Benchmarks / SurveyBench / Why Quaymark's customer score fell / Qwen4 Max

Measured by Spring Prompt

Qwen4 Max: Why Quaymark's customer score fell

The decision: whether to reverse the per-shipment fee and bring back the subscription, invest in shipment tracking and proactive delay alerts, or add account managers for mid-sized customers. The brief and the data →

Verdict
✓ Sound research
Research score
100 of 100
Analysis rating
1,373
Head to head, this task
won 9 of 11

Scorecard

Numbers 12 of 12 right

QuestionIts answerTrue
What is the Q3 2026 wave's NPS across all respondents (percentage scoring 9-10 minus percentage scoring 0-6), unweighted?✓12.912.9
Weighted to the shipments-a-month mix of all 4,720 active accounts, what is the Q3 2026 wave's NPS?✓17.717.8
Re-weighted to the shipments-a-month mix of the Q1 2026 wave's respondents, what is the Q3 2026 wave's NPS?✓20.320.3
Among respondents who scored 0-6 and answered the open question, what percentage of answers mention tracking (not being able to see where shipments are) or late or missed collections or deliveries?✓68.9%68.9%
How many of the open-text answers mention the per-shipment fee or its cost?✓2121
What is the 95% margin of error, in percentage points, on the Q3 2026 wave's unweighted share of respondents scoring 9-10?✓4.1 points4.12 points
Among accounts booking 100 or more shipments a month, is the change in the share scoring 9-10 between the Q1 2026 and Q3 2026 waves statistically significant at the 95% level? (1 for yes, 0 for no)✓NoNo
What percentage of the accounts invited to the Q3 2026 wave completed it?✓11%11%
Among Q3 2026 respondents booking fewer than 20 shipments a month, what percentage said the per-shipment fee is worse for their business than the subscription?✓49.2%49.2%
Among accounts booking fewer than 20 shipments a month, by how many points did NPS fall between the Q1 2026 and Q3 2026 waves?✓3.9 points3.9 points
How many Q3 2026 respondents who send shipments to the Republic of Ireland scored 0-6?✓1313
Weighted to the shipments-a-month mix of all 4,720 active accounts, about how many accounts would score 0-6?✓1,0001,001

Traps in the data 8 of 8 handled

  • ✓The Q1 wave (+24) was phone interviews by account managers; the Q3 wave (+13) was online. Interviewer-administered scores, especially when the interviewer is the customer's own account manager, run higher, so part of the fall may be the change of method rather than a change in loyalty.
  • ✓Accounts booking fewer than 20 shipments a month are 61% of Q3 respondents against 30% in Q1 and 40% of active accounts, and they score lowest (Q3 NPS +5). Re-weighted to Q1's mix the Q3 NPS is +20.3, and to the active-account mix +17.8, so much of the 11-point fall is a change in who answered.
  • ✓The commercial director believes the fee caused the fall. Among detractors who explained their score, 69% mention tracking or delays and 18% the fee; only 39% of all respondents say the fee is worse than the subscription (49% of small accounts, 15% of the largest).
  • ✓Respondents who ship to Ireland (base 14) are nearly all detractors (NPS -93).
  • ✓45% of small accounts say they would be likely to choose a monthly subscription if offered: a stated intention about a hypothetical, with no price given.
  • ✓521 of 4,720 invited accounts completed (about 11%); unhappy customers and small accounts may be more likely to respond online.
  • ✓Among accounts booking 100 or more a month (bases 98 and 48), the promoter share moved from 49% to 52%, NPS +38 to +38: not significant.
  • ✓The 69% figure is a share of the 45 detractors who answered the open question, not of all respondents or all customers.

The analysis

Recommendation

Invest in shipment visibility first: live tracking plus proactive delay alerts, built on fixed haulier event feeds and paired with tighter collection-slot performance management. This is the only option supported by converging, adequately sized evidence (Q4 28.6% dissatisfied, Q7 net 65.6%, 68.9% of detractor comments), and it addresses detractors, passives and the fragile promoter base alike. Do not reverse the per-shipment fee across the board: weighted, only about a third say it is worse for them and a third are indifferent, while 31.2% of 100+ accounts say it is better. Instead, design and price an optional subscription or monthly fee cap aimed at regular low-volume bookers (49.2% of <20 say worse, 44.5% say they would likely switch) and test it as a real priced offer, with margin guardrails, keeping per-shipment as the default. Defer a general account-manager function: 14.2% overall ask for it, and the 31.2% figure among 100+ accounts rests on ~15 people who already give the highest NPS (+38) and want tracking first; if capacity allows, run a small matched pilot rather than a rollout. On Ireland, do not commit spend on 14 responses: pull lane-level operational data and call back the whole small Ireland population, then act on what that shows - the complaints are the same tracking/lateness failures, so the tracking/alerts work is likely to be the remedy anyway. Finally, fix the measurement before the board conversation: the 11-point drop is largely mix and mode, the like-for-like gap is ~3 points and inside noise, and the honest board statement is 'NPS about +18 on the account mix, measured on a new basis; trend not yet established'.

What would change it

Four things would change it. (1) Operational data showing that tracking events are already complete and timely for most hauliers, and that lateness is concentrated in a few carriers - then the priority becomes carrier performance/SLA enforcement rather than a tracking build. (2) A priced subscription test (or billing analysis) showing that a subscription for low-volume bookers is margin-negative, or that take-up is far below the stated 44.5%, or conversely that fee-driven churn among <20 accounts is materially eroding the base - the first would kill the subscription option, the second would justify moving it ahead of tracking. (3) Evidence that Ireland is a genuinely distinct failure mode (customs, lane transit, a specific haulier) affecting a larger, higher-value population than the 2.7%/~127 accounts implied here, with elevated churn - that would justify a dedicated Ireland fix now. (4) Retention or growth data showing 100+ accounts (18% of accounts, likely a much larger share of shipments) are at risk and that human coverage retains them - that would promote the account-manager pilot. A properly designed re-baseline that still shows a large fall in NPS on a like-for-like basis would also change the urgency and the board narrative.

Findings

  1. Do not tell the board loyalty fell 11 points. Most of the gap is sample composition and a change of survey mode, not a measured collapse in customer sentiment; the residual, like-for-like difference is about 3 points and is inside sampling noise.

    High confidenceBase 521

    Q3 online: 35.9% promoters, 23.0% detractors, NPS +12.9 (n=521). Q1 phone: 40.5% / 17.0%, NPS +23.5 (n=412). Re-weighting the Q3 wave to the Q1 respondent volume mix (124/190/98 of 412 = 30%/46%/24%) gives Q3 NPS +20.3, i.e. ~3 points, not 11. The 95% margin on a difference between two NPS of this size is roughly +/-9.5 points. Within bands the changes are small: <20 8.9 to 5.0 (-3.9), 20-99 26 to 21.4 (-4.6), 100+ 38 to 37.5 (-0.5, z=0.35, not significant). Q1 was phone interviews conducted by the customers' own account managers on an AM-chosen list of 640 (64% completion) - a courtesy-bias-prone method; Q3 was self-complete email at 11.0% completion. Q1 also pre-dates the fee change and used account-record volume, not self-report.

  2. All headline numbers must be weighted before use: the achieved sample is dominated by low-volume bookers, which flatters fee anger and depresses NPS. On the true account mix, NPS is about +18 and roughly 1,000 of the 4,720 active accounts are detractors.

    High confidenceBase 521

    Q8 achieved: 61.2% fewer than 20, 29.6% 20-99, 9.2% 100+ (n=521) vs database 40% / 42% / 18%. Weighted NPS = 0.40x5.0 + 0.42x21.4 + 0.18x37.5 = +17.7 (unweighted +12.9). Weighted detractor share = 0.40x26.0 + 0.42x19.5 + 0.18x14.6 = 21.2%, i.e. ~1,000 accounts of 4,720. Weighted 'fee is worse' falls from 39.0% to 32.9%.

  3. The survey does not justify reversing the per-shipment fee for everyone. Fee pain is concentrated in low-volume accounts; high-volume accounts mostly prefer or are indifferent to the new model. Offer a priced subscription option to the segment that wants it rather than switching the whole base back.

    Medium confidenceBase 521

    Q5 (n=521): worse 39.0%, about the same 29.9%, better 20.0%, 7.5% never used the subscription. By band: worse 49.2% of <20 (n=319), 25.3% of 20-99 (n=154), 14.6% of 100+ (n=48); better 31.2% of 100+. Weighted to the account mix: worse 32.9%, same 33.4%, better 22.8% - no majority for reversal. Q6 net 'likely to choose a subscription' 39.5% overall (weighted ~37%): 44.5% of <20, 29.9% of 20-99, 37.5% of 100+; 61.6% among those saying the fee is worse (n=203) but also 27.9% among those saying it is better (n=104) - an unpriced, hypothetical question asked straight after a question framing the change, so treat 39.5% as an upper bound on demand, not a forecast.

  4. The fee mainly caps enthusiasm (7-8 scores); service failures create detractors. Reversing the fee would buy back some passives, not fix the detractor pool - so it is a second-order lever behind service.

    Medium confidenceBase 186

    21 of 186 open answers mention the per-shipment fee or its cost; 13 of those 21 are passives scoring 7-8, 8 are detractors, none are promoters. By contrast 31 of the 45 detractor open answers (68.9%) cite tracking/visibility or late or missed collections/deliveries; only 8 detractor answers mention the fee, and 6 cite support or claims. Q7: lower booking fees 22.6% vs live tracking 41.3% and delay alerts 34.2% (net tracking-or-alerts 65.6%). Open text is self-selected (186 of 521, 35.7%), so these are proportions of comments, not of customers.

  5. Shipment visibility is the weakest part of the experience and the strongest improvement ask across every segment - this is where investment should go first, ahead of pricing reversal or account managers.

    High confidenceBase 521

    Q4 dissatisfied 28.6% (13.4% very + 15.2% fairly) vs Q3 punctuality dissatisfied 22.4%; among detractors 48.4% dissatisfied with visibility (n=120) vs 44.1% with punctuality; only 8.0% of promoters dissatisfied with visibility (n=187). Q7 net tracking-or-alerts 65.6% overall, 69.2% of detractors, 69.6% of passives, 58.8% of promoters, 72.9% of 100+ (n=48). Open text describes statuses stuck at 'booked'/'in transit' for days, blank tracking maps, missing tracking for some hauliers, and constant chasing (e.g. Q094, Q129, Q196, Q331, Q367, Q457, Q521). Closed and open evidence converge, which is why this carries more weight than any single figure.

  6. Tracking and alerts will reduce chasing and anxiety but will not by themselves fix late and missed collections; the investment must be paired with haulier performance management and reliable event feeds, or the NPS gain will disappoint.

    Medium confidenceBase 521

    Q3: 22.4% dissatisfied with punctuality, 10.3% very dissatisfied among <20 (n=319). 29 open answers describe missed or late collections/deliveries (Q009, Q033, Q115, Q150, Q195, Q244, Q283, Q352, Q356, Q367, Q372, Q388, Q448, Q460, Q492 etc.), several with no notification at all. Respondents say tracking data is absent for some hauliers entirely (Q457 'some have a number, some have nothing'; Q521 'tracking just isn't there for most hauliers'), so alert accuracy depends on carrier data quality. 11 answers describe knock-on damage to their own customer promises, including a cancelled order (Q388).

  7. There is no case for a general account-manager layer. At most, run a small targeted pilot for 100+ shipment accounts - and note their first ask is still tracking, not a human contact.

    Low confidenceBase 48

    Q7 'a named account manager' 14.2% overall (n=521), 11.0% of <20 (n=319), 15.6% of 20-99 (n=154), 31.2% of 100+ - about 15 of 48 respondents, with a 95% margin of roughly +/-13 points. The 100+ band already has the highest NPS (+38), is least angry about the fee (14.6% worse, 31.2% better) and picks live tracking most (54.2%). Only one open answer asks for a human owner, and it is about claims ownership (Q486).

  8. Ireland is a signal to investigate, not a sized problem: the -93 NPS rests on 14 people and cannot be reported as an estimate. The complaints are the same tracking/lateness failures seen elsewhere, so treat it as a lane-reliability question and check operational data before spending.

    Low confidenceBase 14

    Q9: 2.7% ship to Ireland (14 of 521; 95% CI on that share 1.3%-4.1%, i.e. roughly 62-194 of 4,720 accounts). Of the 14, 13 score 0-6 and none score 9-10; one respondent moves that NPS by ~14 points. 7 Ireland respondents wrote open answers (Q013, Q033, Q115, Q150, Q295, Q296, Q339); 6 cite tracking or late/missed collections and one the fee - none mention customs or border-specific issues. Non-Ireland NPS is +16 vs +12.9 total, so even on these figures Ireland accounts for at most ~3 points of the overall score; the problem is broad, not Irish.

  9. Loyalty is thin and transactional: the only strong positive is two-minute booking. That is a fragile basis for advocacy and means the promoter base is also at risk from the visibility gap - fixing service protects existing promoters, not just detractors.

    Medium confidenceBase 186

    79 of 186 open answers praise booking speed/ease, overwhelmingly from 9-10 scorers, often with nothing else said (Q114 'Fast. That's really it.', Q212, Q332). 41.1% of the sample are passives (n=214); 17 open answers are 'does the job / nothing special' and 5 say they are too new to judge. Only 19.8% of promoters choose 'none of these' at Q7 - 58.8% still want tracking or alerts. Several 9-10 answers are generic ('no complaints really', Q430), so promoter share should not be read as deep loyalty.

  10. Support responsiveness and damage claims are real but second-order: verify with ticket and claims data before giving them roadmap slots ahead of tracking.

    Medium confidenceBase 521

    Q7: quicker support replies 23.0%, easier claims 13.6% (n=521); 6 open answers on unreachable/slow support (Q063, Q086, Q108, Q275, Q292, Q354) and 5 on damage/claims (Q050, Q455, Q486, Q491, Q514). Some support contact is itself caused by missing tracking (Q137, Q229, Q354), so part of this demand may fall away once visibility works.

Caveats

  • Response rate is 11.0% (521 completes from 4,720 invited accounts; 574 started). No non-response analysis has been done, so the direction of bias is unknown. The achieved sample is heavily skewed: 61.2% book fewer than 20 shipments a month against 40% in the database, and only 9.2% book 100+ against 18%. Every total should be weighted to the account mix before it is quoted.
  • Precision: the 95% margin on the 35.9% promoter share is about +/-4.1 points (n=521, simple-random-sample formula; a finite-population correction would narrow it to ~3.9, but non-response error dominates and is not captured). NPS carries roughly +/-6.5 points, and a difference between the two waves' NPS roughly +/-9.5 points. Band-level figures are much weaker: 100+ n=48 (about +/-14 points on a 50% share), Ireland n=14.
  • The Q1 vs Q3 comparison is not a clean trend. Mode changed from telephone interviews conducted by the customer's own account manager (strong courtesy/social-desirability pressure, likely inflating scores) to self-complete online. Q1's 640 customers were selected by account managers, not sampled from all active accounts, and 412 completed (64%). Q1 fielding (12 Jan-6 Feb) pre-dates the March fee change; Q3 fielding is 7-21 September. Volume band came from account records in Q1 and self-report in Q3. Any of these could produce several points of apparent change on its own.
  • Ireland: 14 respondents, 13 of them detractors. A single respondent shifts that NPS by ~14 points; -93 must not be presented as an estimate of Irish customer sentiment. Ireland status is self-reported (2.1% 'not sure') and the share of Ireland-shipping accounts in the population is unknown, so the ~127-account extrapolation is indicative only.
  • Open text was volunteered by 186 of 521 (35.7%), including only 45 of the 120 detractors. Theme counts describe those who chose to write, not prevalence among customers; they cannot be projected to the base. Q162 ('Costs more now. Full stop.') is counted as a fee/cost mention although it does not name the fee.
  • Q6 measures unpriced, hypothetical choice, asked immediately after Q5 which frames the change as a loss. 27.9% of those who say the fee is better for them also say they would likely choose a subscription - internally inconsistent, a classic hypothetical-bias signal. It is not a demand or revenue forecast. 7.5% never used the subscription and cannot compare at all.
  • S2 screens to the person who does the booking. The finance owner or director who actually feels the fee may be excluded, so fee sentiment could be understated (or overstated by bookers who dislike change); the pricing decision needs the budget-holder's view as well.
  • Q7 is a forced choice from a fixed list with a two-answer cap. 'More reliable hauliers' / 'on-time collections' is not an option, so tracking and alerts may partly be proxies for punctuality; the 65.6% net should not be read as demand for a specific feature spec.
  • All relationships here are cross-sectional associations, not causes: dissatisfied customers rate everything lower, so the visibility-detractor link is consistent with, but does not prove, that fixing visibility raises NPS.
  • The survey contains no churn, revenue, margin, shipment-value or usage data. The financial consequences of reversing the fee, of a subscription price point, or of an account-manager team cannot be estimated from it.
  • Weighting is possible only on shipment volume. Lane, tenure, sector and account value are unknown for the population, so residual bias remains even in weighted figures.

Next steps

  • decisionCorrect the board narrative now: report Q3 NPS as +12.9 unweighted / +17.7 weighted to the account mix, state that the Q1 comparison is confounded by mode and sample mix (like-for-like ~3 points, +/-9.5), and label trend as not yet established.
  • researchAudit operational data for all 4,720 accounts (no sampling error): tracking-event coverage and freshness by haulier, on-time collection and delivery rates, missed collections, 'where is my shipment' contact volumes, and the same metrics for the Ireland lane versus domestic.
  • experimentBuild and randomised-test live tracking plus proactive delay alerts on a treatment group of accounts (stratified by volume and including Ireland shippers), with a held-out control, after fixing the worst carrier event feeds.
  • decisionSet collection-slot and delivery SLAs with partner hauliers, including mandatory scan/event feeds and a missed-collection remedy; review the worst-performing carriers quarterly.
  • experimentDesign an optional subscription / monthly fee cap for regular low-volume bookers and test it as a real priced offer to a randomised subset of <20 and 20-99 accounts, with gross-margin guardrails; keep per-shipment as the default for everyone else.
  • researchRun 12-15 depth interviews with fee-complaining accounts including the budget-holder, plus recent lapsed accounts and Q3 non-respondents, and pair them with billing data on what each account actually paid before and after March.
  • researchCall back the entire population of Ireland-shipping accounts (identify them from shipment records, not self-report) and review the Ireland lane's carriers and transit times.
  • experimentScope, but do not roll out, a six-month named-account-manager pilot for a matched set of 100+ shipment accounts, launched only after the tracking/alerts pilot is live.
  • monitorRe-baseline the tracker: one consistent mode (online), a stratified random sample of all active accounts with over-samples of 100+ and Ireland shippers, reminders to lift response, weighting to the database mix, and a parallel finance-contact question set.
  • researchPull support-ticket and claims cycle-time data to size the responsiveness and damage-claims problems before allocating roadmap capacity.

Open-text themes it coded

Booking is quick and easy (the main positive driver) 79No usable tracking / stale or missing status updates, forcing them to chase hauliers 42Late or missed collections and deliveries (haulier reliability) 29Per-shipment fee costs more / prefers the old subscription (including budgeting certainty) 21Customer support hard to reach or slow to reply 6Damage in transit and a slow, confusing claims process 5Lukewarm or generic endorsement - 'does the job', nothing remarkable 17Too new or too little usage to judge 5Knock-on damage to their own customers and reputation (chasing, missed promises, lost order) 11

The survey it planned

3 screening questions and 10 questions, as the model wrote them.

  1. S1

    Thinking about your business's Quaymark account, which of these best describes your role?

    One answer
    • I am the main person who books our pallet and parcel collections with Quaymark
    • I share the booking equally with a colleague
    • I work for the business, but someone else does the booking
    • I do not work for this business

    Continues if I am the main person who books our pallet and parcel collections with Quaymark; I share the booking equally with a colleague

  2. S2

    Has your business booked at least one pallet or parcel collection through Quaymark in the last three months (June to August 2026)?

    One answer
    • Yes
    • No
    • Don't know

    Continues if Yes

  3. S3

    In a typical month, roughly how many shipments (pallet or parcel collections) does your business book through Quaymark? Your best estimate is fine.

    One answer
    • Fewer than 5
    • 5 to 19
    • 20 to 49
    • 50 to 99
    • 100 to 299
    • 300 or more
    • It varies too much to say

    Continues if Fewer than 5; 5 to 19; 20 to 49; 50 to 99; 100 to 299; 300 or more; It varies too much to say

  4. Q1

    How likely are you to recommend Quaymark to a colleague at another business that needs to send pallets or parcels? Please answer from 0 to 10, where 0 is not at all likely and 10 is extremely likely.

    Scale

    Scale 1-11: Not at all likely to Extremely likely

  5. Q2

    What is the main reason for the score you just gave? Please tell us in your own words.

    Open text
  6. Q3

    Which of these, if any, has your business experienced with Quaymark in the last three months? Please select all that apply.

    Any that apply
    • A collection or delivery was late
    • A collection did not turn up
    • We could not tell where a shipment was
    • We only found out about a delay after we had chased it
    • It took too long to get a problem sorted by customer support
    • There was no one contact who knew our account
    • Goods arrived damaged or went missing
    • Booking a shipment took longer than expected
    • What we were charged was higher than we expected
    • An invoice or charge was wrong
    • None of these
  7. Q4

    Overall, how satisfied or dissatisfied are you with the service Quaymark provides - booking, collection, delivery and support?

    Scale

    Scale 1-5: Very dissatisfied to Very satisfied

  8. Q5

    In a typical month, roughly how many hours in total do you and your colleagues spend chasing updates on Quaymark shipments or sorting out delays? Please enter a whole number, or 0 if none.

    Number
  9. Q6

    Quaymark can only improve a few things over the next six months. From the list below, please rank the three that would matter most to your business: 1 = most important, 2 = second most important, 3 = third most important.

    Ranking
    • Live tracking that shows where each shipment is
    • Automatic alerts about a delay, before you have to chase
    • A named account manager who knows your business and deals with problems
    • More reliable collection and delivery times
    • Faster answers from customer support
    • Lower or more predictable charges
  10. Q7

    Over the next six months, how likely or unlikely is your business to move some or all of its shipments to another booking platform or haulier?

    Scale

    Scale 1-5: Very unlikely to Very likely

  11. Q8

    In March 2026 Quaymark changed how it charges: instead of a monthly subscription (from £25 a month for the smallest accounts), there is now a fee of £3.50 per shipment, reduced for accounts booking 100 or more a month. Since that change, has the total amount your business pays Quaymark in a typical month gone up, gone down or stayed about the same?

    One answer
    • Gone up a lot (about £20 a month or more)
    • Gone up a little (less than £20 a month)
    • Stayed about the same
    • Gone down a little (less than £20 a month)
    • Gone down a lot (about £20 a month or more)
    • I was not aware of this change
    • Don't know
    Judges: assumptive

    Customers who joined after the March change can qualify for the survey but have no pre-change monthly payment to compare. There is no applicable response for these customers; being unaware of the change is not the same as not having been a customer before it.

    Asks how monthly spend changed 'since that change', which assumes the account was with Quaymark before March 2026. Accounts that joined after the change have no before-and-after comparison, and there is no 'we joined after the change' option. These customers would be pushed into 'Don't know' or a guess.

  12. Q9

    Which way of charging would you prefer Quaymark to use for your account?

    One answer
    • A monthly subscription (from £25 a month for the smallest accounts), with no fee per shipment
    • A fee of £3.50 per shipment, reduced for accounts booking 100 or more a month, with no monthly subscription
    • No preference - either would suit us
    • Don't know
  13. Q10

    Thinking about what you pay and what you get, how would you rate Quaymark's overall value for money today?

    Scale

    Scale 1-5: Very poor to Very good

Sample plan

Population and invitation: all 4,720 active UK business accounts in the customer database (a census invitation, not a sample), emailed to the main booking contact, one response per account enforced by a unique tokenised link tied to account ID (if a shared booker responds first, that response stands). No geography screen is needed because Quaymark serves only UK businesses; S1 and S2 confirm role and the 'booked in the last three months' definition, and terminate non-bookers politely. Banding for reporting and quotas comes from the database's average monthly shipment volume (<20, 20-99, 100+); S3 is a cross-check and the fallback where the database field is missing or stale (discrepancies of more than one band are flagged and the database value used, with a sensitivity check using S3). Fieldwork: 12 days in September 2026, invitation plus three reminders to non-openers/non-completers, sent from 'Quaymark Research' rather than account managers to reduce courtesy bias; invitation wording is neutral ('Help us improve Quaymark - about 6 minutes') with no mention of pricing, to avoid differential non-response by customers who are angry about the fee. A £5 charity donation per complete (or equivalent prize-draw entry) is offered to lift response. PRECISION CONFLICT - flagged for the board: ±5pp at 95% for a promoter share within each of three bands needs about 385 completes per band at the worst case (p=50%), i.e. roughly 1,155 completes, before any inflation for weighting. The stated 520 completes cannot deliver this: split three ways (~173 per band) the margin is about ±7.4pp, and worse if one band dominates responses. Two options: (a) keep 520 and report band-level promoter shares as indicative with their actual intervals (recommended only if the board accepts ±7-8pp); (b) raise the target, which is feasible because the finite population is small - applying the finite population correction to actual band counts (to be pulled before fieldwork) cuts the requirement substantially, e.g. with bands of 3,600 / 850 / 270 accounts the needs are ~348 / ~266 / ~159, about 773 completes, or ~890 allowing a weighting design effect of ~1.15. We recommend (b): target ~1,000-1,150 completes (a 21-24% response rate, achievable on a customer list with three reminders and an incentive), with hard minima per band of the FPC- and design-effect-adjusted requirement and a census push (extra reminders, plus a reminder phone call that asks the contact to complete the same online form - never administering it by phone) in the 20-99 and 100+ bands, which are smaller and will not fill naturally. Soft quotas: monitor completes daily by band; do not close the <20 band early, but stop topping it up once its minimum plus 15% is reached so effort goes to the larger-volume bands. Optional add-on for comparability: a dual-mode calibration sub-study of ~150 randomly chosen accounts re-contacted by phone with the Q1 wording, to estimate the mode/administrator effect (see O1). Data quality: drop speeders (<2 minutes), straight-lined scale answers with gibberish open text, duplicates and non-qualifying roles; report the number removed. Non-response and weighting: compare respondents with the full 4,720 on database variables (volume band, tenure, sector, region, and whether they raised a support complaint since March) and post-stratify overall figures to the population on band and complaint incidence, since complaint-driven customers are likely to over-respond; band-level results are reported unweighted with Wilson 95% intervals, overall results weighted. Report achieved margins per band explicitly against the ±5pp target.