independent value from the fact that it is not generally known and not readily ascertainable. See 18 U.S.C. § 1839(3); Life Spine, Inc. v. Aegis Spine, Inc., 8 F.4th 531, 540 (7th Cir. 2021). The parties dispute whether NEXT has described its trade secrets with enough specificity to support a DTSA claim. Our “[c]ase law requires a high level of specificity when a plaintiff makes a claim for misappropriation of a trade secret.” REXA, 42 F.4th at 663 (citations omitted). “[A] plaintiff must show ‘concrete secrets’ rather than ‘broad areas of technology.’” Id. at 662 (quoting Life Spine, 8 F.4th at 540). Whether the plaintiff has provided the requisite level of detail depends on the cir- cumstances of each case. At the least, a plaintiff must present enough specifics for the fact finder to distinguish between in- formation that is generally known and information that is not readily ascertainable and thus qualifies as a statutory secret. See IDX Sys. Corp. v. Epic Sys. Corp., 285 F.3d 581, 584 (7th Cir. 2002) (“[A] plaintiff must do more than just identify a kind of technology and then invite the court to hunt through the de- tails in search of items meeting the statutory definition.” (cit- ing Composite Marine Propellers, Inc v. Van Der Woude, 962 F.2d 1263, 1266 (7th Cir. 1992))). To shoulder this burden, NEXT points to its list of thirty- four software modules and five combination modules of its FAST Tool program. For each module, NEXT provides a title and description of what the feature does, and then it purports to identify the secret of its software program that makes each module work. But NEXT’s descriptions face a fundamental

8 No. 24-1377

problem: NEXT only ever tells us the end result of what its software does, not how it does it. See, e.g., Silvaco Data Sys. v. Intel Corp., 184 Cal. App. 4th 210, 221–22 (Cal. App. Ct. 2010) (the “finished (compiled) product” of what a software pro- gram does is not a trade secret because it is “evident to anyone running the finished program”), disapproved of on other grounds by Kwikset Corp. v. Superior Ct., 51 Cal. 4th 310, 337 (Cal. 2011). NEXT does not identify any specific algorithms, source code, or methodologies underlying the FAST Tool’s functionality. Instead, NEXT defines its modules in vague and generic lan- guage that describes the software’s function. This lack of de- tail makes it impossible to distinguish between the aspects of the FAST Tool that are generally known and ascertainable, and those which NEXT contends are secret and derive value from being kept as such. See IDX Sys., 285 F.3d at 583–84 (soft- ware developer’s tender of a “a 43-page description of the methods and processes underlying and the inter-relation- ships among various features making up IDX's software package” left it “mysterious” what it contended were trade secrets, because it did not separate its purported secrets from the “other information that goes into any software package”). Take, for example, the “Online Self Scheduling” module used as an exemplar by NEXT in its briefing. NEXT states that this module operates on the public-facing side of the FAST Tool and displays a calendar highlighting available technician appointments that a customer can select. The customer can only schedule an appointment if their pre-entered data meets certain “pre-determined criteria.” NEXT then describes what it claims is the trade secret that makes the online self-schedul- ing module work. But its description uses generic functional verbiage that simply describes the process in circular terms. It claims that the module’s “secret” process is that it “manages

No. 24-1377 9

the inventory of appointments” by interfacing with numerous other modules to “present real-time appointment availabil- ity” and “pre-set the number of available appointments based on the customer’s geographic location … [and] the availabil- ity of technicians.” In short, NEXT describes the “Online Self- Scheduling” module generically as allowing customers to schedule appointments based on existing data in the system. But how the software achieves that process is left unspecified. NEXT uses similarly nondescript language for the other FAST Tool modules. For example, the “Email & Text Messag- ing” module is described as distributing notifications and alerts to customers about appointments, but the purported “secret” behind the module is simply that it “generates auto- mated notifications … based on certain triggers.” Other mod- ules allow CLEAResult personnel to generate PDF documents and “track,” “manage,” or “see” appointments and customer data, all through the purported “secret” process of collecting and displaying data across different modules and criteria in the systems. The descriptions of the combination modules are no more specific. For example, the “FAST Tool Residential Appoint- ment Management” is said to “permit[] consumers to sched- ule, cancel, or reschedule appointments online using a public- facing web page that required customers to enter information sufficient to automatically determine eligibility or potential eligibility for a program.” None of the above descriptions articulate a concrete secret that derives economic value from being generally unknown and not readily ascertainable. It is not a trade secret for a soft- ware program to manage appointment inventory, schedule appointments, collect and display data, send notifications, or

10 No. 24-1377

fill out PDFs. Anybody using the FAST Tool application and seeing these features, either on the public-facing or internal web application, would be able to ascertain that the software was acting to perform these tasks. IDX Sys., 285 F.3d at 584 (“things that any user or passer-by sees at a glance,” such as the appearance of a data-entry screen in a billing software program, are “exceedingly hard to call trade secrets”). The fact that NEXT describes what its software does in generic terms that any user can discern means that it has failed to identify any information that might qualify for protection. Id.; see also BondPro Corp. v. Siemens Power Generation, Inc., 463 F.3d 702, 710 (7th Cir. 2006) (“[A] process described in general terms … will usually be widely known and thus not worth incurring costs to try to conceal and so not a trade secret.”). Put another way, NEXT does exactly what we have said a trade secret plaintiff cannot do: it points to broad areas of soft- ware technology and asks us to sort out what aspects of that technology may or may not meet the statutory requirements for protection. But the onus is on NEXT to identify concrete secrets. See REXA, 42 F.4th at 663; IDX Sys., 285 F.3d at 584 (citing Composite Marine, 962 F.2d at 1266). It may be that there is some unknown methodology or process that allows the FAST Tool to manage inventory, dis- play data, and create appointments. The most likely candi- date would be the FAST Tool’s underlying source code, rules, and algorithms. See generally IDX Sys., 285 F.3d at 584 (noting that the algorithms that a software program uses to perform functions may be trade secrets); Fin. Info. Techs., LLC v. iCon- trol Sys., USA, LLC, 21 F.4th 1267, 1273 (11th Cir. 2021) (“As a general matter, software source code is not readily ascertain- able and, accordingly, qualifies for trade-secret protection.”).

No. 24-1377 11

But NEXT has not produced evidence of any such secret source code or unascertainable algorithms. And it even con- ceded that CLEAResult did not have access to the FAST Tool’s code or processing engine. NEXT is thus left with nothing more than its generic descriptions of what its software does in a manner that anybody using it could ascertain. This is not enough to send its claims to a jury. See IDX Sys., 285 F.3d at 584; cf. Silvaco, 184 Cal. App. 4th at 221–22 (contrasting suc- cessful claim based on a software’s source code with a failed claim based on the “22 pages of technical verbiage” describ- ing “various features, functions, and characteristics of the de- sign and operation” of the software); Syntel Sterling Best Shores Mauritius Ltd. v. TriZetto Grp., Inc., 68 F.4th 792 (2d Cir. 2023) (plaintiff described its software trade secrets with sufficient specificity where it provided documentation identifying the underlying source code and architecture behind the claimed software trade secrets). As the district court explained, “[d]escribing the software functions without disclosing the underlying methods is like saying someone stole your top secret apple pie recipe, but never identifying the secret recipe itself.” NEXT Payment Sols., Inc. v. CLEAResult Consulting, Inc., No. 1:17-cv-08829, 2020 WL 2836778, at[*15] (N.D. Ill. May 31, 2020). NEXT’s failure to iden- tify the programs, methodology, or processes underlying its software makes it impossible for a fact finder to determine whether the functions the FAST Tool performs are protectable secrets. 3

3 To be clear, we do not hold that a plaintiff who brings a claim for misappropriation of trade secrets in their software products must always provide their underlying source code or algorithms to survive summary

12 No. 24-1377

NEXT’s counterarguments are unavailing. It first argues that its software modules are trade secrets because they are kept on the non-public side of the FAST Tool. NEXT seems to be suggesting (without any case support) that the fact that the internal web application accessed by CLEAResult was not available to the general public necessarily renders the soft- ware modules on that non-public side a secret. But we are not aware of any authority holding that “not readily ascertaina- ble” or “not generally known” means not accessible to the public writ large. NEXT sold its software to clients such as CLEAResult, and the functions performed by that software (scheduling appointments, managing inventory, etc.) would be obvious and apparent to any client that was using the in- ternal web application. That the functions are obvious to any user means they do not qualify for protection. See IDX Sys., 285 F.3d at 584; see also iControl Sys., 21 F.4th at 1273 (“aspects

judgment. In practical terms, pointing to source codes or algorithms is of- ten the most efficient way to identify a protectable secret. See iControl Sys., 21 F.4th at 1273; Silvaco, 184 Cal. App. 4th at 221–22. But we do not exclude the possibility that a plaintiff may, under some circumstances, be able to come forward with other kinds of evidence demonstrating that its soft- ware functions in a manner that is not generally ascertainable and derives value from being kept secret. See, e.g., Integrated Cash Mgmt. Servs., Inc. v. Digital Transactions, Inc., 920 F.2d 171, 174 (2d Cir. 1990) (finding a protect- able trade secret in a software package that combined various generic pro- grams where an expert testified that, although each individual program was in the public domain, the architecture of the combined end product “could not be readily duplicated without the secret information acquired by [plaintiff] through years of research.”). It is enough to say that NEXT has offered nothing from which we can identify a protectable secret here, as its generic descriptions of its software functions—the only thing it has put forward—are insufficient on their own.

No. 24-1377 13

of computer software that are readily ascertainable don't qualify” for trade secret protection). 4 Second, NEXT argues that FAST Tool modules, such as its appointment scheduling functions, performed something “new and valuable.” It asserts that they were tailored to the specific needs of CLEAResult’s utility customers and allowed those customers to be matched with technicians based on real- time appointment availability in set geographic areas. We are unpersuaded. We will take it as true that its pro- gram, in allowing utility companies to match customers with technicians to schedule appointments, does something “new and valuable”—though we are skeptical of this unsupported claim given the ubiquity of scheduling applications across a variety of industries. But regardless, something being “new and valuable” does not itself render it a trade secret. See Sil- vaco, 184 Cal. App. 4th at 222 (the finished software product “might have distinctive characteristics resulting from that de- sign—such as improved performance—[but] they cannot con- stitute trade secrets because they are not secret but are evident to anyone running the finished program”). Protectable trade secrets must be generally unknown, not readily ascertainable, and must be described with specificity. These requirements are lacking here.

4 We acknowledge CLEAResult personnel were subject to nondisclo-

sure agreements in their use of the FAST Tool web application. The exist- ence of those agreements might be relevant to misappropriation, i.e., whether CLEAResult personnel could lawfully use what they saw in the FAST Tool to make their own software. But here, the issue is not misap- propriation but whether there was a protectable trade secret at all.

14 No. 24-1377

In sum, we conclude that NEXT failed to come forward with evidence from which a jury could conclude that its FAST Tool software modules qualify as trade secrets. The district court therefore properly entered summary judgment on this claim. II. NEXT’s Unjust Enrichment Claim We turn next to the district court’s ruling in limine, which precluded NEXT from arguing a theory of unjust enrichment based on misappropriation of “other proprietary infor- mation.” We review de novo the district court’s legal conclu- sions underlying a ruling in limine, though we review its ulti- mate decision for abuse of discretion. United States v. Wade, 962 F.3d 1004, 1011 (7th Cir. 2020). The district court provided several independent grounds for its ruling, but we focus on just one here. Specifically, the court found that NEXT’s unjust enrichment theory based on misuse of “other proprietary information” was a new theory of liability not previously pursued by NEXT, and that it would prejudice CLEAResult and delay the proceedings to al- low NEXT to add the claim after summary judgment. Because we find this provided an adequate basis to grant the motion in limine, we do not discuss the court’s alternative holdings. District courts enjoy wide discretion to refuse attempts to raise new theories of liability at late stages of the case if such changes will prejudice the opposing party or unduly delay the proceedings. See Crest Hill Land Dev., LLC v. City of Joliet, 396 F.3d 801, 804 (7th Cir. 2005) (citing Fort Howard Paper Co. v. Standard Havens, Inc., 901 F.2d 1373, 1380 (7th Cir. 1990)) (“‘Surprises’ such as new arguments or defense theories prop- agated after the completion of discovery and filing of

No. 24-1377 15

summary judgment are wisely discouraged.”); Bethany Phar- macal Co. v. QVC, Inc., 241 F.3d 854, 862 (7th Cir. 2001) (“[A]llowing [plaintiff] to add [a new claim] would have re- quired additional delays in the resolution of the case to allow [defendant] to respond to a new theory of liability. We do not require a district court to tolerate such delays.”); see also Soltys v. Costello, 520 F.3d 737, 743 (7th Cir. 2008) (“Eleventh hour additions ... [are] bound to produce delays that burden not only the parties to the litigation but also the judicial system and other litigants.” (citation omitted)). The district court was firmly within its discretion in holding that NEXT was improp- erly raising a new theory of liability shortly before trial and that allowing it to do so would cause undue delays and prej- udice. NEXT notably does not make any argument regarding de- lay or prejudice but simply asserts that its theory was not new at all. It argues it pleaded this theory in the operative com- plaint, which alleged that CLEAResult had been “unjustly en- riched by having access to NEXT trade secrets and other pro- prietary information” and that CLEAResult “misused NEXT’s trade secrets and proprietary information to enable persons and entities with no right to have access to NEXT’s trade secrets and proprietary information to create competitive sys- tems…that [CLEAResult] subsequently used to service [its] customers.” (emphasis added). As NEXT sees it, these allega- tions create two distinct theories of unjust enrichment: one based on misuse of trade secrets, one based on misuse of “other proprietary information.” NEXT thus maintains that it always had a separate claim for unjust enrichment based on misuse of proprietary information, and that it should have been allowed to proceed to trial on that theory irrespective of whether it could maintain a trade secret claim.

16 No. 24-1377

We are not convinced. NEXT’s argument misrepresents how its amended complaint reads as a whole. Cf. Scott v. City of Chicago, 195 F.3d 950, 952 (7th Cir. 1999) (“Whether a com- plaint provides notice [of a claim], however, is determined by looking at the complaint as a whole.”). NEXT’s complaint never identifies any “proprietary information” that is sepa- rate and distinct from its claimed trade secrets. The term “pro- prietary information” only appears four times, and three times it is used in the unjust enrichment claim in conjunction with trade secrets in the phrase “trade secrets and proprietary information.” The other time NEXT refers to its “proprietary information,” it does so in the context of describing its “NEXT System Back End,” which is the software system that NEXT identifies throughout the complaint as its protectable trade se- cret. In other words, when the complaint refers to “proprie- tary information,” it always does so in tandem with its trade secrets. It never distinguishes “other proprietary infor- mation” as representing some category of information sepa- rate and apart from its trade secrets. There is thus no indica- tion in the complaint that NEXT intended a free-standing claim based on misuse of proprietary information apart from its trade secrets. And if NEXT intended to plead an independent unjust en- richment claim based on some proprietary information dis- tinct from its trade secrets, it never made that position known in the lengthy history of this case. The district court’s in limine ruling provides that history in exacting detail, and we need not recite it all again. See NEXT Payment Sols., Inc. v. CLEARe- sult Consulting, Inc., No. 1:17-cv-08829, 2023 WL 7196125, at *2–*13. (N.D. Ill. Apr. 17, 2023). There were multiple occasions during discovery and the summary judgment proceedings where NEXT argued (expressly or implicitly) that the scope

No. 24-1377 17

of its unjust enrichment claim included, at most, two theories: one based on unpaid invoices, the other based on misappro- priation of trade secrets. The district court itself confirmed this understanding in its first summary judgment ruling, de- scribing NEXT’s unjust enrichment claims as seeking “redress for Defendant’s misappropriation of trade secrets related to the FAST Tool and failure to pay Plaintiff’s invoices for use of the FAST Tool.” It was only after the court granted summary judgment and dismissed NEXT’s trade secrets claim that it sought to invent a new theory based on some other “propriety information.” In sum, the district court reasonably concluded that NEXT was attempting to raise a new and undeveloped theory of li- ability based on other proprietary information that it had never previously presented nor even identified. It was within its discretion to exclude that new theory of liability on the grounds that it would prejudice CLEAResult and delay the trial. The court did not abuse its discretion in precluding NEXT from arguing this theory at trial. CONCLUSION For the foregoing reasons, we AFFIRM the judgment of the district court.

MALDONADO, Circuit Judge. CLEAResult Consulting, Inc. hired NEXT Payment Solutions, Inc. to develop appointment scheduling software. After a few years of successful partner- ship and use of NEXT’s application, CLEAResult decided to shift to different scheduling software. Before making the tran- sition, and without telling NEXT, CLEAResult modified and refined its new program based on the functionality and

2 No. 24-1377

features of NEXT’s application. CLEAResult then terminated its relationship with NEXT and stopped using its scheduling tool. NEXT sued CLEAResult for misappropriation of trade se- crets under federal law and unjust enrichment under Illinois common law. 1 After years of contentious litigation, the dis- trict court entered summary judgment against NEXT on the trade secrets claim because NEXT failed to define its secrets with enough specificity. The case moved to pretrial proceed- ings on the unjust enrichment claim, and the district court granted a motion in limine limiting the scope of the claim. NEXT then voluntarily dismissed the claim, and the district court entered final judgment in favor of CLEAResult. We af- firm. BACKGROUND I. Factual Background NEXT designs and develops customer service software for businesses. NEXT’s primary software platform is appropri- ately called the “NEXT System.” NEXT offers customizable versions of the NEXT System for its customers’ business needs. CLEAResult provides North American utilities with en- ergy efficiency programs and services for utility customers. As part of its services, CLEAResult offers utility customers the opportunity to schedule in-home appointments for services like home energy-efficiency assessments.

1 NEXT asserted several other common law claims that are not rele- vant to this appeal.

No. 24-1377 3

In 2014, CLEAResult hired NEXT to develop an online- scheduling tool. NEXT worked with CLEAResult to under- stand its needs and created a customized software application for booking appointments called the “FAST Tool.” The FAST Tool contained a public-facing side, where cus- tomers could access their accounts and schedule appoint- ments, and an internal side, where CLEAResult personnel could manage appointments and access customer infor- mation and data. There were also underlying software me- chanics, including rules, algorithms, and source code to which only NEXT had access. CLEAResult used the FAST Tool for several years without issue. Then in early 2017, CLEAResult acquired a separate green technology company that had its own cloud-based ap- plication, known as the DSMTracker. CLEAResult decided to replace the FAST Tool with that new platform. According to CLEAResult, it wanted to analyze how the DSMTracker needed to be refined to meet its functionality needs. So CLEAResult launched “Project Renaissance,” an internal analysis of the gaps in the DSMTracker. As part of that effort, CLEAResult developers looked at the internal-facing side of the FAST Tool—the feature used by customer service person- nel to manage appointments and customer data—to ensure that the DSMTracker covered the same functionality. NEXT presents a different picture of Project Renaissance. According to NEXT, CLEAResult surreptitiously, and with- out NEXT’s approval, stole secret information from the FAST Tool to reverse engineer and replicate its features into the DSMTracker. Notably, however, NEXT concedes that CLEAResult personnel never had access to the underlying processing engine or rules software of the FAST Tool. They

4 No. 24-1377

further agree that there is no evidence CLEAResult ever ac- cessed the FAST Tool’s underlying source code. In November 2017, after CLEAResult finished Project Re- naissance and transitioned several customers to the DSMTracker, it informed NEXT it was ending their relation- ship. II. Procedural History Shortly thereafter, NEXT initiated this lawsuit asserting, among other things, claims for misappropriation of trade se- crets under the Defend Trade Secrets Act, 18 U.S.C. §§ 1836 et seq. (“DTSA”), and unjust enrichment under Illinois common law. NEXT alleged that CLEAResult violated the DTSA by misappropriating trade secret information from the FAST Tool to better the DSMTracker’s functionality. As to unjust enrichment, NEXT pled two separate theories of recovery: that CLEAResult refused to pay certain invoices, and that CLEAResult had “access to NEXT trade secrets and other pro- prietary information” and was unjustly enriched when it “misused NEXT’s trade secrets and proprietary information … to create competitive systems.” Following discovery, CLEAResult twice moved for sum- mary judgment on NEXT’s DTSA claim arguing that NEXT failed to identify its software secrets with sufficient specific- ity. In its first summary judgment ruling, the district court agreed with CLEAResult in part that NEXT’s secrets were too nonspecific and narrowed the DTSA claim to “those parts of the FAST Tool that were not present in the DSMTracker be- fore Defendant transitioned” to its new program. Later, the court ordered NEXT to provide more infor- mation on the features of the FAST Tool that it contended

No. 24-1377 5

were misappropriated and to explain why they were trade se- crets. NEXT produced a spreadsheet identifying a list of thirty-four software “modules,” and five “combinations of modules and features” that it said CLEAResult took from the FAST Tool. The modules—discussed in more detail below— had titles such as “Online Self Scheduling” and “Dashboard Client.” NEXT describes the modules with generic language such as “manages the inventory of appointments,” “pre- sent[s] real-time appointment availability,” and “create[s] open appointment slots, to update appointment availability.” CLEAResult, unsatisfied with NEXT’s updated descrip- tions, again moved for summary judgment. This time, the dis- trict court granted the motion in full. The court explained that NEXT had been given multiple opportunities to identify “spe- cific and concrete” misappropriated secrets in its software but came forward with nothing more than “broad descriptions and jargon-laden terminology” describing what its software did, not how it did it, which made it “hazy what, exactly, NEXT is claiming as its alleged trade secrets.” More than a year later—during pretrial proceedings on NEXT’s remaining claims—a dispute arose over the scope of NEXT’s unjust enrichment claim. In a pretrial status report, NEXT asserted for the first time that it intended to bring to trial an independent claim for unjust enrichment based on the theory that CLEAResult misused “other proprietary infor- mation” separately and apart from misusing NEXT’s trade se- crets. On a motion in limine from CLEAResult, however, the dis- trict court excluded any evidence or argument of misuse of “other proprietary information” as a basis for unjust enrich- ment. The court found that this theory of unjust enrichment

6 No. 24-1377

had never before been presented as a separate theory from misappropriation of trade secrets. The court held such a late change was improper and that NEXT was, in essence, trying to put forth a new theory on the eve of trial as an end-run around the dismissal of its DTSA claim. After the ruling in limine, NEXT voluntarily dismissed its remaining claims with prejudice. It appeals the district court’s dismissal of its DTSA claim and the in limine order on its claim for unjust enrichment. DISCUSSION I. Summary Judgment on NEXT’s DTSA Claim We review de novo the district court’s grant of summary judgment. REXA, Inc. v. Chester, 42 F.4th 652, 662 (7th Cir. 2022). Summary judgment is appropriate when, drawing all reasonable inferences in the nonmovant’s favor, there is no genuine dispute of material fact, so the moving party is enti- tled to judgment as a matter of law. See id. To prevail on a DTSA misappropriation claim, a plaintiff must show that (1) their information was a trade secret; (2) it was misappropriated; and (3) it was used in the defendant’s business. See Learning Curve Toys, Inc. v. PlayWood Toys, Inc., 342 F.3d 714, 721 (7th Cir. 2003) (citations omitted). 2 This case turns on the first element: whether the software modules

2 While this case involves the DTSA, the parties agree that we may

rely on caselaw applying identical state trade secrets statutes (e.g., Illinois, Wisconsin, and California), which, like the DTSA, are based on the Uni- form Trade Secrets Act. See generally Mickey's Linen v. Fischer, No. 17 C 2154, 2017 WL 3970593, at *8 n.1 (N.D. Ill. Sept. 8, 2017) (collecting cases discussing the overlap between the DTSA and various other state statutes based on the UTSA).

No. 24-1377 7

NEXT identified in the FAST Tool constitute protectable trade secrets. Under the DTSA, information qualifies as a “trade se- cret” if (1) the owner has taken “reasonable measures” to keep it secret, and (2) the information derives independent value from the fact that it is not generally known and not readily ascertainable. See 18 U.S.C. § 1839(3); Life Spine, Inc. v. Aegis Spine, Inc., 8 F.4th 531, 540 (7th Cir. 2021). The parties dispute whether NEXT has described its trade secrets with enough specificity to support a DTSA claim. Our “[c]ase law requires a high level of specificity when a plaintiff makes a claim for misappropriation of a trade secret.” REXA, 42 F.4th at 663 (citations omitted). “[A] plaintiff must show ‘concrete secrets’ rather than ‘broad areas of technology.’” Id. at 662 (quoting Life Spine, 8 F.4th at 540). Whether the plaintiff has provided the requisite level of detail depends on the cir- cumstances of each case. At the least, a plaintiff must present enough specifics for the fact finder to distinguish between in- formation that is generally known and information that is not readily ascertainable and thus qualifies as a statutory secret. See IDX Sys. Corp. v. Epic Sys. Corp., 285 F.3d 581, 584 (7th Cir. 2002) (“[A] plaintiff must do more than just identify a kind of technology and then invite the court to hunt through the de- tails in search of items meeting the statutory definition.” (cit- ing Composite Marine Propellers, Inc v. Van Der Woude, 962 F.2d 1263, 1266 (7th Cir. 1992))). To shoulder this burden, NEXT points to its list of thirty- four software modules and five combination modules of its FAST Tool program. For each module, NEXT provides a title and description of what the feature does, and then it purports to identify the secret of its software program that makes each module work. But NEXT’s descriptions face a fundamental

8 No. 24-1377

problem: NEXT only ever tells us the end result of what its software does, not how it does it. See, e.g., Silvaco Data Sys. v. Intel Corp., 184 Cal. App. 4th 210, 221–22 (Cal. App. Ct. 2010) (the “finished (compiled) product” of what a software pro- gram does is not a trade secret because it is “evident to anyone running the finished program”), disapproved of on other grounds by Kwikset Corp. v. Superior Ct., 51 Cal. 4th 310, 337 (Cal. 2011). NEXT does not identify any specific algorithms, source code, or methodologies underlying the FAST Tool’s functionality. Instead, NEXT defines its modules in vague and generic lan- guage that describes the software’s function. This lack of de- tail makes it impossible to distinguish between the aspects of the FAST Tool that are generally known and ascertainable, and those which NEXT contends are secret and derive value from being kept as such. See IDX Sys., 285 F.3d at 583–84 (soft- ware developer’s tender of a “a 43-page description of the methods and processes underlying and the inter-relation- ships among various features making up IDX's software package” left it “mysterious” what it contended were trade secrets, because it did not separate its purported secrets from the “other information that goes into any software package”). Take, for example, the “Online Self Scheduling” module used as an exemplar by NEXT in its briefing. NEXT states that this module operates on the public-facing side of the FAST Tool and displays a calendar highlighting available technician appointments that a customer can select. The customer can only schedule an appointment if their pre-entered data meets certain “pre-determined criteria.” NEXT then describes what it claims is the trade secret that makes the online self-schedul- ing module work. But its description uses generic functional verbiage that simply describes the process in circular terms. It claims that the module’s “secret” process is that it “manages

No. 24-1377 9

the inventory of appointments” by interfacing with numerous other modules to “present real-time appointment availabil- ity” and “pre-set the number of available appointments based on the customer’s geographic location … [and] the availabil- ity of technicians.” In short, NEXT describes the “Online Self- Scheduling” module generically as allowing customers to schedule appointments based on existing data in the system. But how the software achieves that process is left unspecified. NEXT uses similarly nondescript language for the other FAST Tool modules. For example, the “Email & Text Messag- ing” module is described as distributing notifications and alerts to customers about appointments, but the purported “secret” behind the module is simply that it “generates auto- mated notifications … based on certain triggers.” Other mod- ules allow CLEAResult personnel to generate PDF documents and “track,” “manage,” or “see” appointments and customer data, all through the purported “secret” process of collecting and displaying data across different modules and criteria in the systems. The descriptions of the combination modules are no more specific. For example, the “FAST Tool Residential Appoint- ment Management” is said to “permit[] consumers to sched- ule, cancel, or reschedule appointments online using a public- facing web page that required customers to enter information sufficient to automatically determine eligibility or potential eligibility for a program.” None of the above descriptions articulate a concrete secret that derives economic value from being generally unknown and not readily ascertainable. It is not a trade secret for a soft- ware program to manage appointment inventory, schedule appointments, collect and display data, send notifications, or

10 No. 24-1377

fill out PDFs. Anybody using the FAST Tool application and seeing these features, either on the public-facing or internal web application, would be able to ascertain that the software was acting to perform these tasks. IDX Sys., 285 F.3d at 584 (“things that any user or passer-by sees at a glance,” such as the appearance of a data-entry screen in a billing software program, are “exceedingly hard to call trade secrets”). The fact that NEXT describes what its software does in generic terms that any user can discern means that it has failed to identify any information that might qualify for protection. Id.; see also BondPro Corp. v. Siemens Power Generation, Inc., 463 F.3d 702, 710 (7th Cir. 2006) (“[A] process described in general terms … will usually be widely known and thus not worth incurring costs to try to conceal and so not a trade secret.”). Put another way, NEXT does exactly what we have said a trade secret plaintiff cannot do: it points to broad areas of soft- ware technology and asks us to sort out what aspects of that technology may or may not meet the statutory requirements for protection. But the onus is on NEXT to identify concrete secrets. See REXA, 42 F.4th at 663; IDX Sys., 285 F.3d at 584 (citing Composite Marine, 962 F.2d at 1266). It may be that there is some unknown methodology or process that allows the FAST Tool to manage inventory, dis- play data, and create appointments. The most likely candi- date would be the FAST Tool’s underlying source code, rules, and algorithms. See generally IDX Sys., 285 F.3d at 584 (noting that the algorithms that a software program uses to perform functions may be trade secrets); Fin. Info. Techs., LLC v. iCon- trol Sys., USA, LLC, 21 F.4th 1267, 1273 (11th Cir. 2021) (“As a general matter, software source code is not readily ascertain- able and, accordingly, qualifies for trade-secret protection.”).

No. 24-1377 11

But NEXT has not produced evidence of any such secret source code or unascertainable algorithms. And it even con- ceded that CLEAResult did not have access to the FAST Tool’s code or processing engine. NEXT is thus left with nothing more than its generic descriptions of what its software does in a manner that anybody using it could ascertain. This is not enough to send its claims to a jury. See IDX Sys., 285 F.3d at 584; cf. Silvaco, 184 Cal. App. 4th at 221–22 (contrasting suc- cessful claim based on a software’s source code with a failed claim based on the “22 pages of technical verbiage” describ- ing “various features, functions, and characteristics of the de- sign and operation” of the software); Syntel Sterling Best Shores Mauritius Ltd. v. TriZetto Grp., Inc., 68 F.4th 792 (2d Cir. 2023) (plaintiff described its software trade secrets with sufficient specificity where it provided documentation identifying the underlying source code and architecture behind the claimed software trade secrets). As the district court explained, “[d]escribing the software functions without disclosing the underlying methods is like saying someone stole your top secret apple pie recipe, but never identifying the secret recipe itself.” NEXT Payment Sols., Inc. v. CLEAResult Consulting, Inc., No. 1:17-cv-08829, 2020 WL 2836778, at[*15] (N.D. Ill. May 31, 2020). NEXT’s failure to iden- tify the programs, methodology, or processes underlying its software makes it impossible for a fact finder to determine whether the functions the FAST Tool performs are protectable secrets. 3

3 To be clear, we do not hold that a plaintiff who brings a claim for misappropriation of trade secrets in their software products must always provide their underlying source code or algorithms to survive summary

12 No. 24-1377

NEXT’s counterarguments are unavailing. It first argues that its software modules are trade secrets because they are kept on the non-public side of the FAST Tool. NEXT seems to be suggesting (without any case support) that the fact that the internal web application accessed by CLEAResult was not available to the general public necessarily renders the soft- ware modules on that non-public side a secret. But we are not aware of any authority holding that “not readily ascertaina- ble” or “not generally known” means not accessible to the public writ large. NEXT sold its software to clients such as CLEAResult, and the functions performed by that software (scheduling appointments, managing inventory, etc.) would be obvious and apparent to any client that was using the in- ternal web application. That the functions are obvious to any user means they do not qualify for protection. See IDX Sys., 285 F.3d at 584; see also iControl Sys., 21 F.4th at 1273 (“aspects

judgment. In practical terms, pointing to source codes or algorithms is of- ten the most efficient way to identify a protectable secret. See iControl Sys., 21 F.4th at 1273; Silvaco, 184 Cal. App. 4th at 221–22. But we do not exclude the possibility that a plaintiff may, under some circumstances, be able to come forward with other kinds of evidence demonstrating that its soft- ware functions in a manner that is not generally ascertainable and derives value from being kept secret. See, e.g., Integrated Cash Mgmt. Servs., Inc. v. Digital Transactions, Inc., 920 F.2d 171, 174 (2d Cir. 1990) (finding a protect- able trade secret in a software package that combined various generic pro- grams where an expert testified that, although each individual program was in the public domain, the architecture of the combined end product “could not be readily duplicated without the secret information acquired by [plaintiff] through years of research.”). It is enough to say that NEXT has offered nothing from which we can identify a protectable secret here, as its generic descriptions of its software functions—the only thing it has put forward—are insufficient on their own.

No. 24-1377 13

of computer software that are readily ascertainable don't qualify” for trade secret protection). 4 Second, NEXT argues that FAST Tool modules, such as its appointment scheduling functions, performed something “new and valuable.” It asserts that they were tailored to the specific needs of CLEAResult’s utility customers and allowed those customers to be matched with technicians based on real- time appointment availability in set geographic areas. We are unpersuaded. We will take it as true that its pro- gram, in allowing utility companies to match customers with technicians to schedule appointments, does something “new and valuable”—though we are skeptical of this unsupported claim given the ubiquity of scheduling applications across a variety of industries. But regardless, something being “new and valuable” does not itself render it a trade secret. See Sil- vaco, 184 Cal. App. 4th at 222 (the finished software product “might have distinctive characteristics resulting from that de- sign—such as improved performance—[but] they cannot con- stitute trade secrets because they are not secret but are evident to anyone running the finished program”). Protectable trade secrets must be generally unknown, not readily ascertainable, and must be described with specificity. These requirements are lacking here.

4 We acknowledge CLEAResult personnel were subject to nondisclo-

sure agreements in their use of the FAST Tool web application. The exist- ence of those agreements might be relevant to misappropriation, i.e., whether CLEAResult personnel could lawfully use what they saw in the FAST Tool to make their own software. But here, the issue is not misap- propriation but whether there was a protectable trade secret at all.

14 No. 24-1377

In sum, we conclude that NEXT failed to come forward with evidence from which a jury could conclude that its FAST Tool software modules qualify as trade secrets. The district court therefore properly entered summary judgment on this claim. II. NEXT’s Unjust Enrichment Claim We turn next to the district court’s ruling in limine, which precluded NEXT from arguing a theory of unjust enrichment based on misappropriation of “other proprietary infor- mation.” We review de novo the district court’s legal conclu- sions underlying a ruling in limine, though we review its ulti- mate decision for abuse of discretion. United States v. Wade, 962 F.3d 1004, 1011 (7th Cir. 2020). The district court provided several independent grounds for its ruling, but we focus on just one here. Specifically, the court found that NEXT’s unjust enrichment theory based on misuse of “other proprietary information” was a new theory of liability not previously pursued by NEXT, and that it would prejudice CLEAResult and delay the proceedings to al- low NEXT to add the claim after summary judgment. Because we find this provided an adequate basis to grant the motion in limine, we do not discuss the court’s alternative holdings. District courts enjoy wide discretion to refuse attempts to raise new theories of liability at late stages of the case if such changes will prejudice the opposing party or unduly delay the proceedings. See Crest Hill Land Dev., LLC v. City of Joliet, 396 F.3d 801, 804 (7th Cir. 2005) (citing Fort Howard Paper Co. v. Standard Havens, Inc., 901 F.2d 1373, 1380 (7th Cir. 1990)) (“‘Surprises’ such as new arguments or defense theories prop- agated after the completion of discovery and filing of

No. 24-1377 15

summary judgment are wisely discouraged.”); Bethany Phar- macal Co. v. QVC, Inc., 241 F.3d 854, 862 (7th Cir. 2001) (“[A]llowing [plaintiff] to add [a new claim] would have re- quired additional delays in the resolution of the case to allow [defendant] to respond to a new theory of liability. We do not require a district court to tolerate such delays.”); see also Soltys v. Costello, 520 F.3d 737, 743 (7th Cir. 2008) (“Eleventh hour additions ... [are] bound to produce delays that burden not only the parties to the litigation but also the judicial system and other litigants.” (citation omitted)). The district court was firmly within its discretion in holding that NEXT was improp- erly raising a new theory of liability shortly before trial and that allowing it to do so would cause undue delays and prej- udice. NEXT notably does not make any argument regarding de- lay or prejudice but simply asserts that its theory was not new at all. It argues it pleaded this theory in the operative com- plaint, which alleged that CLEAResult had been “unjustly en- riched by having access to NEXT trade secrets and other pro- prietary information” and that CLEAResult “misused NEXT’s trade secrets and proprietary information to enable persons and entities with no right to have access to NEXT’s trade secrets and proprietary information to create competitive sys- tems…that [CLEAResult] subsequently used to service [its] customers.” (emphasis added). As NEXT sees it, these allega- tions create two distinct theories of unjust enrichment: one based on misuse of trade secrets, one based on misuse of “other proprietary information.” NEXT thus maintains that it always had a separate claim for unjust enrichment based on misuse of proprietary information, and that it should have been allowed to proceed to trial on that theory irrespective of whether it could maintain a trade secret claim.

16 No. 24-1377

We are not convinced. NEXT’s argument misrepresents how its amended complaint reads as a whole. Cf. Scott v. City of Chicago, 195 F.3d 950, 952 (7th Cir. 1999) (“Whether a com- plaint provides notice [of a claim], however, is determined by looking at the complaint as a whole.”). NEXT’s complaint never identifies any “proprietary information” that is sepa- rate and distinct from its claimed trade secrets. The term “pro- prietary information” only appears four times, and three times it is used in the unjust enrichment claim in conjunction with trade secrets in the phrase “trade secrets and proprietary information.” The other time NEXT refers to its “proprietary information,” it does so in the context of describing its “NEXT System Back End,” which is the software system that NEXT identifies throughout the complaint as its protectable trade se- cret. In other words, when the complaint refers to “proprie- tary information,” it always does so in tandem with its trade secrets. It never distinguishes “other proprietary infor- mation” as representing some category of information sepa- rate and apart from its trade secrets. There is thus no indica- tion in the complaint that NEXT intended a free-standing claim based on misuse of proprietary information apart from its trade secrets. And if NEXT intended to plead an independent unjust en- richment claim based on some proprietary information dis- tinct from its trade secrets, it never made that position known in the lengthy history of this case. The district court’s in limine ruling provides that history in exacting detail, and we need not recite it all again. See NEXT Payment Sols., Inc. v. CLEARe- sult Consulting, Inc., No. 1:17-cv-08829, 2023 WL 7196125, at *2–*13. (N.D. Ill. Apr. 17, 2023). There were multiple occasions during discovery and the summary judgment proceedings where NEXT argued (expressly or implicitly) that the scope

No. 24-1377 17

of its unjust enrichment claim included, at most, two theories: one based on unpaid invoices, the other based on misappro- priation of trade secrets. The district court itself confirmed this understanding in its first summary judgment ruling, de- scribing NEXT’s unjust enrichment claims as seeking “redress for Defendant’s misappropriation of trade secrets related to the FAST Tool and failure to pay Plaintiff’s invoices for use of the FAST Tool.” It was only after the court granted summary judgment and dismissed NEXT’s trade secrets claim that it sought to invent a new theory based on some other “propriety information.” In sum, the district court reasonably concluded that NEXT was attempting to raise a new and undeveloped theory of li- ability based on other proprietary information that it had never previously presented nor even identified. It was within its discretion to exclude that new theory of liability on the grounds that it would prejudice CLEAResult and delay the trial. The court did not abuse its discretion in precluding NEXT from arguing this theory at trial. CONCLUSION For the foregoing reasons, we AFFIRM the judgment of the district court.