*1334Before the court are the Final Results of Redetermination Pursuant to CIT Order, ECF No. 63 (May 21, 2018) ("Final Remand Results"), of the U.S. Department of Commerce ("Commerce" or "the Department") in its review of the request for a new shipper review ("NSR") by Inner Mongolia Jianlong Biochemical Co., Ltd. ("Jianlong"). As part of its antidumping duty investigation of xanthan gum from the People's Republic of China, Jianlong had requested a NSR, which was ultimately rescinded by Commerce. Xanthan Gum from the People's Republic of China , 81 Fed. Reg. 56,586 (Dep't Commerce Aug. 22, 2016) (rescission of NSR) and accompanying Issues & Decision Mem ("I & D Mem."). The court remanded to Commerce for reconsideration of that rescission, as well as other findings made by the Department in its rescission decision. See Inner Mongolia Jianlong Biochemical Co. v. United States , 41 CIT ----, 279 F.Supp.3d 1332 (2017) (" Jianlong I "). On remand, Commerce continues to find support for its decision to rescind Jianlong's NSR, albeit on separate, alternative grounds. See generally Final Remand Results. The court now sustains Commerce's findings as to the atypicality of certain aspects of the sale in question. However, without substantial evidence supporting its selection of input sales, the court is powerless to sustain Commerce's ultimate bona fide finding. Therefore, the court remands to Commerce: 1) for further consideration of the sales price, with reference to the entire Fufeng AR2 dataset, and 2) for a finding as to the bona fide nature of the NSR transaction that incorporates Commerce's additional sales price analysis.
BACKGROUND
The appeal previously presented to the court arose out of Commerce's review of Jianlong's NSR, the rescission of which was found to be unsupported by substantial evidence. Jianlong I , 279 F.Supp.3d at 1342. That order faulted the Department for: 1) determining that Jianlong had failed to meet the regulatory requirements for requesting a NSR, 19 C.F.R. § 351.214(b)(2)(iv)(A) ; 2) failing to support with substantial evidence its finding that the sole transaction reported in the period of review ("POR") was non-bona fide ; and 3) rejecting certain documents (Exhibits 1, 2, and 3 submitted in response to new factual information placed on the record in the Department's Preliminary Bona Fide Sales Analysis ("Exhibits 1, 2, and 3") ) without considering the stated purposes for which the exhibits were offered. The court ordered that Commerce reconsider each issue. Id.
In reexamining its decision to rescind Jianlong's NSR on the basis of Jianlong's failure to report its sample shipments, Commerce has now determined that it lacks substantial evidence to support that determination. Final Remand Results at 6-7. As such, Commerce does "not, in this instance, consider Jianlong's failure to report the[ ] sample shipments ... as a failure to meet" the regulatory requirements for requesting a NSR. Id. at 7.
The Department therefore has permitted the NSR to proceed under 19 C.F.R. § 351.214(b)(2)(iv)(A), but now bases its rescission on its prior, alternative basis: that Jianlong's only sale during the POR was non-bona fide. See id. at 12-23, 36-52. "Commerce reexamined all of the evidence on the record of the proceeding and the totality of circumstances surrounding Jianlong's *1335NSR sale.... [and] identified additional record evidence that supports the conclusion that Jianlong's single NSR sale is not a bona-fide sale." Id. at 36. Specifically, the Department determined that four factors support this conclusion: 1) the establishment of Jianlong's U.S. affiliate, Jianlong USA; 2) the timing of the sale in the context of the POR; 3) the sales price; and 4) a new additional factor, namely the actions of [ [ ] ] after the sale was completed. Final Remand Results at 12-23. Commerce compared Jianlong's one reported sale to those the Department selected from sales reported by Neimenggu Fufeng Biotechnologies Co., Ltd. ("Fufeng") in the second administrative review ("AR2") of the antidumping duty order on xanthan gum from China.1 Commerce used only a subset ("input sales" or "Fufeng subset") of the complete dataset of sales reported by Fufeng ("Fufeng AR2 dataset"). The comparison done by Commerce indicated that Jianlong's sales price was high. The evidence, "when considered with the high price" of the transaction in question, indicates to Commerce that Jianlong's normal sales process was not followed, which in turn, suggests to Commerce that the sale in question is non-bona fide. Id. at 52.
Last, Commerce continues to reject Exhibits 1, 2, and 3 offered by Jianlong as clarifying information. Id. at 25-28, 52-56. Commerce has "considered the power of Jianlong's new factual information to rebut, clarify or correct existing factual information and [finds] that [it is rejected because] it does not serve any of these functions." Id. at 55.
In the instant appeal, Jianlong does not contest Commerce's change of position as to the reporting of sample shipments, but now argues that the Department's other determinations are unsupported by substantial evidence. Pl.'s Comments on Final Results of Redetermination Pursuant to Remand, ECF No. 66 (June 20, 2018) ("Pl.'s Comments"). Accordingly, Jianlong asks the court for a remand so that Commerce can further consider whether the sale to [ [ ] ] was bona fide as well as whether the Department has improperly rejected Jianlong's clarifying documents. Because Commerce's selection of the Fufeng subset lacks substantial evidence, the court remands for further consideration in accordance with this opinion. All other determinations as to the atypicality of the sole NSR transaction are sustained.
Having reviewed the Department's findings and conclusions of law, the court grants Jianlong's motion in part, remanding to Commerce for the limited basis of conducting a further sales price comparison analysis and a corresponding finding as to the bona fide nature of the transaction in question.
JURISDICTION AND STANDARD OF REVIEW
The court has jurisdiction pursuant to 28 U.S.C. § 1581(c) and will sustain Commerce's determinations unless they are "unsupported by substantial evidence on the record, or otherwise not in accordance with law." 19 U.S.C. § 1516a(b)(1)(B)(i). "Substantial evidence requires 'more than a mere scintilla,' but is satisfied by 'something less than the weight of the evidence.' " Altx, Inc. v. United States , 370 F.3d 1108, 1116 (Fed. Cir. 2004) (citations omitted). If "a reasonable mind might accept the evidence as sufficient to support the finding,"
*1336Maverick Tube Corp. v. United States , 857 F.3d 1353, 1359 (Fed. Cir. 2017), the substantial evidence threshold is likely met so long as the Department has provided a reasoned basis for its decision. See Motor Vehicle Mfrs. Ass'n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co. , 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983). Commerce's consideration should reflect a sound decision-making process, see Burlington Truck Lines v. United States, 371 U.S. 156, 168, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962), taking into account all evidence on the record, including that which may detract from the ultimate conclusion, CS Wind Vietnam Co. v. United States , 832 F.3d 1367, 1373 (Fed. Cir. 2016). But, whatever the result, the agency's rationale must not be arbitrary, capricious, or an abuse of discretion. Wheatland Tube Co. v. United States , 161 F.3d 1365, 1369 (Fed. Cir. 1998).
DISCUSSION
For the reasons outlined below, the court sustains Commerce's determinations regarding: 1) Jianlong's reporting of sample shipments and 2) the individual findings of typicality pursuant to 19 U.S.C. § 1675(a)(2)(B)(iv) (II-VII). Yet, the court remands for a further consideration of Jianlong's sales price that considers the entirety of the Fufeng AR2 dataset, as well as an ultimate bona fide finding that accounts for the price comparison to be done on remand. It is recommended that, on remand, the Department admit Exhibits 1, 2, and 3 submitted by Jianlong so as to provide for a proper determination, supported by substantial evidence, that Commerce has selected all relevant input sales for its sales price analysis.
I. Entries for Consumption
In the court's prior decision, it ordered Commerce to conduct "a more fulsome consideration of Jianlong's sample shipments as entries for consumption." Jianlong I , 279 F.Supp.3d at 1339. Having now acknowledged that there is not substantial evidence on the record to support its prior rescission of Jianlong's NSR under 19 C.F.R. § 351.214(b)(iv)(A), the Department "will not, in this instance, consider Jianlong's failure to report these sample shipments in its NSR request as a failure to meet the requirements of 19 C.F.R. § 351.214(b)(iv)(A)."2 Jianlong agrees that "the final remand results correctly confirm that Jianlong did not fail to meet the regulatory requirements of 19 C.F.R. § 35l.214(b)(2)(iv)(A)." Pl.'s Comments at 4. As this finding is no longer in dispute, the court sustains Commerce's determination here.
II. Non-Bona Fide Transaction
As part of its review of Jianlong's NSR request, Commerce has conducted an analysis of the sale in question to determine if it is bona fide. The court's prior remand ordered Commerce to conduct its analysis, considering the totality of the circumstances, and make a determination "sufficiently supported by substantial evidence, explaining how the establishment of Jianlong USA, the timing of the sale, and the sales price support a finding that the transaction in question was, or was not, bona fide ." Jianlong I , 279 F.Supp.3d at 1342. Complying with that directive,3 Commerce *1337has considered each factor and its determination seeks to explain that the totality of the circumstances indicates that the sale was non-bona fide. Yet, at this juncture, Commerce has not sufficiently supported its determination so as to allow the court to sustain its ultimate bona fide finding. As a result, the court remands for further proceedings in accordance with this opinion while also sustaining certain elements of the Department's examination.
Because the Department's individual findings as to the atypicality of Jianlong's single sale are supported by substantial evidence, the court sustains Commerce's findings in this area. See infra sections II.B, II.C, II.E. However, the Department's methodology and findings as to the sales price do not find sufficient support in the record. The Department's determination that the sales price of the NSR transaction was high is a crucial one that can greatly impact the bona fide sale analysis. As part of this assessment, Commerce must spell out its selection of the Fufeng subset against which it compared Jianlong's sale. On remand, Commerce is to reconsider its sales price analysis, with reference to the full Fufeng AR2 dataset, and adjust its bona fide finding accordingly.
A. Legal Framework
Congress has charged Commerce with calculating dumping margins under 19 U.S.C. § 1675(a)(2). As part of that mandate, the Department will conduct a NSR when requested to do so by an exporter not subject to a current antidumping duty order. That assessment requires that Commerce consider only bona fide sales-that is, those that Commerce determines to be such by reference to several factors:
(I) the prices of such sales;
(II) whether such sales were made in commercial quantities;
(III) the timing of such sales;
(IV) the expenses arising from such sales;
(V) whether the subject merchandise involved in such sales was resold in the United States at a profit;
(VI) whether such sales were made on an arms-length basis; and
(VII) any other factor the administering authority determines to be relevant as to whether such sales are, or are not, likely to be typical of those the exporter or producer will make after completion of the review.
19 U.S.C. § 1675(a)(2)(B)(iv)(I-VII). "[B]ecause the ultimate goal of the new shipper review is to ensure that the U.S. price side of the antidumping calculation is based on a realistic figure, any factor which indicates that the sale under consideration is *1338not likely to be typical of those which the producer will make in the future is relevant." Tianjin Tiancheng Pharm. Co. v. United States , 29 CIT 256, 260, 366 F.Supp.2d 1246, 1250 (2005).
Ultimately, though, Commerce must consider the totality of the circumstances-as outlined by the factors under 19 U.S.C. § 1675(a)(2)(B)(iv) -in order to assess the commercial reasonableness of the transaction in question. In each NSR, the factors that have the most weight will vary based on the circumstances, but one aspect will remain constant throughout each review: the aim of the process is "to ensure that a producer does not unfairly benefit from an atypical sale to obtain a lower dumping margin than the producer's usual commercial practice would dictate." Huzhou Muyun Wood Co. v. United States, Slip Op. 18-89, 324 F. Supp. 3d 1364, 1376, 2018 WL 3455350, at *8 (CIT July 16, 2018) (" Huzhou Muyun II "). Evidence of only a single sale during the POR may raise certain suspicions; as such, "[w]hile a single sale is not inherently commercially unreasonable, it will be carefully scrutinized to ensure that new shippers do not unfairly benefit from unrepresentative sales." Tianjin Tiancheng , 29 CIT at 275, 366 F.Supp.2d at 1263 (citation omitted).
The goal of the bona fide sales analysis is to reveal whether the transaction is "unrepresentative or extremely distortive." Tianjin Tiancheng , 29 CIT at 259, 366 F.Supp.2d at 1249-50. If the entirety of the § 1675(a)(2)(B)(iv) factors indicate that a non-bona fide transaction may be afoot, that transaction is to be discarded so that the new shipper is unable to manipulate the NSR for the purposes of obtaining a lower rate than it is entitled. In the absence of any bona fide sales, Commerce will rescind the NSR.
In reviewing Commerce's bona fide determinations, the court is sensitive to the congressional concerns that motivated adoption of the § 1675(a)(2)(B)(iv) factors: the potential for abuse of the NSR process "to avoid antidumping and countervailing duties." H.R. REP. NO. 114-114(I), pt. 1, at 89 (2015). As a result, the bona fide assessment seeks to determine the typicality of the transaction in question, thus allowing for review only of legitimate transactions with no signs of "abuse." See id. ; see also generally Huzhou Muyun II , 324 F. Supp. 3d 1364, 2018 WL 3455350.
Here, Commerce considered: 1) Jianlong's formation of a U.S. entity, Jianlong USA; 2) the timing of the sale by Jianlong to [ [ ] ]; 3) the sales price, as compared to Fufeng's relevant reported sales; and 4) [ [ ] ] continued behavior after the POR. Each factor is discussed in detail below.
B. Formation of Jianlong USA
Commerce found that the circumstances surrounding the establishment of Jianlong USA indicate that the transaction in question was not typical of those Jianlong would engage in in the future. In light of the stated sales process provided by Jianlong and that which it undertook in its sale to [ [ ] ], the court sustains Commerce's finding as reasonable.
On the record, Commerce requested that Jianlong describe how it "sets the prices of the merchandise" and the way in which price negotiations take place. Section A Resp. A-5, ECF No. 50, Confidential J.A. Tab 4, C.R. 6-9 (Sep. 30, 2015) ("Section A Resp."). Jianlong's response indicated the following:
Jianlong and Jianlong USA set prices based on market conditions, competitive price information and cost of production considerations. Jianlong and Jianlong USA have full autonomy in negotiating prices with U.S. customers, free of intervention from any entity outside the company.
*1339With authorization from Jianlong, Jianlong USA negotiates directly with the unaffiliated customers. The purchase order and e-mail correspondence included in the sale trace in Exhibits [sic ] A-5 provides evidence [sic ] independent price negotiations.
Id. Exhibit A-5 included emails and their attachments detailing a back and forth between Jianlong USA and [ [ ] ]. See id. at ex. A-5. Other responses given by Jianlong indicated that Jianlong USA also had the responsibility of "identifying U.S. customers." Id. at A-17. The response suggested that Jianlong USA's sales staff would accomplish this "by several different means: through trade fairs, exhibitions, requests for proposals, and contacts within the industry." Id.
From the chain of emails provided by Jianlong, Commerce has inferred "that the sales process reported to be the normal sales process for Jianlong USA was not followed for the sale in question." Final Remand Results at 13. The Department finds that these emails indicate a sales process devoid of normal price negotiations. Id. at 13-14. Additionally, Commerce finds the lack of Jianlong USA's sales expenses particularly troubling "because the sale under review, allegedly made by Jianlong USA, was reported as a constructed export price (CEP) sale," the calculation of which requires the examination of the expenses involved in a typical transaction. Id. at 15. The expenses reported did not include those required to "identify or communicate with potential customers." Id. at 14. Collectively, this, the Department finds, "is not representative of Jianlong/Jianlong USA's normal business practices, which is one factor that supports a finding that the NSR sales transaction is not bona fide. " Id. at 16.
For its part, Jianlong contests Commerce's findings, stating that "the record evidence contradicts these conclusions." Pl.'s Comments at 5. With regard to price negotiations, Jianlong points out that there was a line of communication that: 1) altered the price from the invoice to the purchase order and 2) resulted in technical specifications and certain modifications to the order. This, Jianlong argues, demonstrates that "Jianlong USA clearly acted on its customer's behalf to provide superior services for its U.S. customer," id. 6, while "Jianlong participated in many aspects of the sale to ensure clear communication ... and to ensure that there was an enhanced meeting of the minds regarding all terms of sales and other aspects of the NSR transaction," id. at 5. Jianlong also argues that Commerce failed to consider that its "normal office expenses," including those expenses for "salaries and office equipment," may in fact cover the sales expenses the Department found to be lacking from the record-particularly those associated with communicating with potential customers. Id. at 6.
As an initial matter, the court finds that whether or not a newly formed business has followed its normal sales process is a reasonable factor for Commerce's consideration under 19 U.S.C. § 1675(a)(2)(B)(iv)(VII). As part of its bona fide analysis, Commerce may consider "any other factor [it] determines to be relevant as to whether such sales are, or are not, likely to be typical of those the exporter or producer will make after completion of the review." 19 U.S.C. § 1675(a)(2)(B)(iv)(VII). Clearly, where the circumstances indicate that the establishment of a U.S. entity is indicative of an atypical transaction, it is within Commerce's discretion to consider that factor as a part of its totality of the circumstances analysis.
To support its non-bona fide finding, the Department cites a mismatch between the *1340reported process and that which Jianlong followed in executing this transaction. Where there is but a single sale reported in the POR, Commerce obviously does not have the benefit of comparing the transaction in question to other transactions the new shipper has undertaken; as a result, the best evidence upon which the Department can rely may very well be the normal sales process reported by the new shipper. Certainly, a failure to follow a firm's own reported sales process may indicate that the sale in question is atypical. Here, the evidence in the record supports the view that the mismatch suggests that Jianlong's sale to [ [ ] ] was non-bona fide.
As Jianlong reported that "Jianlong USA negotiates directly with the unaffiliated customers," Section A Resp. at A-5, a genuine price negotiation is highly probative of whether the sale in question is, or is not, "likely to be typical" of sales Jianlong will make in the future. Here, the absence of "independent price negotiations," Section A Resp. at A-5, proves fatal to Jianlong's claim that the sale to [ [ ] ] followed its normal sales process. Surely, a genuine negotiation typically includes some combination of modifications related to price, quantity, or means of delivery. Indeed, Jianlong claims that it has proven Jianlong USA's involvement in "determining and confirming the unit price" and addressing certain "technical specifications." Pl.'s Comments at 5. However, Commerce reasonably determined that the terms of the sale had "already been established," Final Remand Results at 39, and that Jianlong USA's role did not reflect the typical sales process reported by Jianlong, id. at 42.
What's more, Commerce cites a lack of evidence to show that Jianlong USA engaged in one of its most crucial roles, identifying potential customers. Final Remand Results at 40-41 (citing Section A Resp. at A-17). Rather, Commerce found that the customer, [ [ ] ], had already been identified through personal contacts and that Jianlong, not Jianlong USA, had engaged in a prolonged period of relationship-building prior to the sale. Id. at 16-17, 37. While Jianlong attempts to demonstrate Jianlong USA's participation in customer identification activities through reported expenses, Pl.'s Comments at 5-6, Commerce found no support for that explanation in the record. Final Remand Results at 40. No other evidence exists on the record to show that Jianlong USA engaged in customer identification.
Resultantly, Commerce finds no evidence in the record demonstrating that Jianlong USA engaged in either of its two primary functions: price negotiations and customer identification.4 Therefore, substantial evidence supports the notion that Jianlong did not follow its typical sales process by establishing Jianlong USA to conduct the sale and this factor calls into question the bona fide nature of transaction as a whole. In the absence of evidence that Jianlong USA undertook its reported responsibilities during the transaction in question, Commerce was left with no choice but to conclude that the sales process was not reflective of sales Jianlong is likely to engage in in the future. As a result, the court sustains Commerce's findings related to this factor.
*1341C. Timing of Sale
Likewise, Commerce supported with substantial evidence its determination that the timing of Jianlong's sole transaction suggested that the sale was non-bona fide. The Department reasonably concluded that the timing of the sale "called into question whether the sale was made and timed specifically to obtain an NSR rate." Final Remand Results at 16. As such, the court sustains Commerce's findings as to this particular element of the Department's bona fide analysis.
Commerce viewed the timing of the sale significant because it signaled that the "focus was on making a U.S. sale of subject merchandise during the POR to obtain a NSR." Final Remand Results at 46-47. This, Commerce viewed, was supported by the following circumstances: "(1) a sales process executed within 4 days at the end of the POR, even though Jianlong was in contact with the U.S. customer [ ] a year and a half earlier; (2) [ [ ] ]; (3) [ [ ] ]; and, (4) [ [ ] ] instruction to Jianlong USA [ [ ] ]. Id. at 47. Ultimately, that all suggested to Commerce that the sale was atypical and, thus, not bona fide. Id. at 47.
To this, Jianlong responds by questioning Commerce's presumption "that a sale made in the last month of a period of review raises concerns regarding the bona fide nature of the sale." Pl.'s Comments at 8. Additionally, Jianlong contends that the fact that no sample was provided before shipment is of no moment (because samples had already been provided), that there is no record evidence to support the conclusion that rushing to make shipment is not typical, and that "there is no evidence ... that shipment of the merchandise actually was expedited" because the shipment was made by "normal sea freight" and not some other expedited means. Id.
By its terms, 19 U.S.C. § 1675(a)(2)(B)(iv)(III) enables Commerce to consider the timing of a sale in its totality of the circumstances analysis. That consideration enjoys its due weight and the Department may consider when the sale was made both in relation to other sales and the timing relative to the POR itself. The timing of a single sale may give the Department pause as it may indicate that the sale was made in an expedited manner so as to complete the transaction within a given POR.
Jianlong asserts that the Department's findings in this space are lacking in that they rely on the sort of "suspicion and innuendo" the court had previously found deficient. Pl.'s Comments at 6-7 (citing Jianlong I , 279 F.Supp.3d at 1340 ). Specifically, Jianlong argues that Commerce relies on no evidence to conclude that "there was no communication between the parties over 450 days" and that the parties "rushed to complete the sale ... in 37 days." Id. at 7. (citing Final Remand Results at 17). Jianlong contends that Commerce disregarded "the long history and relationship between Jianlong and its U.S. customer" and that "[i]t is not reasonable ... for Commerce to consider a year-and-a-half long process to be a 'rush to complete a sale.' " Id.
Regardless, the Department's finding that the timing of the sale is suggestive of a non-bona fide sale is reasonable and, therefore, sustained by the court. Certainly, that "the xanthan gum was shipped [ [ ] ] days prior to, and invoiced [ [ ] ] days prior to, the end of the POR," Final Remand Results at 42, is a relevant consideration. That timing becomes questionable when combined with the following: 1) there were interactions between the firms at least a year and a half prior to the reported sale; 2) there was no evidence in the record that Jianlong and [ [ ] ] had any communication for 450 days before the sale; and 3) the communications that ensued *1342between [ [ ] ] and Jianlong USA spanned over only four days, during which time Jianlong represents that Jianlong USA negotiated and completed a sale. See id. at 43. That narrative has left Commerce to consider an initial relationship, a period of silence, and then a quickly executed sale near the end of the POR. Under such circumstances, it is reasonable for Commerce to conclude that the sale in question was timed specifically to obtain an NSR and, therefore, atypical.
Further, Commerce considered several other reviews in which the timing of a sale similarly suggested that a given NSR transaction was non-bona fide. Final Remand Results at 19 (citing Certain Preserved Mushrooms from the People's Republic of China , 82 Fed. Reg. 1,317 (Dep't Commerce Jan. 5, 2017) (rescission of NSR); Crystalline Silicon Photovoltaic Cells , Whether or Not Assembled into Modules, from the People's Republic of China , 80 Fed. Reg. 55,090 (Dep't Commerce Sept. 14, 2015) (rescission of NSR); Certain Steel Nails from the People's Republic of China , 76 Fed. Reg. 56,147 (Dep't Commerce Sept. 12, 2011) (prelim. results & prelim. rescission of NSR) ).5 In both Crystalline Silicon Photovoltaic Cells and Certain Preserved Mushrooms , Commerce determined that the lack of sales activity occurring between firms followed by a sale in the last month of a POR raised suspicions as to the bona fide nature of a transaction. Commerce thus has demonstrated its consistent, reasonable practice such that the court declines to overturn its reasoned determination.
Accordingly, the court sustains Commerce's findings as to the timing of Jianlong's sales.
D. Sales Price
Next, Commerce has marched through a methodology for comparing Jianlong's transaction to Fufeng's sales, but has failed to support with substantial evidence the selection of the input sales. As the sales price is, most often, the most important factor in the Department's bona fide analysis, the need for substantial evidence supporting any determination in this area is of paramount importance to the court's consideration. The analysis provided to this point remains unsatisfactory to the court. Without substantial evidence supporting Commerce's selection of input sales, the court cannot sustain the Department's conclusions. As a result, the proceedings are remanded for further consideration in accordance with this opinion.
At the outset of the proceedings, Commerce placed on the record certain sales from Fufeng for comparison against Jianlong's one sale; ultimately, that comparison produced results indicating to Commerce that the sales price was atypical and, thus, another factor showing that this was a non-bona fide sale. Having provided a reasonable rubric for the comparison of Jianlong's NSR transaction with similar Fufeng sales, Commerce's methodology is sustained. However, Commerce has left the record bereft of any meaningful indication of how the Fufeng subset was selected. The court, therefore, remands to Commerce *13431) for additional consideration of the input sales placed on the record that takes the entire Fufeng AR2 dataset into account and 2) to engage in the same price comparison methodology, if appropriate, using any new sales information.
In general, Commerce compared Jianlong's sales against the input sales, concluding that Jianlong's U.S. sales price was high and, therefore, indicative of a non-bona fide sale. Commerce's methodology "reduced Jianlong USA's sales price for all appropriate transportation expenses, customs duties, taxes, and U.S. direct and indirect selling expenses," Final Remand Results at 48-49; declined to make adjustments for customer types, instead following its practice to use "control numbers (CONNUMs) for product matching purposes that reflect the physical characteristics of the merchandise," id. at 49-50; and followed its practice of not adjusting "for cash deposit rates when calculating net prices in its price comparisons," id. at 50.
Where there is substantial variability between product characteristics, it is reasonable for Commerce to choose a comparison product and evaluate the price at which that comparison product was sold. See, e.g., Huzhou Muyun Wood Co. v. United States , 41 CIT ----, ----, 279 F.Supp.3d 1215, 1227-28 (2017) (" Huzhou Muyun I "). When making this choice, Commerce is to support with substantial evidence its choice of the comparison product as well as any adjustments made to the price comparison. See id. As is the case with all questions of substantial evidence, the court is only able to sustain Commerce's determinations on the grounds invoked by the agency, State Farm , 463 U.S. at 43, 103 S.Ct. 2856 ; if the record is devoid of any indication of the reasons for a certain determination, the court is powerless to sustain that finding.
As discussed below, while its means for comparing two sets of data was reasonable in this context, Commerce's selection of input sales remains suspect. Accordingly, pursuant to this opinion, on remand Commerce must: 1) articulate its selection of the Fufeng subset with reference to the entire Fufeng AR2 dataset, 2) engage in a proper sales comparison, and 3) make an ultimate finding as to the bona fide nature of the transaction at issue.
1. Input Sales
The court agrees that sales of the same grade of xanthan gum are the relevant sales for Commerce's price comparison, but finds insufficient evidence on the record to verify the Department's representation that it extracted all relevant sales from the complete Fufeng AR2 dataset. Commerce cites to no methodology for its selection of the Fufeng subset upon which it relies in its price comparison, but rather merely asserts that the Department has placed the complete relevant subset of Fufeng sales on the record. See I & D Mem. at cmt. 2 (stating Commerce "believe[d] [it] ha[d] the complete relevant sales data from AR2 on the record" because it placed on the record only "the reported sales of xanthan gum of the same grade as the grade of xanthan gum sold by [Jianlong]."); Final Remand Results at 51 ("The complete relevant subset of Fufeng's AR2 sales data, which were extracted from Fufeng's full U.S. sales database and used in the comparison, is on the record."). Without a more developed record, articulating the methodology used for identifying input sales with reference to the entire Fufeng AR2 dataset, the court cannot sustain on the grounds invoked by the Department, namely that it has placed on the record the relevant input sales and reliance on the Fufeng subset was reasonable. Because this process falls short of the substantial evidence threshold required for *1344the court to sustain, the proceedings are remanded to Commerce for further consideration of the sales price and how that further analysis impacts the Department's ultimate bona fide finding.
While it is true that this court typically does not "upset Commerce's reasonable choice" of sales data, see Huzhou Muyun I , 279 F.Supp.3d at 1228 (quoting Ad Hoc Shrimp Trade Action Comm. v. United States , 36 CIT ----, ----, 828 F.Supp.2d 1345, 1354 (2012) ), the court must be presented with some reliable basis upon which it can discern the methodology employed by Commerce. See generally Allied Tube & Conduit Corp. v. United States , 31 CIT 1090, 1094, 2007 WL 2040695, at *4 (2007) ("Commerce has the discretion to choose whatever methodology it deems appropriate, as long as it is reasonable and its conclusions are supported by substantial evidence ." (emphasis added) ). While Commerce claims that it has "compared the price of Jianlong's NSR sales transaction to the average export price and maximum export price of the same type of xanthan gum sold by Fufeng," Final Remand Results at 20, the Department provides no explanation of how it initially determined which of Fufeng's reported sales belonged in the subset of input sales. When Commerce placed the Fufeng subset on the record, it merely described its methodology as follows: "We compared the quantity and unit price of the sale under review to the quantities and unit prices of sales of similar subject merchandise, with similar sales terms, reported by the mandatory respondents in [AR2] in this proceeding, which covers the same period as this NSR." Prelim. Bona Fide Sales Analysis 4, ECF No. 50, Confidential J.A. Tab 10, P.R. 136, C.R. 73-75 (Mar. 15, 2016) ("Prelim. Sales Analysis"). Clearly lacking from this statement-and the Department's successive discussions on the topic-is any indication of how Commerce identified the relevant sales.6 This flaw calls into question the results of the Department's sales comparison. Without an understanding of the choice of input sales and substantial evidence supporting their selection, the court cannot sanction the output produced by Commerce's methodology.
In a NSR sales price comparison, the starting point for Commerce is often to place on the record all sales, subsequently using only a relevant subset of those sales to compare to the new shipper's reported transactions. See, e.g., Zhengzhou Huachao Indus. Co. v. United States , 37 CIT ----, ----, 2013 WL 3215181, at *4 (May 14, 2013) ("Commerce placed on the record [Customs] data containing all entries of merchandise exported to the United States from the PRC during the POR ...." (emphasis added) ). Here, rather than place the full dataset on the record, Commerce merely asserted that it had chosen the relevant sales for consideration. See I & D Mem. at cmt. 2; Final Remand Results at 51.
Tellingly, Commerce seems to acknowledge that its chosen dataset may be under-inclusive. See Final Remand Results at 26 ("Commerce only compared the price of Jianlong's [ [ ] ] sales transaction during the POR of xanthan gum with Fufeng's sales of [ [ ] ] xanthan gum during AR2. Therefore, unless there were additional sales of [ [ ] ] xanthan gum made by Fufeng in AR2, that were omitted from the price-comparison data Commerce used in *1345its bona fide analysis, submitting the complete AR2 Fufeng sales database on the record does not rebut, clarify, or correct the factual information place on the record by Commerce." (emphasis added) ). Commerce cannot rely on Jianlong's burden of proof, 28 U.S.C. § 2639(a)(1), to discharge its duty to "establish[ ] antidumping margins as accurately as possible." Shakeproof Assembly Components, Div. of Il. Tool Works, Inc. v. United States , 268 F.3d 1376, 1382 (Fed. Cir. 2001) ; see also Lasko Metal Prods., Inc. v. United States , 43 F.3d 1442, 1443 (Fed. Cir. 1994) ("[T]here is much in the statute that supports the notion that it is Commerce's duty to determine margins as accurately as possible, and to use the best information available to it in doing so."). Commerce's own admission that the reliability of its dataset only furthers the court's doubts that the Department's selection of data may be unreliable. See Allied Tube & Conduit Corp. , 31 CIT at 1096, 2007 WL, at *5. As a result, it appears that both Commerce and Jianlong agree that there may be Fufeng sales of [ [ ] ] xanthan gum that were not included in the subset of input sales.
However inartful its discussion of the problem may be, See Pl.'s Comments at 11-12; Final Remand Results at 51, Jianlong is at least correct that without an articulation of Commerce's methodology to select the data upon which it relies, the court is left with no tools from which it can assess Commerce's determination. Because the court is unable to gauge the reliability of the input sales, the Department's omission amounts to a failure to "articulate any rational connection between the facts found and the choice made." Burlington Truck Lines , 371 U.S. at 168, 83 S.Ct. 239. The Fufeng subset does not, on its own, contain the requisite indicators of reliability. See Prelim. Sales Analysis at attach. III. Commerce's assurances that it "believes" it has placed all "relevant" sales on the record likewise provide little comfort.
As such, the court cannot sustain Commerce's finding that the sales price is indicative of an atypical sale. Without substantial evidence supporting the Department's view, this factor does not support Commerce's non-bona fide finding. Accordingly, the court remands this portion of the proceedings to Commerce for a further sales price analysis that indicates the methodology employed by Commerce for selecting the input sales. Commerce is to conduct this assessment with reference to the complete Fufeng AR2 dataset. Once that final piece of the bona fide analysis is complete, the Department should be able to resolve the bona fide sales inquiry informed both by reference to the factors sustained here, as well as a sales price determination (hopefully) supported by substantial evidence.
2. Methodology
Notwithstanding the deficiencies in Commerce's selection of input sales, the court sustains the Department's methodology for comparing sales-if not the result-employed by the Department. Following additional consideration of the complete Fufeng AR2 dataset, if appropriate, Commerce should utilize the same methodology, along with a reasoned decision explaining its selection of input sales-and possibly a new subset of input sales, depending on the results of Commerce's price determination-in the remand proceedings.
Commerce's methodology compared sales of the same grade of xanthan gum-[ [ ] ] xanthan gum-with the same terms of sale. Commerce began by narrowing the Fufeng dataset to only [ [ ] ] xanthan gum sales. Prelim. Sales Analysis at attach. III. By "limiting [its] comparison to similar products," Commerce appropriately limited *1346the scope of Fufeng's sales that it would consider. Final Remand Results at 51. Then, Commerce "reduced Jianlong USA's sales price for all appropriate transportation expenses, customs duties, taxes, and U.S. direct and indirect selling expenses."7 Id. at 48-49. This second step enabled Commerce to either "compare[ ] prices of sales with the same terms [ [ ] ], after making adjustments, or ma[ke] conservative comparisons with Fufeng's [ [ ] ] sales." Id. The result of Commerce's methodology produced a comparison showing that Jianlong's sale price of [ [ ] ] was high when compared to Fufeng's average of [ [ ] ], Prelim. Sales Analysis at attach. II-that is, "[ [ ] ] than the average price of the same grade of xanthan gum sold by Fufeng during the POR," Final Remand Results at 51.
Further, Commerce's view that "differences in timing, sales terms, customer type, and dumping duties" are not to be included in the comparison is reasonable. See Final Remand Results at 47-48. With respect to the selection of the input sales, "Commerce placed on the record data from AR2 for all the reported sales by Fufeng of xanthan gum that matched the grade of xanthan gum sold by Jianlong," id. at 51, and compared only Fufeng's sales of the particular grade of xanthan gum which Jianlong sold to [ [ ] ]. As to the timing of the sales, Commerce reviewed a substantial number of Fufeng transactions from the same period, [ [ ] ], and the average gross unit sales price of Fufeng's sales were similar across different time periods, [ [ ] ]. Id. at 48. That indicated to Commerce that the timing of Fufeng's sales across the POR remained stable. Commerce also explained that its methodology is to "use[ ] control numbers (CONNUMs) for product matching purposes that reflect the physical characteristics of the merchandise, not differences in customer type," id. at 49, and Jianlong has failed to raise a significant reason to depart from this practice. Last, Commerce dismissed "Jianlong's claim that Commerce should adjust for dumping cash deposit rates" by explaining that the Department has a "longstanding practice [ ] not to adjust for cash deposit rates when calculating net prices in its price comparisons." Id. at 50.
In each instance detailed above, Commerce either articulated a specific reason why departing from its regular practice was not appropriate or supported its choice with substantial evidence. For considerations of customer type and dumping cash deposit rates, Commerce followed its respective practices in these spaces. As those practices reflect a reasonable interpretation of the Department's duties, the court finds the agency's reasonable interpretations to be in accordance with law. As to the issues of timing and sales terms, the Department supported its given approaches with convincing rationales and reasonable adjustments. Accordingly, the court sustains those determinations as well.
All in all, Commerce's methodology itself was reasonable. Commerce reasonably explained its methodology and the reasons that it did not undertake to make the adjustments Jianlong sought. However, as explained above, the court cannot now say that this process resulted in a reasonable comparison. Accordingly, the court remands *1347for further consideration of the sales price of the [ [ ] ] transaction and comparison to Fufeng's relevant sales.
E. Other Factors
Although the court remanded to conduct a totality of the circumstances analysis "sufficiently supported by substantial evidence, explaining how the establishment of Jianlong USA, the timing of the sale, and the sales price support a finding that the transaction in question was, or was not, bona fide ," Jianlong I , 279 F.Supp.3d at 1342, Commerce furthered its analysis by considering additional factors beyond those three referred to in the court's remand order. See, e.g., Final Remand Results at 22 (commercial quantities); id. at 39 (expenses arising from the sale). The totality of the circumstances test is a flexible one, allowing the Department to prioritize certain factors depending on the context, but it at least requires that Commerce consider as many factors as are relevant.
Jianlong now challenges a new, additional factor considered by Commerce: "that [ [ ] ] ... did not purchase xanthan gum after the sale in question ...." Final Remand Results at 23. Commerce found it atypical that a firm like [ [ ] ] would purchase the subject merchandise from Jianlong at a high price given the circumstances.8 Id. Jianlong maintains that it was improper for Commerce to consider these factors because they relate not to Jianlong's sale, but "to the customer's own situation and to the developing market situation of xanthan gum downstream products in the United States." Pl.'s Comments at 12. The court sees no reason to disturb the normal discretion afforded Commerce to rely on "any other factor the [Department] determines to be relevant ...." 19 U.S.C. § 1675(a)(2)(B)(iv)(VII). Jianlong suggests neither the ways in which Commerce's determination here lacks substantial evidence nor why the Department's reasoning was arbitrary or capricious. Accordingly, the court sustains Commerce's finding as to this additional factor.
III. Rejection of Certain Documents
Last, the court reaches Commerce's rejection of Exhibits 1, 2, and 3 submitted to clarify the subset of Fufeng sales placed on the record. As the court held previously, "Commerce's characterization of Exhibits 1, 2, and 3 as mere confirmation of factual information is unreasonable." Jianlong I , 279 F.Supp.3d at 1341. In this proceeding, Commerce does not appear to have altered its position but asks that the court sustain its rejection of this documentation anyway. As its decision reflects an abuse of discretion, the court will not do so and remands for additional proceedings.
The court reviews Commerce's rejection of information offered to clarify for an abuse of discretion, which may be evidenced by a decision that "is clearly unreasonable, arbitrary, or fanciful, is based on an erroneous conclusion of law, rests on clearly erroneous fact findings, or follows from a record that contains no evidence on which Commerce could rationally base its decision." An Giang Fisheries Imp. & Exp. Joint Stock Co. v. United States , 41 CIT ----, ----, 203 F. Supp. 3d 1256, 1288 (2017) (citing Gerritsen v. Shirai , 979 F.2d 1524, 1529 (Fed. Cir. 1992) ). Because it is *1348clearly unreasonable for Commerce to reject the documentation in question on the grounds invoked by the agency, the court remands for further consideration in accordance with this opinion.
In its preliminary findings, Commerce found that, based on comparator sales, the price of Jianlong's sale was atypically high. See Prelim. Sales Analysis at 4. As part of that price comparison, the Department placed new information on the record as to sales reported by mandatory respondents in AR2, including Fufeng. See id. ; see also Deadline for Submission of Comments on New Factual Information, ECF No. 50, Confidential J.A. Tab 11, P.R. 137 (Mar. 17, 2016). Jianlong challenged Commerce's sales price finding, arguing that the Department should take a more holistic view of the sales price comparison and should place on the record "the full sales data reported by the mandatory respondents." Jianlong's Resp. to CP Kelco's Request to Reject Jianlong's Submission 2, ECF No. 50, Confidential J.A. Tab 12, P.R. 141 (Apr. 4, 2016) ("Jianlong's Resp. to Request to Reject Submission"). Doing so, Jianlong argued, would clarify the input sales and allow that sales data to serve as a reliable means of comparison. Id. Commerce itself described Exhibits 1, 2, and 3 as "contain[ing] the full U.S. sales databases and responses to Section C of the Department's questionnaire of the mandatory respondents in AR2 and the Section A supplemental questionnaire response of one of the mandatory respondents in AR2," but nevertheless rejected Jianlong's submission as a mere confirmation of the information the Department had already placed on the record. Rejection Mem. 1-2, ECF No. 50, Confidential J.A. Tab 15, P.R. 148 (Apr. 25, 2016).
Here, Commerce purports to give additional reasons for the rejection of this documentation,9 stating that Jianlong failed to show that the information placed on the record was "not understandable or comprehensible" and therefore in need of clarification. Final Remand Results at 54. The Department rebukes Jianlong's alleged failure to "explicitly state[ ] that there were errors that needed to be corrected," id. ,10 but Commerce has missed the forest for the trees. In painting Jianlong's submission as one aimed at "confirm[ing] the accuracy of the subset of sales from AR2 which Commerce placed on the record," id. at 25, the Department overlooks how Exhibits 1, 2, and 3 may clarify the nature of the input sales. Moreover, Commerce appears to demand that Jianlong meet the artificial burden of "assert[ing] that there were errors in the data Commerce used." Def.'s Comments at 27.
The Department has seemingly ignored-yet again-Jianlong's stated purpose *1349for submitting Exhibits 1, 2, and 3: "the Department's claim that it compared [Jianlong's] sale under review to sales of similar subject merchandise, with similar sales terms, cannot be relied upon unless the full sales data reported by the mandatory respondents (and a complete description and examination of those data) are available for comparison." Jianlong's Resp. to Request to Reject Submission at 2-3. Jianlong asked not that the entire Fufeng dataset be accepted so that the parties could confirm what was already known, but so that the information could clarify at least the means by which the Department selected the information on which it relied.
The Department's explicit justification for rejecting Exhibits 1, 2, and 3 relies on its view that those documents can only serve to confirm that which is already on the record. Given the circumstances, that is clearly unreasonable and is an abuse of discretion. The full Fufeng AR 2 dataset would certainly serve to clarify which of those sales Commerce selected as input sales.
Further, Commerce's determination implicitly rejected the notion that the entire dataset cannot clarify its selected subset. On its face, such a reading of the regulation's use of the word "clarify" is unreasonable. The regulation is designed to ensure that the Department "obtain the information it needs in its antidumping investigations," see Guangzhou Maria Yee Furnishings, Ltd. v. United States , 29 CIT 1470, 1475, 412 F.Supp.2d 1301, 1306 (2005), and that information certainly can include that from which input data is derived. The regulation does not free Commerce from its duty to support its determinations with substantial evidence nor does it grant the Department carte blanche to reject information simply because the data on the record is not explicitly erroneous.
In a situation such as this, where the Department has left the record bereft of its reasoning and supplementing the record may elucidate Commerce's rationale, rejecting such documentation may be an abuse of discretion. That is the precise situation the court now faces. Accordingly, the Department's decision to reject Exhibits 1, 2, and 3 due to their failure to clarify reflects an abuse of discretion and the court remands for the purposes of verifying that the proper subset was chosen from among the complete Fufeng sales dataset.
The court recommends that Commerce reopen the record on remand and accept Exhibits 1, 2, and 3 as clarifying information. See Thai Plastic Bags Indus. Co. v. United States , --- CIT ----, ----, 895 F.Supp.2d 1337, 1346 (2013) ; see also Qingdao Sea-Line Trade Co. v. United States , Slip Op. 13-102, 2013 WL 4038618, at *5 (CIT Aug. 8, 2013) ("Reopening the record is particularly appropriate when, as here, doing so clearly advances the purposes of the remand."). Such a step seems to be the most straightforward-and quite possibly the best-way to support the selection of input sales with substantial evidence.
In any event, the court remands the case to Commerce with directions to articulate its methodology for choosing the Fufeng subset, with reference to the entire Fufeng AR2 dataset. With that information in tow, the court will be properly situated to assess the reasonableness of Commerce's findings with regard to sales price.
CONCLUSION AND ORDER
For the foregoing reasons, the court remands on a limited basis: with reference to the complete Fufeng AR2 dataset, Commerce shall articulate a reasonable method, supported by substantial evidence and in accordance with law, for selecting the *1350input sales for the purposes of its sales price analysis. As part of that determination, the court recommends that Commerce reopen the record and accept Exhibits 1, 2, and 3 as a means of clarifying the Fufeng subset. Once it has chosen that subset and supported that selection with substantial evidence, Commerce is to employ the same comparison methodology, if appropriate, approved here: 1) to determine whether, with regard to the entire Fufeng AR2 dataset, the sales price was atypical and 2) to make a finding under 19 U.S.C. § 1675(a)(2)(B)(iv) as to the bona fide nature of the transaction in question. That bona fide inquiry shall incorporate the Department's new sales price analysis as well as all other factors here sustained.
Accordingly, it is hereby:
ORDERED that the Rescission is remanded to Commerce for redetermination in accordance with this Opinion and Order; it is further
ORDERED that Commerce issue a redetermination in accordance with this Opinion and Order that is in all respects supported by substantial evidence and in accordance with law; it is further
ORDERED that Commerce explain, with reference to the entire Fufeng AR2 dataset, how it chose the subset of Fufeng [ [ ] ] xanthan gum sales and use that subset to conduct a sales price comparison; it is further
ORDERED that Commerce conduct a "totality of the circumstances" analysis using the findings here sustained combined with its new sales price finding to determine whether the transaction in question was, or was not, bona fide ; it is further
ORDERED that the Department reconsider its rejection of Exhibits 1, 2, and 3; it is further
ORDERED that all other challenged determinations of Commerce are sustained; it is further
ORDERED that Commerce shall have ninety (90) days from the date of this Opinion and Order in which to file its redetermination, which shall comply with all directives in this Opinion and Order; that the Plaintiff and Defendant-Intervenor shall have thirty (30) days from the filing of the redetermination in which to file comments thereon; and that the Defendant shall have thirty (30) days from the filing of Plaintiff's and Defendant-Intervenor's comments to file comments.