GRIFFIN, Circuit Judge.
Plaintiff- Jeanne Wallner, a twenty-seven-year employee who was fired by defendant Hilliard Lyons nine days after returning to work from taking leave under the Family and Medical Leave Act (“FMLA”), appeals the district court’s entry of summary judgment against her FMLA interference and retaliation claims. We affirm the entry of summary judgment in defendant’s favor on Wallner’s FMLA interference claim because she has neither alleged nor demonstrated that she was denied a statutory benefit to which she was entitled. But, because we conclude that a reasonable jury could find that Wallner’s exercise of her FMLA rights was a motivating factor for defendant’s decision to terminate her, we reverse the entry of judgment against her FMLA retaliation claim and remand for further proceedings.
I.
Wallner was hired by Hilliard Lyons, a financial brokerage firm, in 1982 and worked as an options trader in its three-person options department for twenty-seven years. In 2009, her supervisor was Denny Moorman, while Cassondra Dunning — a trader like Wallner — rounded out the department.
In January 2009, Wallner was preparing to leave for vacation but learned on the evening before her planned departure that her domestic flight had been rescheduled to leave the next day at 7:00 a.m. rather than at 7:00 p.m. Wallner called Moorman that evening to request the next day off of work due to the rescheduled flight, and he granted her request. When she returned from vacation, however, Moorman, who had earlier communicated to human resources his concerns that Wallner habitually arrived late to work, handed Wallner a written warning for “chronic tardiness and abuse of unscheduled absences.” The “unscheduled absence” to which the warning referred was the day that Moorman had given her permission to take off. According to Moorman, each workday ended at 5:00 p.m., meaning that her call in the evening was the “next day as far as I’m concerned, which would be [an unscheduled] same day [request for time off].” Despite the warning’s additional allegation of habitual tardiness, Moorman had never mentioned chronic tardiness as a deficien[548]*548cy in Wallner’s previous performance appraisals.
In June 2009, Wallner requested FMLA leave in order to have knee-replacement surgery. Her physician completed the pertinent paperwork, estimating that Wall-ner would be unable to work for two months: from August 11 until “approximately]” October 11, 2009. As Hilli-ard Lyons concedes, Wallner’s FMLA leave was approved through the end date noted by Wallner’s physician.
Wallner underwent her scheduled surgery and began taking FMLA leave while recuperating. Wallner’s short-term disability insurance carrier, however, had calculated Wallner’s return-to-work date as September 22, 2009. On September 14 and 15, Sharon Landgraf, a human resources manager at Hilliard Lyons, telephoned Wallner and informed her that she was expected to return to work on September 22. Wallner, who was still convalescing from surgery, tried to explain to Landgraf the difference between short-term disability and FMLA leave and that she could not return to work until her doctor cleared her to do so. Landgraf nevertheless insisted that Wallner needed to return on September 22. Courtesies degenerated, and Wallner’s husband, who had been listening to the exchange in the background, began shouting profanities that Landgraf could hear over the phone.
After conversing with Landgraf, Wallner contacted her short-term disability carrier and was able to extend her coverage through October 1. Upon receiving the update, Landgraf sent Wallner a letter that again erroneously conflated short-term disability leave with FMLA leave, informing Wallner that her FMLA leave had been extended through October 1, 2009, with a return-to-work date of October 2.
Because her FMLA leave was not coterminous with her short-term disability coverage, Wallner — who did not receive Land-graf s letter until after October 2 — was not required to return to work on the day that her disability coverage expired and did not do so. Her next physician’s appointment was on Monday, October 5, at which her doctor cleared her to return to work— almost a week earlier than he had initially estimated would be possible. Immediately upon receiving clearance from her doctor to return to work, Wallner contacted Hilli-ard Lyons with the news that she would be back at work the next day.
During Wallner’s absence on FMLA leave, Landgraf had twice informed Moor-man that Wallner would be returning to work when she actually would not — first on September 22, then on October 2. Landgraf does not recall ever informing Moorman that she was incorrect about Wallner’s return-to-work date, and Moor-man does not remember being told that Landgraf had been mistaken. In fact, Moorman had been “suspicious” that Wall-ner would not show up on September 22, and when she did not, he felt that his suspicions had been confirmed. “[T]he dates were changing,” and Moorman was simply going to “wait until she got back” before he “believed” that she would actually show up on any given date. In the meantime, running a three-person department with only two people was difficult for Moorman, who later indicated that the department was so busy that he could not leave it during Wallner’s FMLA absence to attend meetings in other locations or even, at times, to take lunch or bathroom breaks.
On October 6 — the day of Wallner’s return from FMLA leave-three things happened. First, she was reinstated to her options trader position. Second, she was issued a final written warning for her behavior during the September 15 phone call with Landgraf and for purportedly failing [549]*549to stay in contact with Hilliard Lyons af-terwards. The warning was styled a “final” written warning despite the fact that Wallner had not received any intermediate warning for her behavior, as was required by company policy. And third, Wallner’s colleague, Cassondra Dunning, began making informal records of when Wallner arrived for work each morning.
According to Dunning’s records — which she claims that she gave to Landgraf at some point but cannot recall whether it was before or after Wallner’s termination — Wallner arrived a few minutes after 8:00 a.m. on five days following her return to work: specifically, at 8:05, 8:02, 8:06, 8:09, and 8:05 a.m. Moorman himself did not keep any documentation of Wall-ner’s arrival times, despite the fact that Landgraf had instructed him to do so if her tardiness was a problem. When Wall-ner arrived at 8:05 a.m. on October 15 (video footage later established that she had actually arrived at 8:02:52), Moorman decided that she should be fired. Moor-man later testified that “to be five minutes late is not necessarily a swing factor,” but that “five to 15 minutes late consistently when it affects other employees is a major problem.”
Moorman and several human resources personnel, including Landgraf, terminated Wallner on the afternoon of October 15, nine days after she returned from FMLA leave. Moorman later compiled a “document of deficiencies” that he indicated “were basically the reasons” for Wallner’s termination. It read:
It was company policy to be at work at 8:00 AM. [Wallner] was persistently late (5-15 minutes) even in view of repeated warnings. This created a morale problem within the department.
She would occasionally call in the morning before work to inform us that she would be absent from work that day. Unscheduled absences were not permit[t]ed.
She failed two (2) Series 4 (Registered Option Principal) exams. It wasn’t required that she take the exams ...
She had a hip replacement. I believe there was a standard allotment for recuperation of four (4) weeks. However, in case it was required an additional two (2) weeks could be allowed in ease it was necessary. There was never any communication from her, to my knowledge, as to when she would return to work. In a three[-] person department, an absence involved 33% of the work force. Options trading and surveillance are specialized and critical activities within the daily operations of the department and can’t be adequately replaced by a worker (next door).
An accumulation of deficiencies became intolerable.
Moorman later indicated that the “accumulation of deficiencies” referenced at the document’s conclusion referred to “the ones that are listed in [the] document.”
Wallner eventually filed suit against Hil-liard Lyons in the district court, asserting claims for interference and retaliation under the FMLA, proceeding on the latter claim on a single-motive theory or, in the alternative, a mixed-motive theory. Hilli-ard Lyons moved for summary judgment. The district court, construing Wallner’s FMLA interference claim as an FMLA retaliation claim, analyzed them together and ultimately granted Hilliard Lyons’s motion in full. Wallner appeals.
II.
Wallner contends on appeal that the district court erred in granting summary judgment against her FMLA interference and retaliation claims. We review the district court’s summary judgment de[550]*550termination de novo. Thomas M. Cooley Law Sch. v. Kurzon Strauss, LLP, 759 F.3d 522, 526 (6th Cir.2014). Summary judgment is appropriate only if “the mov-ant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(a). A factual issue is genuinely in dispute if a reasonable fact-finder could resolve it either way. See Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986). Where the disputed issue of fact is material to liability, therefore, premature entry of summary judgment inappropriately supplants the role of the fact-finder in adjudicating liability. See id. at 248-49, 106 S.Ct. 2505. Denial of summary judgment where there is no genuine dispute of material fact, on the other hand, improperly permits a claim to go to the fact-finder, even though there can be only one possible outcome. See Celotex Corp. v. Catrett, 477 U.S. 817, 327, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986); Anderson, 477 U.S. at 250-52, 106 S.Ct. 2505. In determining “whether the evidence presents a sufficient disagreement to require submission to a jury or whether it is so one-sided that one party must prevail as a matter of law,” Anderson, 477 U.S. at 251-52, 106 S.Ct. 2505, the court must view the evidence and draw all reasonable inferences in favor of the nonmoving party. Id. at 255, 106 S.Ct. 2505; Shreve v. Franklin Cnty., 743 F.3d 126, 132 (6th Cir.2014).
III.
Wallner first argues that the district court erred in granting summary judgment against her FMLA interference claim, primarily asserting that the district court wrongly concluded that Wallner failed to assert any FMLA interference claim that was distinct from her FMLA retaliation claim. On this issue, we affirm the district court and disagree with Wallner. “There are two theories of recovery under the FMLA: an interference (or entitlement) theory and a retaliation (or discrimination) theory.” Hunter v. Valley View Local Sch., 579 F.3d 688, 691 (6th Cir.2009). An interference claim stems from 29 U.S.C. § 2615(a)(1), which provides that an employer may not “interfere with, restrain, or deny the exercise of or the attempt to exercise, any right provided under [the FMLA].” Id.; see Wysong v. Dow Chem. Co., 503 F.3d 441, 446 (6th Cir.2007). Because an employer interferes with an employee’s exercise of FMLA rights whenever the employee does not receive the rights that are due to her under the statute, the intent of the employer is irrelevant to whether an FMLA violation has occurred under the interference theory. See Seeger v. Cincinnati Bell Tel. Co., 681 F.3d 274, 282 (6th Cir.2012); Arban v. West Publ’g Corp., 345 F.3d 390, 401 (6th Cir.2003). Although an employer may defeat liability by showing that the FMLA benefit did not accrue to the plaintiff because the plaintiff was discharged for some other, legitimate reason, see Donald v. Sybra, Inc., 667 F.3d 757, 762 (6th Cir.2012); Arban, 345 F.3d at 401, the only things that a plaintiff must show to establish an FMLA interference claim are that she was entitled to benefits under the FMLA, notified her employer of her intent to exercise her FMLA rights, and was denied the FMLA benefits to which she was entitled. See Edgar v. JAC Products, Inc., 443 F.3d 501, 507-08 (6th Cir.2006). As long as the plaintiff shows that she was harmed, she does not need to show that the employer intended to retaliate against her for exercising her rights. See id.
An FMLA retaliation claim, by contrast, arises from 29 U.S.C. § 2615(a)(2), which prohibits employers from discriminating [551]*551against anyone “for opposing any practice made unlawful by [the FMLA].” Id.; see Grace v. USCAR, 521 F.3d 655, 669 (6th Cir.2008). We have routinely construed this language as barring retaliation against anyone who has exercised his rights under the FMLA. See, e.g., Bryant v. Dollar Gen. Corp., 538 F.3d 394, 400-01 (6th Cir. 2008). Unlike in an FMLA interference claim, an employer’s motive is relevant to án FMLA retaliation claim “because retaliation claims impose liability on employers that act against employees specifically because those employees invoked their FMLA rights.” Edgar, 443 F.3d at 508. Thus, an FMLA retaliation claim is established where the plaintiff shows that she availed herself of a protected right under the FMLA, that she suffered an adverse employment action, and that there was a causal connection between the two. Id. Again, an employer may defeat liability by showing that it would have made the same decision anyway, even if it had not been impermissibly motivated. See Hunter, 579 F.3d at 692.
Although our cases have sometimes asserted that the interference provision— § 2615(a)(1) — also prohibits “retaliatory discharge for taking leave,” this does not mean that all claims under § 2615(a)(1) are synonymous with retaliation claims under § 2615(a)(2). Chandler v. Specialty Tires of Am., Inc., 283 F.3d 818, 825 (6th Cir.2002). The seminal case cited for this proposition is Skrjanc v. Great Lakes Power Serv. Co., 272 F.3d 309 (6th Cir.2001). There, the plaintiff alleged that he was discharged before he could take his requested FMLA leave. We analyzed these allegations as raising a retaliation claim, on the theory that the employer could have been retaliating against the plaintiff for notifying it of his intent to exercise his right to FMLA leave in the future. Id. at 314. See Seeger, 681 F.3d at 282; Weimer v. Honda of Am. Mfg., Inc., 356 Fed.Appx. 812, 815 (6th Cir.2009); Wysong, 503 F.3d at 447; Arban, 345 F.3d at 403; Chandler, 283 F.3d at 825.
As in Skrjanc, of course, a single act may potentially constitute both interference and retaliation: firing someone who has just requested FMLA leave before he can take it could be construed both as retaliation for having asked for leave and as interference with the employee’s ability to take it. See Skrjanc, 272 F.3d at 314. We have therefore held that “a plaintiff ha[s] not waived a claim based on the interference theory where the complaint alleged general violations of 29 U.S.C. § 2615 that could apply to both interference and retaliation claims.” Seeger, 681 F.3d at 282 (citation and internal quotation marks omitted) (construing Wysong, 503 F.3d at 446).
But the mere fact that two statutory provisions may both be implicated by a single act does not mean that the two provisions are perfectly coextensive. Wé reiterated in Seeger the long-standing substantive distinction between interference claims and retaliation claims. See id. As a rule of thumb, where a plaintiff does not dispute that she has been granted all of the FMLA leave that she requested, has received all of the leave to which she was entitled, and has been reinstated upon her return to work, “the essence” of the plaintiffs claim is likely a retaliation claim, not an interference claim, regardless of the conclusory labels under which it has been pled. Id.
Here, although Wallner’s complaint labeled a portion of her allegations as an “interference” claim, all of the supporting allegations proceeded upon a retaliation theory without mentioning the deprivation of any FMLA benefit to which she was entitled. This trend continued at summary judgment, where the district court [552]*552correctly perceived that Wallner’s theory of recovery was exclusively retaliation-oriented rather than entitlement-based. The district court properly analyzed Wallner’s assertions as raising a retaliation claim, not an interference one. See Seeger, 681 F.3d at 282.
Even to the extent that Wallner belatedly amplified her interference theory and now contends that Hilliard Lyons illegally truncated her FMLA leave by depriving her of five days to which she was entitled, she is incorrect on the merits. Wallner’s doctor determined that she was able to return to work as of October 6, and Wall-ner notified Hilliard Lyons that she would do so. The FMLA does not guarantee Wallner the right to stay at home until the 11th, despite the fact that both she and her doctor agreed that she was able to return to work before then. See 29 U.S.C. § 2612(a)(1)(D) (guaranteeing an eligible employee up to twelve weeks of unpaid leave if she has a “serious health condition” that made her “unable to perform the functions” of her position); Edgar, 443 F.3d at 506. See also Novak v. Metro-Health Med. Ctr., 503 F.3d 572, 578 (6th Cir.2007) (“The employer may require an employee to provide a doctor’s certification confirming the existence of a serious health condition.”) (citing 29 U.S.C. § 2613(a)). As of October 6, therefore, Wallner was no longer entitled to FMLA leave, meaning that her interference claim — to the extent that she ever raised one — is without merit. See also Ragsdale v. Wolverine World Wide, Inc., 535 U.S. 81, 89, 122 S.Ct. 1155, 152 L.Ed.2d 167 (2002) (noting that the FMLA “provides no relief unless the employee has been prejudiced by the violation”); Harris v. Metro. Gov’t of Nashville & Davidson Cnty., 594 F.3d 476, 484 (6th Cir.2010). We therefore affirm the district court’s judgment in favor of defendant on Wallner’s interference claim.
IV.
Next, Wallner contends that the district court erred in entering summary judgment in favor of defendant on her FMLA retaliation claim. We agree.
Although Wallner has pursued an FMLA retaliation claim on both single-motive and mixed-motive theories — arguing them in the alternative — we see no need to inquire whether Wallner has marshaled enough evidence to withstand summary judgment on a single-motive theory. We have observed previously that the pertinent portion of the FMLA — 29 U.S.C. § 2615(a) — prohibits an employer from using an employee’s exercise of FMLA rights as “a negative factor” against the employee when making an employment decision. Hunter, 579 F.3d at 692 (quoting 29 C.F.R. § 825.220(c)). Because this standard “envisions that the challenged employment decision might also rest on other, permissible factors,” we have held that “the FMLA, like Title VII, authorizes claims in which an employer bases an employment decision on both permissible and impermissible factors.” Id. See also Lewis v. Humboldt Acquisition Corp., 681 F.3d 312, 321 (6th Cir.2012) (en banc) (declining to revisit Hunter). Because the operative statutory language permits recovery on a mixed-motive theory, we see no reason to produce dicta about whether Wallner possesses even more evidence than is required by the statute.
The parties agree that Wallner’s mixed-motive FMLA claim is governed by the Price Waterhouse1 burden-shifting approach that was applied in Hunter. Under that scheme, if the plaintiff has presented evidence to establish that her exercise of [553]*553her FMLA rights “played a motivating part in an employment decision,” then the defendant may avoid liability “only by proving by a preponderance of the evidence that it would have made the same decision even if it had not taken [the plaintiffs exercise of her rights] into account.” Gross v. FBL Fin. Servs., Inc., 557 U.S. 167, 173-74, 129 S.Ct. 2343, 174 L.Ed.2d 119 (2009) (distilling Price Waterhouse); Hunter, 579 F.3d at 692. Plaintiff may rely upon either direct or circumstantial evidence to establish that defendant used her exercise of FMLA rights against her as a negative factor in an employment action. See Desert Palace, Inc. v. Costa, 539 U.S. 90, 99-100, 123 S.Ct. 2148, 156 L.Ed.2d 84 (2003); Hunter, 579 F.3d at 692 n. 2.
Under Hunter, we agree with Wallner that the district court erred in entering summary judgment against her mixed-motive FMLA retaliation claim. We note, however, that this is not because Wallner possesses direct evidence of discrimination. Although Wallner contends that the “document of deficiencies” created by Moorman is direct evidence that Hilliard Lyons used her FMLA leave as a negative factor in determining whether to terminate her, she overstates the strength of her case. Contrary to Wallner’s argument, the document does not require, if believed, “the conclusion that unlawful discrimination was at least a motivating factor in the employer’s actions.” Demyanovich v. Cadon Plating & Coatings, LLC, 747 F.3d 419, 432 (6th Cir.2014) (defining “direct evidence”) (citation omitted). Although Wallner’s FMLA leave featured prominently in Moorman’s document and certainly could be inferred as one of the accumulated “deficiencies” that contributed to his termination decision, the document could also be read as referencing her failure to communicate while on leave, not merely her having taken the leave in the first place. Because the evidence requires an inferential step before it proves that unlawful discrimination occurred, it is not direct evidence of an FMLA violation and therefore does not on its own shift the burden to the employer “to prove by a preponderance of the evidence that it would have made the same decision absent the impermissible motive.” Daugherty v. Sajar Plastics, Inc., 544 F.3d 696, 707 (6th Cir.2008) (citation omitted).
But even though Wallner lacks direct evidence of discrimination, she has presented circumstantial evidence of retaliation sufficient to forestall summary judgment on her claim.2 Hilliard Lyons relies [554]*554on only one putatively legitimate reason for firing Wallner: her “excessive tardiness between October 7 and October 15, 2009.” But, again, Wallner may succeed on her FMLA retaliation claim even if multiple factors contributed to Hilliard Lyons’s decision to discharge her; she is entitled to a jury so long as a reasonable one could find that both permissible and impermissible factors motivated the termination decision. Hunter, 579 F.3d at 692. And on the evidence presented by Wallner, a reasonable jury could find that the decision to terminate Wallner did not depend upon her tardiness alone. Several considerations convince us that the district court erred when it concluded that no reasonable jury could find that Hilliard Lyons used Wallner’s exercise of her FMLA rights as “a negative factor” when deciding to terminate her employment. Id.
First, the district court understated the strength of the causal connection suggested by the facts of this case. Temporal proximity between “the plaintiffs exercise of protected rights” and an adverse action is relevant to whether the former is causally connected to the latter. Nguyen v. City of Cleveland, 229 F.3d 559, 563 (6th Cir.2000); see Mickey v. Zeidler Tool & Die Co., 516 F.3d 516, 525 (6th Cir.2008). The district court focused on the length of time between Wallner’s initial request to take FMLA leave in June 2009 and her discharge in October 2009,
As Skrjanc in fact suggests, the causality determination should instead be informed by what we have always advised: namely, by reference to the temporal proximity between “the plaintiffs exercise of protected rights” and the subsequent adverse action. Nguyen, 229 F.3d at 563. The fact that Wallner took FMLA leave rather than continue working — and the fact that she persisted in taking FMLA leave until the earlier of (1) the end of the twelve-week statutory allotment or (2) the abatement of her serious medical condition — was an exercise of her FMLA rights. After all, if merely asking for FMLA leave is a protected activity, then surely actually taking the time off is, as well. Skrjanc, 272 F.3d at 314. The temporal proximity that is relevant to Wallner’s FMLA retaliation claim, therefore, is the length of time between when she was absent from work on FMLA leave and when she was discharged. This was a total of nine days, not four months. Thus, the inference of causality generated by temporal proximity is stronger than recognized by the district court.
Second, the “document of deficiencies” created by Moorman could reasonably be interpreted as identifying Wallner’s FMLA-related absence as one of several reasons for her termination. The document, which Moorman characterized as comprising essentially “the reasons” that Wallner was discharged, described at least four separate reasons for her termination. It concludes that “[a]n accumulation of deficiencies became intolerable.”
Although the document certainly could be read as denoting other factors (such as lapses in communication) as reasons for Wallner’s termination, a jury could also readily conclude that the document’s three paragraphs describing the negative impact of Wallner’s FMLA absence on Moorman’s department were identifying her FMLA-related absence as one of the “accumu-lat[ed] ... deficiencies” that motivated her termination. In fact, Moorman later admitted that he was irritated by Wallner’s constantly changing return-to-work date, which prompted him to be “suspicious” that she would not return from leave when she allegedly was required to. This suspi[555]*555cion was, of course, engendered by Hilliard Lyons’s own conduct, when Landgraf communicated several erroneous return-to-work dates to Moorman without ever notifying him of her error.
It would be reasonable for a jury to infer from this evidence that Wallner’s FMLA leave generated ill-will toward her from Moorman, her direct supervisor, and that it factored into his decision to terminate her. After all, Moorman’s dissatisfaction with Wallner over her failure to return from FMLA leave when he expected her to is not far removed from his stated rationale for ultimately terminating her— both involve her perceived unwillingness to be at work when she was supposedly required to be there. Although Hilliard Lyons now contends that the document’s reference to Wallner’s FMLA absence was only to indicate dissatisfaction with her failure to communicate with her supervisors regarding her return-to-work date, Hilliard Lyons’s FMLA policy does not require any such communication. At most, Hilliard Lyons sent Wallner a letter requiring her to keep it apprised of her “status and intention to return to work,” and there is no dispute that Wallner informed Hilliard Lyons as soon as her status changed — i.e., when her doctor cleared her for a return to work.
None of this, in any event, changes the fact that the document could reasonably be interpreted by a jury as an admission by Moorman that Wallner was terminated partially because of the fact that she took FMLA leave, not merely because she failed to communicate her return date to him. On its face, the document is easily susceptible to the interpretation that Wall-ner’s FMLA-related absence was one of four or five reasons motivating her discharge, and it certainly does not support Hilliard Lyons’s argument that tardiness was the sole factor contemplated by Moor-man when he decided to fire Wallner. The document is, on its own, fairly significant evidence that more went into Hilliard Lyons’s decision to terminate Wallner than solely her continuance of a years-long trend of tardiness.
Hilliard Lyons nevertheless insists that no reasonable jury could conclude anything other than that Wallner would have been fired anyway, even if her FMLA leave had not been considered. See Gross, 557 U.S. at 173-74,129 S.Ct. 2343 (explaining the Price Waterhouse standard); Hunter, 579 F.3d at 693-94.3 We think [556]*556not. Although Wallner does not dispute that she arrived late to work on five of the seven days after her return from FMLA leave, a jury could reasonably find that Hilliard Lyons is overstating how seriously it viewed her late arrivals. Notably, Moorman himself did not keep any records of Wallner’s arrival times, despite the fact that human resources personnel had instructed him to do so if her tardiness was ■a problem. According to the contemporaneous estimates kept by Wallner’s coworker Dunning, Wallner arrived at 8:05, 8:02, 8:06, 8:09, and 8:05 a.m. on the five days in question. But neither Landgraf nor Moorman claims to have received Dunning’s estimates prior to Wallner’s termination, meaning that Wallner ostensibly was fired for tardiness without rebanee upon any actual record of how often or how late she arrived. Further, video footage later established that Wallner actually entered the building on the days in question earlier than Dunning claimed she did — at 8:04:37, 8:01:41, 8:05:04, 8:08:55, and 8:02:52, respectively. Despite the fact that Hilliard Lyons claims that this is proof positive that Wallner needed to be fired, Moorman himself testified that “to be five minutes late is not necessarily a swing factor; five to 15 minutes late consistently when it affects other employees is a major problem.” According to Hilliard Lyons’s own evidence, Wallner was more than five minutes late on only two days during the time in question — exceeding Moorman’s cut-off once by five minutes and once by four seconds. She did not exceed the five-minute acceptable limit on the day that Moorman decided to fire her. Although Hilliard Lyons certainly could have fired Wallner for arriving late to work (or for any other legal reason, for that matter), a jury weighing these facts could well be suspicious of the argument that Wallner’s tardiness was the only reason that she was fired.
Moreover, despite Hilliard Lyons’s assertion that Wallner was “chronically” late for work, the evidence on this is thin, consisting mainly of Moorman’s assertion that he had verbally warned Wallner about it for years. Wallner’s recent performance evaluations, for example, made no mention of chronic tardiness. But even assuming that Wallner’s tardiness did contribute to the ultimate decision, the timing is suspicious. If she had been chronically late for twenty-seven years, why fire her only now? Why did her co-worker begin chronicling Wallner’s arrival times only on the very morning that she returned to work from FMLA leave? Why did Wallner receive a “final” written warning on the day she returned from FMLA leave for conduct indisputably associated with her assertion of FMLA rights when the reprimand was not consistent with protocol, which required an intermediate written warning prior to a “final” written warning? Of the possible competing explanations, it would not be unreasonable for a jury to select the one that Wabner urges: that, particularly in light of Moorman’s “document of deficiencies,” Hilliard Lyons’s conduct was at least partiaby responsive to her decision to take FMLA leave and the significant conflict over her correct return-to-work date that erupted while she was exercising her statutorily protected rights.
Of course, this is not to say that Hilliard Lyons must necessarily be disbelieved when it argues that Wallner’s decision to take FMLA leave, despite Hilliard Lyons’s repeated attempts to compel her to return before she was required to, had nothing to do with her termination nine days after she returned. Hilliard Lyons is free to make that argument to a jury. But, as [557]*557indicated, we do not agree with Hilliard Lyons that a jury would necessarily be unreasonable if it failed to accept Hilliard Lyons’s version of events. A reasonable jury could find that Wallner’s exercise of her FMLA rights may have been a motivating factor in Hilliard Lyons’s decision to terminate her. See Gross, 557 U.S. at 173-74, 129 S.Ct. 2343; Hunter, 579 F.3d at 693-94. That is all that is required for Wallner to defeat summary judgment on her FMLA retaliation claim, and the district court erred in concluding otherwise. See Desert Palace, 539 U.S. at 101, 123 S.Ct. 2148; Hunter, 579 F.3d at 692.
V.
For these reasons, we affirm the entry of summary judgment against Wallner’s FMLA interference claim, reverse the entry of judgment against her FMLA retaliation claim, and remand for further proceedings consistent with this opinion.