Skip to content
OBBBA

GILTI is now NCTI: the OBBBA rewrite of §951A

OBBBA renamed GILTI to NCTI inside §951A, dropped the §250 deduction to 40%, eliminated the QBAI offset, and cut the §960(d) FTC haircut to 10%. Here's the substantive-vs-cosmetic boundary, the effective-rate math, and where the AI tools are getting this wrong.

AI Tax Practitioner Editorial

Published ·17 min read ·Last reviewed

The lede in one paragraph: OBBBA renamed GILTI to NCTI in §951A and restructured the computation at the same time. The audience reflex is to assume cosmetic relabel — that read is wrong. The rename is statutory and real, the deduction percentage dropped, the QBAI offset that defined the “intangible” framing was struck, and the FTC haircut improved. The net economic effect depends heavily on the CFC’s effective foreign tax rate — high-tax CFCs see a meaningfully better FTC picture; low-tax or zero-tax CFCs face higher effective US rates on the tested income. This is the working-practitioner read of what changed, what the transition workflow looks like, which planning patterns survive, and what the AI tools are getting wrong.

The 30-second answer

Seven facts a practitioner advising any US shareholder of a CFC needs before briefing the client or running the numbers:

  • The rename is statutory, not regulatory. OBBBA §70323 amended IRC §951A so that the statutory title now reads “Net CFC tested income.” The Treasury Regulations under §951A — including the tested-income computation rules, the high-tax exclusion election, and the allocation framework — have not been re-issued.
  • The §250(a)(1)(B) deduction dropped from 50% to 40%. OBBBA §70321 amended IRC §250. At the 21% §11(b) corporate rate, the effective US rate on NCTI moves from 10.5% (pre-OBBBA) to roughly 12.6% (post-OBBBA). The TCJA-scheduled step-down would have taken the deduction to 37.5% and the effective rate to 13.125% absent OBBBA — so OBBBA improved the rate relative to the step-down but worsened it relative to the legacy TCJA rate.
  • The QBAI / DTIR offset was eliminated. Under TCJA, the tested income inclusion was reduced by the Deemed Tangible Income Return — 10% of Qualified Business Asset Investment. OBBBA §70323 struck this — specifically by striking the §951A subsections containing the DTIR formula. NCTI now equals the CFC’s gross tested income without any return-on-tangibles deduction. Asset-heavy CFC structures that historically benefited from the QBAI offset face wider inclusion bases under the new framework, with no replacement mechanism.
  • The §960(d)(1) FTC haircut dropped from 20% to 10%. OBBBA §70312 amended IRC §960. Pre-OBBBA, only 80% of the CFC’s foreign taxes allocable to tested income were creditable against the US inclusion. Post-OBBBA, 90% are creditable. For CFCs in high-tax jurisdictions, this improvement materially offsets the deduction reduction. Confirmed in IRS Notice 2025-77.
  • A new mid-year transition date applies to certain PTEP distributions. A new §960(d)(4) disallows 10% of foreign taxes attributable to §951A Previously Taxed Earnings and Profits distributions occurring after June 28, 2025 — confirmed by IRS Notice 2025-77.
  • The full TY2026 regime hits for tax years beginning after December 31, 2025. TY2025 returns being filed in 2026 still use GILTI, the 50% deduction, and Form 8992 as currently published. TY2026 returns — filed in 2027 — are the first to carry the NCTI label, 40% deduction, no QBAI offset, and the improved FTC haircut.
  • The AI tools are unreliable on this for now. Every major tax-research tool is structurally exposed to the statute-vs-regulation asymmetry. The detail is in the AI tools section below.

The pre-OBBBA GILTI framework — what’s being replaced

TCJA created the GILTI regime in 2017 as a minimum US tax on the earnings of US-owned CFCs. The conceptual design was straightforward: if a US shareholder’s aggregate CFC tested income exceeded a 10%-of-QBAI “routine return” floor, the excess was a GILTI inclusion under §951A(b). The C-corp US shareholder then got a §250(a)(1)(B) deduction of 50% against that inclusion, producing an effective US rate of 10.5% at the 21% corporate rate. A foreign tax credit under §960(d) was available against the inclusion, but the pre-OBBBA §960(d)(1) haircut allowed only 80% of the CFC’s foreign taxes on tested income to be creditable. The combination — 10.5% effective rate, 80% FTC passthrough — meant CFCs in zero- or very-low-tax jurisdictions generated meaningful US tax while CFCs paying foreign taxes near or above 10.5% could often eliminate or minimize the US GILTI liability through §960(d) credits.

The “intangible” name was always something of a misnomer. TCJA’s QBAI floor notionally stripped out the return on tangible assets on the theory that only intangible-income-generating activity deserved the preferential rate. But the computation didn’t track intangible assets directly — it used a rough 10%-of-tangible-asset-value floor as a proxy. Asset-heavy manufacturers or distributors with substantial tangible-property bases got proportionally more QBAI offset. IP-holding structures with no tangible assets got none. The result was that two CFCs with the same tested income could produce very different US inclusions depending on their asset composition.

TCJA scheduled the §250(a)(1)(B) deduction to step down to 37.5% for tax years beginning after December 31, 2025 — pushing the effective rate to 13.125%. The scheduled step-down kept the GILTI regime in the practitioner-anxiety lane through the first half of 2025, alongside the parallel FDII step-down. OBBBA landed on both at once.

The §951A high-tax exclusion election under Treas. Reg. §1.951A-2(c)(7) allowed US shareholders to exclude from tested income any CFC tested income subject to a foreign effective tax rate exceeding 90% of the US corporate rate — historically 18.9% at the 21% corporate rate. That election survived OBBBA substantively. With the post-OBBBA US effective rate on NCTI now at roughly 12.6%, the high-tax exclusion threshold has implicitly shifted in economic terms — the election now kicks out high-tax CFC income from a 12.6% US rate exposure rather than a 10.5% exposure.

What OBBBA actually changed — the rename plus the substantive recalibration

The §951A rename and computation changes sit in OBBBA §70323. The §250 deduction-rate change is in §70321. The §960 FTC haircut change is in §70312. The headline numbers, QBAI elimination, FTC improvement, and mid-year PTEP transition date are summarized in the table below. Cornell LII’s current statutory text at 26 USC §951A and the post-OBBBA practitioner-canonical commentary (WilmerHale, BDO, Mayer Brown, Cooley, Holland & Hart, RSM) converge on the same parameters.

The §951A / §250(a)(1)(B) framework before, during the scheduled step-down, and after OBBBA
ParameterPre-OBBBA (TY < 1/1/2026)Scheduled (TY ≥ 1/1/2026, absent OBBBA)Post-OBBBA (TY ≥ 1/1/2026)
Statutory labelGILTI (Global Intangible Low-Taxed Income)GILTI (no rename scheduled)NCTI (Net CFC Tested Income)
§250(a)(1)(B) deduction50%37.5%40%
Effective US rate at 21%10.5%13.125%~12.6%
QBAI / DTIR offset10% of QBAI(same)eliminated
§960(d)(1) FTC haircut20%20%10%
Creditable foreign taxes80% of allocable tested-income taxes80%90%
§904(b)(5) allocation ruleInterest + R&E allocated against NCTI basket(same)No longer allocated
§960(d)(4) PTEP distribution——10% disallowance for PTEP distributions after June 28, 2025
High-tax exclusion electionTreas. Reg. §1.951A-2(c)(7) at >18.9%(same)Survives; threshold same; economic effect shifted

Three observations worth holding before drafting any client memo.

First, the framing depends entirely on the baseline. A client comparing post-OBBBA to pre-OBBBA sees a rate increase: 10.5% to 12.6%. A client comparing to the scheduled step-down sees a rate cut: 13.125% would have become 12.6%. Both are technically accurate. Holland & Hart’s analysis of the FDDEI side uses the same dual-baseline framing; it applies with equal force here. The client will anchor on whichever number they hear first — document both in the engagement memo.

Second, the QBAI elimination and the FTC improvement interact, and the net economic effect runs in opposite directions depending on the CFC’s asset and tax profile. For asset-heavy CFCs with substantial QBAI, the QBAI elimination widens the inclusion base — meaning more income is now NCTI than was GILTI under the same facts. The FTC haircut improvement from 20% to 10% partially or fully offsets this, but only if the CFC is paying material foreign taxes. A low-tax or zero-tax CFC loses the QBAI offset with no FTC offset — net effect is more US tax. A high-tax CFC that was generating large §960(d) credits at 80% passthrough now passes through at 90% — the FTC improvement may more than offset the deduction reduction, leaving the US shareholder better off overall. The Mayer Brown and WilmerHale analyses both make this distinction explicitly.

Third, the new §904(b)(5) allocation change mirrors the FDDEI-side allocation rule. Interest and R&E expenses are no longer allocated against the NCTI foreign tax credit basket. This widens the NCTI FTC basket, meaning more foreign taxes stay creditable rather than being pushed into excess-credit limitation. For US shareholders of CFCs in foreign jurisdictions with significant tax expense, this is a genuine improvement in the FTC utilization picture.

Who is affected and the math

NCTI is a regime for US shareholders of CFCs. Under §951A(a), a US shareholder includes in gross income its pro-rata share of the CFC’s net tested income for the year. The US shareholder is defined in §951(b) as a US person who owns 10% or more of the voting stock (or value) of the CFC.

In solo and small-firm practice, the NCTI conversation typically involves one of these client types:

  • A US founder / owner with a foreign operating subsidiary. The most common pattern in technology, e-commerce, and professional-services practices with international revenue.
  • A US C-corp with offshore manufacturing, distribution, or IP licensing through a CFC.
  • A high-net-worth individual making a §962 election to access the §250 deduction and §960(d) credits against their NCTI inclusion.
  • A client exiting a CFC structure and facing GILTI recapture or inclusion mechanics under §951A on the final tested income.

The effective-rate math for a calendar-year C-corp US shareholder in TY2026, using representative numbers:

Example — low-tax CFC. CFC has $1,000 of tested income. No QBAI (or QBAI is now irrelevant — full $1,000 is NCTI). No material foreign tax. US shareholder’s NCTI inclusion: $1,000. §250(a)(1)(B) deduction: 40% × $1,000 = $400. Taxable NCTI: $600. US tax at 21%: $126. Effective US rate: 12.6%. Under legacy GILTI with $200 of QBAI: tested income was $1,000 minus $20 DTIR = $980 GILTI. §250 deduction 50% = $490. Taxable: $490. US tax: $102.90. Effective rate: ~10.3%. Net result: $23.10 more US tax per $1,000 of tested income for the low-tax CFC under post-OBBBA rules.

Example — high-tax CFC. CFC has $1,000 of tested income and pays $250 of foreign taxes (25% effective rate). Post-OBBBA: NCTI inclusion $1,000. §250 deduction $400. Taxable NCTI $600. US tax before FTC: $126. §960(d) credit: 90% × $250 = $225. Net US tax: max(0, $126 - $225) = $0 (FTC eliminates the US liability). Under legacy GILTI: §250 deduction $500. Taxable GILTI: $500. US tax before FTC: $105. §960(d) credit: 80% × $250 = $200. Net US tax: max(0, $105 - $200) = $0. Both result in zero US tax, but the post-OBBBA path passes 90% of the foreign taxes into the FTC basket versus 80%, which matters for overall FTC utilization if the shareholder has other FTC limitation constraints.

For the §962-electing individual, the same mechanics apply. The election causes the §250 deduction to be available against the NCTI inclusion, and §960(d) credits flow through at the 90% rate for TY2026 inclusions. Whether the §962 election remains optimal in TY2026 requires re-running the pre-election vs post-election comparison under the new deduction percentage and FTC passthrough rate. The Withum analysis of post-OBBBA §962 considerations addresses this directly; the Forvis Mazars analysis is a parallel reference that walks through the worked numbers.

Planning angles practitioners should re-run

Several planning patterns that worked under GILTI need to be re-evaluated under NCTI.

High-tax exclusion election under Treas. Reg. §1.951A-2(c)(7). The election excludes from tested income any CFC income taxed above 90% of the US corporate rate — 18.9% at the 21% rate. That threshold did not change. But the economic calculus of making the election shifted. Under GILTI at a 10.5% effective rate, a CFC paying 12% foreign tax could generate positive US tax (depending on QBAI), making the election valuable at a relatively low foreign-tax rate. Under NCTI at a 12.6% effective rate, the breakeven foreign tax rate where the FTC alone eliminates the US liability has risen. Clients with CFCs in medium-tax jurisdictions — say 12-18% effective foreign rates — should re-model whether the election is optimal under the new framework or whether the 90% FTC passthrough now covers the US exposure without the election.

QBAI-optimized structures. Some GILTI planning under TCJA involved increasing QBAI (investing in tangible CFC assets) to expand the DTIR offset and reduce the tested income. That planning logic evaporates for TY2026 and beyond. Structures that were maintained primarily to build QBAI should be reviewed. Conversely, asset-light CFC structures that previously had minimal QBAI and thus minimal DTIR offset face no structural change on the inclusion-calculation side — they were already including their full tested income as GILTI and will continue to do so as NCTI.

FDDEI coordination. The §250 deduction is a single computation. A C-corp US shareholder claiming both §250(a)(1)(A) deductions on FDDEI (the domestic export side) and §250(a)(1)(B) deductions on NCTI inclusions runs both through the §250(a)(2) taxable-income limitation. That limitation scales back the §250 deduction if the combined FDDEI + NCTI deduction exceeds taxable income computed without §250. Under the new post-OBBBA percentages — 33.34% on FDDEI and 40% on NCTI — the pro-rata reduction mechanic survives, but the absolute deduction amounts shifted. Practitioners with C-corp clients running at low taxable income relative to the combined deduction pool should re-model the §250(a)(2) cap.

§1411 NIIT interaction. The §962-electing individual’s NCTI inclusion and §960(d) credits interact with the net investment income tax under §1411. Generally, NCTI inclusions are not themselves net investment income (NII) under the §1411(c)(1)(A) structure, but the downstream PTEP distributions from a CFC can enter the NII analysis depending on the §1411(b)(1) rules. The higher FTC passthrough (90% vs 80%) does not directly affect the §1411 analysis on inclusions, but practitioners should verify whether their clients’ fact patterns involve PTEP distributions in the §960(d)(4) window — distributions after June 28, 2025 take the 10% disallowance — and whether those distributions enter the NII calculation.

§245A interaction. The §245A dividends-received deduction is available to a C-corp US shareholder on actual dividends from a specified 10%-owned foreign corporation, but only to the extent the dividend is not PTEP. Post-OBBBA, the interplay between the wider NCTI inclusion base (no QBAI offset), the §960(d) FTC improvement, and the §245A DRD on non-PTEP distributions is worth re-modeling for any CFC structure that was optimizing between §245A and §951A. The BDO and WilmerHale analyses both flag the NCTI base-widening as the dominant structural change in multi-CFC structures.

State conformity. States vary widely in their treatment of §951A inclusions. Some conform to federal GILTI / NCTI and will adopt the OBBBA changes by rolling reference. Others have decoupled from GILTI entirely. New Jersey and Iowa — both of which issued explicit guidance on the FDDEI / NCTI rename — are the clearest state-level conformity data points as of May 2026. For any multi-state C-corp with CFC tested income, the state-conformity question is a separate computation from the federal NCTI analysis.

Where the AI tools currently get this wrong

This is the section practitioner readers landed on this article to find. The honest answer as of May 2026: every retrieval-augmented and non-retrieval AI tax-research tool a US practitioner realistically uses is structurally exposed to the same failure mode on NCTI — and the shape of the failure is the same as the FDII → FDDEI failure mode.

The structural problem. Cornell LII’s text of 26 USC §951A now reads “Net CFC tested income” with the OBBBA-amended body text reflecting the renamed provision and the elimination of the DTIR offset. But Treas. Reg. §1.951A-1 and the entire §1.951A regulation series — the operational machinery for computing tested income, high-tax exclusion, and §960(d) credits — still uses “GILTI” throughout and has not been re-issued post-OBBBA. IRS Form 8992 and its instructions still bear the GILTI label. Any AI tool that retrieves across the current statutory text, the pre-OBBBA regulations, and the current IRS form will surface mixed terminology by design. This is not a failure of the tool; it is a reflection of the current state of primary authority.

The knowledge-cutoff overlay. The OBBBA was signed July 4, 2025. General-purpose LLMs trained before that date have zero pre-training awareness of the §951A rewrite. Models trained through late 2025 have some corpus coverage, but that corpus is dominated by pre-rename GILTI commentary from the 2017-2024 period. The CBS News piece quoting American University’s Caroline Bruckner documents the general OBBBA-era currency problem; the §951A rename sits squarely in the same lane.

Where the named tax-AI tools sit in May 2026.

CoCounsel (Thomson Reuters) runs over the Checkpoint corpus. Checkpoint’s editorial layer is updated proactively, and post-OBBBA international tax content is generally current at the editorial tier. The residual risk is mixed-corpus blending — Checkpoint contains both pre-OBBBA GILTI analysis and post-OBBBA NCTI analysis, and the generation layer may surface or blend older content unless the prompt specifies the controlling tax year. CCH AnswerConnect has explicitly retitled its §951A reference content to reflect the NCTI rename, consistent with the CCH AnswerConnect FDDEI and NCTI page. Bloomberg Tax has similarly acknowledged the rename in its headline GILTI / NCTI comparison page.

Blue J is statute-anchored, which makes it structurally best-positioned among the named tax-research tools to return the correctly-rebranded NCTI terminology. But any Blue J answer pulling regulatory-level detail will surface GILTI language, because Treas. Reg. §1.951A-1 still says GILTI and the regulation is the operational authority for the computation. TaxGPT and Hive Tax are more variable; both are reported by practitioners to lean on IRS-publication-level content over statute, and IRS Form 8992 and its instructions still carry the GILTI label.

Consumer general-purpose LLMs — ChatGPT, Claude without web search, Perplexity with stale indices, Gemini, Copilot — are the highest-risk tools on this question. The Wealth AI Audit documented that all five major consumer engines returned stale TCJA-era answers on post-OBBBA estate-planning questions. The §951A rename and rate change is the same structural problem. A consumer LLM returning “GILTI, 50% deduction, 10.5% effective rate” on a TY2026 §951A query is giving the wrong framework.

The practitioner workflow implication. AI-tool output on NCTI should be treated as a first-draft pointer, not a citable answer. The verification step against the current Cornell LII text of §951A and §250 is non-optional. Circular 230 §10.22 treats AI-tool output as “others’ work product” — the practitioner who signs the return owns the answer regardless of which tool produced it.

§10.22 due-diligence under the rename

Circular 230 §10.22 imposes a due-diligence obligation on any practitioner preparing, approving, or advising on a position that materially affects a return. The proposed Circular 230 modernization adds an explicit technological-competency requirement at proposed §10.35 — not yet final as of May 2026, but the direction is set. (Federal Register, December 26, 2024 Notice of Proposed Rulemaking.)

For §951A / NCTI work specifically, due-diligence in the post-OBBBA environment means:

  • Apply the correct effective date. TY2025 returns use GILTI, 50% deduction, and Form 8992 as published. TY2026 returns use NCTI, 40% deduction, no QBAI offset, and the 90% FTC passthrough. The cutover date — tax years beginning after December 31, 2025 — is the operative line. For fiscal-year CFCs with non-calendar tax years, the effective date turns on when their tax year begins, not the January 1, 2026 calendar date.
  • Re-run the high-tax exclusion election analysis. The election threshold at Treas. Reg. §1.951A-2(c)(7) is 90% of the US corporate rate — unchanged. But the economic breakeven for making the election shifted because the post-OBBBA US effective rate is 12.6%, not 10.5%. Clients for whom the election was previously non-optimal may now find it optimal; clients for whom it was optimal may find the 90% FTC passthrough renders it unnecessary. Verify the election position on every CFC in the NCTI basket.
  • Verify the QBAI offset assumption. If any NCTI / GILTI workpaper uses a DTIR offset, flag it. For TY2026 returns, the QBAI offset does not exist. Prior-year workpapers that show a DTIR reduction in the tested income computation are correct for TY2025 and earlier; they are wrong templates for TY2026.
  • Apply the §960(d)(4) PTEP date rule. Distributions of §951A PTEP occurring after June 28, 2025 take a 10% disallowance on the associated foreign taxes. Workpapers should reflect the distribution date on any PTEP distribution in the relevant window.
  • Verify AI-tool output against statute. Any AI-tool answer returning “GILTI, 50% deduction, QBAI offset” on a TY2026 §951A question is giving the pre-OBBBA framework. Any answer returning “NCTI, 40% deduction, no QBAI, 90% FTC” is structurally correct — but verify the percentages and mechanics against the current Cornell LII §951A and §250 text before relying on it.

The §7216 consent discipline applies to NCTI work to the same degree it applies to any other client-return position. Pasting a CFC’s tested-income schedule, QBAI data, or §960(d) computation into a public LLM almost certainly counts as a §7216 disclosure under Treas. Reg. §301.7216-1(b)(5), regardless of vendor privacy claims. The defensible workflow remains a signed §7216 consent in the engagement letter plus a vendor with a DPA, no-training clause, and SOC 2 Type II.

Frequently asked questions

Is the GILTI → NCTI rename just cosmetic?

No. The rename is statutory and the substance changed alongside it. The §250(a)(1)(B) deduction dropped from 50% to 40%. The QBAI / DTIR offset was eliminated. The §960(d)(1) FTC haircut dropped from 20% to 10%. A new §960(d)(4) PTEP distribution disallowance applies for distributions after June 28, 2025. The reason “GILTI” remains in circulation — in IRS Form 8992, in the §1.951A regulation series, and in most AI-tool outputs — is that the Treasury Regulations and IRS forms have not been re-issued for the statutory rename.

What is the new effective US tax rate on NCTI?

At the 21% §11(b) corporate rate: 21% × (1 - 0.40) = 12.6%, before FTC. The pre-OBBBA GILTI rate was 10.5%. The TCJA-scheduled step-down that would have applied absent OBBBA was 13.125%. Which framing matters depends on the client’s baseline. Factor in the §960(d) FTC passthrough at 90% to arrive at the net US liability — which is zero for most high-tax CFCs.

When does NCTI first appear on a return?

TY2026 returns, filed in 2027, for calendar-year filers. TY2025 returns filed in 2026 still use GILTI, 50% deduction, and Form 8992 as currently published. The §960(d)(4) PTEP distribution disallowance for distributions after June 28, 2025 applies within TY2025 returns where the distribution occurred after that date.

What happened to the QBAI offset?

It was eliminated entirely for TY2026 and subsequent years. NCTI equals the gross tested income of the CFC without any return-on-tangibles deduction. For TY2025 and prior returns, the DTIR / QBAI mechanics still apply. Prior-year GILTI workpapers are correct templates for TY2025; they are wrong templates for TY2026.

Does the §962 election still make sense after OBBBA?

It depends on the fact pattern. The election is available to individual US shareholders of CFCs to access the §250(a)(1)(B) deduction (now 40%) and §960(d) credits (now at 90% passthrough) against their NCTI inclusion. Whether the election is optimal under the new parameters requires running the pre-election and post-election scenarios under the 40% deduction and 90% FTC passthrough. The Withum analysis of post-OBBBA §962 considerations is a good starting point.

Is the high-tax exclusion election still available?

Yes. Treas. Reg. §1.951A-2(c)(7) survived OBBBA substantively. The threshold — 90% of the US corporate rate, currently 18.9% — did not change. But the economic calculation of whether to make the election shifted, because the post-OBBBA US effective rate on NCTI (12.6%) is higher than the pre-OBBBA rate (10.5%). Re-model the election for every CFC in the NCTI basket.

Can I rely on Blue J, CoCounsel, or TaxGPT for NCTI research right now?

Treat their output as a first-draft pointer requiring verification against the current Cornell LII text of §951A and §250. Blue J is structurally best-positioned because its retrieval is statute-anchored. CCH AnswerConnect and Bloomberg Tax have both updated their editorial content to reflect the NCTI rename. Consumer LLMs without primary-source verification should not be relied on for §951A / NCTI planning work.

Sources and citations

All URLs listed for verification. Primary statutory sources are Cornell LII (returns HTTP 200). Practitioner-firm explainers are from post-OBBBA client alerts and analysis published July–September 2025. The article was assembled from statutory text, regulatory authority, IRS forms, and post-OBBBA practitioner-canonical commentary.

Statutory text and Treasury Regulations:

IRS guidance and forms:

Practitioner-canonical post-OBBBA commentary:

State conformity:

AI-tool accuracy and currency:

This is general information, not advice for your engagement. Before relying on any NCTI computation for a client return, verify current parameters against the Cornell LII §951A and §250 text and confirm your AI-tool vendor’s compliance with §7216, Treas. Reg. §301.7216-2, and your engagement letter’s consent provisions. Circular 230 §10.22 due-diligence remains with the practitioner regardless of tool output.

Pricing and feature availability for AI tools current as of May 2026. Verify before purchase decisions.