Winning Strategies for Section 2(d) Office Actions: What the Data Shows

A Section 2(d) likelihood of confusion refusal is the most common substantive ground for refusing a trademark registration. Under 15 U.S.C. 1052(d), the USPTO will refuse registration if a mark so resembles a previously registered mark that it is likely to cause confusion among consumers. Roughly one in five trademark applications receives this refusal.

The legal standard comes from In re E.I. du Pont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973), which established the thirteen "DuPont factors" that examining attorneys weigh when evaluating confusion. Not all factors are relevant in every case — TMEP 1207.01 makes clear that "each case must be decided on its own facts" and only factors supported by evidence need be considered.

But which arguments actually move the needle? To answer that question empirically, we analyzed 3,459 Section 2(d) office action cases that reached a terminal outcome — registration on the Principal Register or abandonment. The results challenge some conventional assumptions about how to fight likelihood of confusion refusals.

Data current as of April 2026.

The Numbers at a Glance

Metric	Value
Total cases analyzed	3,459
Cases using argue strategies	58%
Cases using comply strategies	42%
Total arguments extracted	10,447
Argue arguments	6,017
Comply arguments	4,430

The 58/42 argue-comply split is significant. More than half of successful applicants chose to fight the refusal with substantive legal arguments rather than simply narrowing their goods or services. But comply strategies — particularly goods amendments — remain the single most common individual tactic, appearing in 3,110 winning cases.

This suggests the most effective approach is often a hybrid: narrow the identification to reduce the apparent overlap, then argue that the remaining goods or services are sufficiently distinct.

Top Winning Argue Strategies

1. Commercial Impression (1,393 cases)

The most frequently used argue strategy. This argument corresponds to DuPont Factor 1 — "the similarity or dissimilarity of the marks in their entireties as to appearance, sound, connotation and commercial impression." Applicants who win argue that the marks, considered as a whole, convey sufficiently different meanings, connotations, or overall impressions to consumers.

The legal standard is important here: the marks must be compared in their entireties, not dissected into individual components. As the Federal Circuit has held, "the commercial impression of a trademark is derived from it as a whole, not from its elements separated and considered in detail." This is why composite marks with shared elements can still have different commercial impressions — the additional words, design elements, or conceptual associations change the overall effect.

Our data shows commercial impression is the most commonly deployed argue strategy, but it is also frequently attempted in losing cases (see below). The difference between success and failure appears to hinge on specificity — winning responses cite concrete textual, visual, or conceptual differences rather than making abstract claims.

Winning case: BADASS WOMEN (SN 88471559) Overcame a Section 2(d) refusal by demonstrating that the overall commercial impression was distinct from the cited registration. Registered on the Principal Register in January 2021, 319 days after the office action response, with no additional office actions required.

Winning case: FIREFLY FUN HOUSE (SN 88602436) Successfully distinguished its commercial impression despite shared terminology with the cited mark. The response demonstrated that the composite mark created a distinct overall impression that consumers would not confuse. Registered in February 2025.

2. Mark Distinction (1,245 cases)

The second most common argue strategy focuses on structural, visual, and phonetic differences between the marks. Where commercial impression addresses the overall meaning and feel, mark distinction zeroes in on objective differences in appearance, sound, and structure — the first three components of DuPont Factor 1.

This argument works best when the marks have measurable differences: different word counts, different syllable structures, different leading elements, or the presence of design features that fundamentally change the visual impression. It is weaker when the marks are phonetically similar or when the dominant portion of the applicant's mark is identical to the cited registration.

Winning case: SECUREKNOX (SN 88370540) Argued that the combination of elements created a structurally and phonetically distinct composite that consumers would not confuse with the cited registration. The portmanteau structure differentiated it from marks sharing one component. Registered in May 2021, 398 days after responding.

Winning case: THE L.A.B. LYRICS ARTS BEATS (SN 88421878) Used the acronym structure and full expansion to distinguish the mark. The response emphasized that the mark as a whole — with its dual reading of "L.A.B." and its expanded descriptor — communicated a fundamentally different message. Registered in March 2021.

3. Goods Coexistence (859 cases)

This argument targets DuPont Factor 2 ("the similarity or dissimilarity and nature of the goods or services") and Factor 3 ("the similarity or dissimilarity of established, likely-to-continue trade channels"). Even when marks are similar, applicants can prevail by demonstrating that the goods or services are sufficiently different in nature, purpose, or channels of trade.

Under TMEP 1207.01(a)(vi), the examining attorney must consider whether the goods are related enough that consumers encountering them under similar marks would assume a common source. Applicants can rebut this by showing distinct trade channels, different classes of purchasers, or fundamentally different product categories.

Winning case: GIANTS (SN 79324750) Despite sharing a common term with the cited registration, successfully argued that the goods operated in sufficiently distinct trade channels and targeted different consumer segments. Registered in June 2023, 354 days after the response.

Winning case: PROGRESS POWERED BY ECOVADIS (SN 79388908) Argued that the applicant's sustainability assessment services were sufficiently different from the cited registration's services. The response demonstrated distinct trade channels and purchasing contexts. Registered in March 2026 in just 122 days.

Winning case: ZENPLUS (SN 79412789) Demonstrated that the goods coexisted in the marketplace without actual confusion, leveraging DuPont Factor 8 (length of time and conditions under which marks have been used concurrently without evidence of actual confusion). Registered in January 2026.

Top Comply Strategies

Goods Amendment (3,110 cases)

The most common strategy of any kind. Narrowing the identification of goods or services to carve out the overlap with the cited registration directly addresses the examiner's concern: if the goods no longer overlap, the DuPont Factor 2 analysis changes fundamentally. Under TMEP 1402.01, applicants may clarify or limit their identification but may not expand it beyond the original scope.

This is often the fastest path to registration. Many practitioners recommend combining a goods amendment with substantive arguments — narrow first to reduce the apparent conflict, then argue that the remaining goods are distinct enough to avoid confusion.

Class Deletion (523 cases)

When a multi-class application includes a class that directly conflicts with the cited registration, deleting that class eliminates the conflict entirely. This is a more aggressive concession — the applicant gives up coverage in that class — but it resolves the refusal without requiring the examiner to evaluate substantive arguments.

What Doesn't Work

Our analysis of losing cases — applications that received Section 2(d) refusals, responded with arguments, and ultimately abandoned — reveals which strategies fail most often.

Argument	Usage in Losing Cases	Avg Effectiveness
Commercial impression	4,046 cases	0.47
Goods amendment	3,021 cases	0.50
Mark distinction	3,087 cases	0.43
Weakness of cited mark	2,231 cases	0.51
Goods coexistence	2,093 cases	0.44

The pattern is clear: mark distinction is the argument most likely to fail, with an average effectiveness score of just 0.43 in losing cases. When applicants lead with "our marks look different" against an examiner who has already concluded they are similar, the argument frequently falls flat. The examiner has already compared the marks and reached the opposite conclusion — restating the differences without new evidence or a new legal framework rarely changes the analysis.

Commercial impression arguments also fail frequently in losing cases (0.47 effectiveness). However, the gap between winning and losing commercial impression arguments is narrower, suggesting this is a viable strategy that requires strong execution — not a fundamentally weak approach.

Weakness of cited mark arguments have the highest effectiveness even in losing cases (0.51), suggesting this is a legitimate strategy that fails primarily on the facts rather than the legal framework. When a cited mark genuinely has limited scope or strength, this argument can be powerful.

Key Takeaways for Practitioners

1. Don't rely on mark distinction alone. It's the argument most likely to fail. If you're going to argue the marks are different, pair it with goods coexistence or commercial impression arguments that shift the analysis to different DuPont factors.

2. Goods amendment is the single most effective tactic. Over 3,100 winning cases used it. Under TMEP 1402.01, you can narrow your identification to carve out the overlap. If you can do this without losing essential commercial coverage, do it first.

3. The best responses combine argue and comply. The 58/42 split across winning cases means many successful applicants both narrowed their goods AND argued on the merits. Amend first to reduce the overlap, then argue that what remains is sufficiently distinct — this combination is more effective than either approach alone.

4. Commercial impression wins when it's specific. Don't just assert the marks create different impressions — explain exactly why, with reference to specific words, design elements, conceptual associations, or connotation differences that change the overall commercial impression under DuPont Factor 1.

5. Use 3-5 focused arguments. Our data across all refusal types shows that 3-5 arguments is the optimal range. Responses with more than 5 arguments correlate with higher failure rates — they suggest the applicant is throwing everything at the wall rather than building a focused, coherent case.

6. Consider consumer sophistication. DuPont Factor 4 addresses "the conditions under which and buyers to whom sales are made." For specialized or expensive goods, consumers exercise greater care in purchasing, making confusion less likely even with similar marks. This factor is underutilized in our dataset.

Related Strategy Guides

• How Trademark Applicants Actually Win Office Actions: A Data-Driven Analysis — the full statistical overview across all 14 refusal types
• Winning Strategies for Identification of Goods/Services Refusals — the refusal type most often paired with Section 2(d)
• Winning Strategies for Disclaimer Refusals — frequently co-occurring with likelihood of confusion

How GleanMark Can Help

GleanMark analyzes your specific office action against this dataset of 3,459 Section 2(d) cases to recommend the strategies most likely to succeed. The system identifies relevant winning precedents, evaluates your DuPont factors, and drafts a response that leads with the arguments our data shows work best for cases like yours.

Try it with your next Section 2(d) refusal — the analysis is complimentary during the beta.

---

This analysis is based on 3,459 Section 2(d) office action cases analyzed by GleanMark's winning arguments pipeline. Only verified Principal Register registrations are included. Data is updated quarterly. For the legal framework underlying these strategies, see TMEP 1207.01 and In re E.I. du Pont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973).