The quiet leak at the trade counter

Executive summary

Owner-run fastener distributors are losing margin in two places their reports never show

The minutes spent rekeying orders, and the regulars who stop ordering without saying why. The fastener itself is a small part of what a trade customer pays for. What they really buy is the right part, first time, without chasing. That makes the order-to-reorder loop, from the emailed part list to the next top-up, the place where a small distributor wins or quietly loses.

1. Costs have risen by close to half, and margins are measured in pennies per hundred. UK producer prices for basic metals and fabricated metal products rose 47% between 2019 and September 2026¹. The National Living Wage rose 46% between April 2020 and April 2026², and employer National Insurance rose to 15% in April 2025³.
2. Customers use ten channels, but the order still arrives as an email someone has to retype. B2B buyers now use an average of ten channels⁴, yet emailing a rep was the most preferred way to place an order (39%) in McKinsey’s survey of 3,600 decision makers⁵.
3. Rekeying is expensive, and fastener orders are built for errors. APQC benchmarks put the median cost of a traditional-channel order at $10.00 against $3.10 for digital channels⁶. A controlled study found 0.95% of values keyed by hand were wrong⁷; on a 12-line part list that compounds quickly.
4. The largest leak is silent: regulars who drift away without complaint. Bain’s research, cited by HBR, links a 5% rise in retention to a 25% to 95% rise in profit⁸. More than half of B2B buyers say they are likely to switch after a poor experience across channels⁴.
5. The usual fixes do not reach the gap, and the cost is material. Gartner expects more than 70% of recent ERP initiatives to fall short of their business case⁹; 71% of UK firms trying to hire report difficulty¹⁰. In our illustrative model a 12-person distributor loses around £44,067 a year, mostly in margin on regulars who go quiet.

• Regular

  A trade account that orders on a recognisable pattern, weekly, monthly or on a kanban cycle.

• Drift

  A regular ordering less often or less widely without telling you, usually before leaving altogether.

• Reorder rhythm

  The normal interval between a regular’s orders, used to judge when an order is overdue.

• RFQ

  Request for quotation: a customer’s list of parts they want priced, often by email.

• Kanban and VMI

  Bin-based replenishment and vendor-managed inventory, where you keep a customer’s stock topped up.

• First-time accuracy

  The share of orders keyed, picked and delivered correctly without any later correction.

01 · The market

A fragmented, consolidating market where turnover is rising faster than the number of firms

There is no official count of fastener distributors. The nearest proxy is SIC 46.74, wholesale of hardware, plumbing and heating equipment and supplies, where most fixings and fasteners wholesalers are classified. In March 2024 it contained 3,495 VAT or PAYE registered enterprises employing 52,511 people, an average of about 15 per business¹¹. A further 7,070 enterprises sit in SIC 46.69, wholesale of other machinery and equipment, which includes many industrial and engineering supply houses¹¹.

• +35%Nominal turnover growth in SIC 46.74 between 2022 and 2024, while the number of enterprises fell by about 2%¹¹. Part of this is price inflation, but it also points to sales concentrating in fewer, larger hands.
• 155UK enterprises classified as manufacturers of fasteners and screw machine products (SIC 25.94) in 2024¹¹. IBISWorld puts that industry’s revenue at about £790m in 2026¹². Domestic production is small; most of what distributors sell is bought in.

Consolidators are investing in exactly the work small firms still do by hand

The scale players are buying up independents and investing in systems. In 2022, United Tooling Solutions acquired three UK fastener suppliers in one move, taking its annual revenue to £85m and its network to 120 branches, with finance for new vending units to distribute to customers¹³. A buyer who is used to that level of service from a national group judges the independent by the same yardstick: how fast the quote comes back, and whether the order arrives right.

“Fed up with regular customers going quiet before anyone notices, because the reps are stuck in spreadsheets working out who to ring.”

02 · Costs and the 15-85 rule

Metal and labour costs have both risen by close to half, while the part itself is a small share of what the customer pays for

Two cost lines have moved together. Producer prices for metal products jumped 16% in 2022 alone and have kept climbing, reaching 171.1 in September 2026 against 116.3 in 2019¹. On the labour side, the wage floor has risen every April, and since April 2025 employers pay 15% National Insurance on earnings above £5,000 a year, down from a £9,100 threshold³. The British Chambers of Commerce reports that 71% of firms trying to hire in Q3 2026 faced difficulties¹⁰.

Customers are buying the 85%, not the 15%

Bossard, one of Europe’s largest fastener specialists, describes the economics with its “rule of 15-85”: on average the fastener itself makes up only around 15% of total cost, with the other 85% coming from procurement, inventory, logistics, assembly and related processes¹⁴. The figure is Bossard’s own estimate, but the logic holds for any trade customer: a bolt that arrives wrong or late costs far more than its price.

For a small distributor this cuts both ways. It is the strongest argument for paying a little more to a reliable local supplier. It is also why admin errors and slow quotes hurt so much: they land squarely in the 85% the customer cares about.

• 15 / 85Share of total fastening cost in the part itself versus procurement, stock, logistics and handling, on Bossard’s estimate¹⁴.

03 · Imports and trade remedies

Most bolts on UK shelves are imported, so a small distributor manages more suppliers, longer lead times and trade rules set elsewhere

The UK imported $613m of bolts and screws in 2024, weighing 122.7 million kilograms¹⁵. Asian origins (China, Taiwan, India, Japan, Thailand, Korea, Vietnam and Malaysia) supplied 62% of that weight. The value and weight shares diverge sharply because unit values do: Chinese product averaged about $2.34 a kilogram against $9.53 for German product¹⁵. Most distributors stock both ends, which means more suppliers, more purchase orders and more lines to price.

Trade policy adds a further layer. In 2022 the EU imposed definitive anti-dumping duties of 22.1% to 86.5% on many Chinese steel fasteners¹⁶. The earlier EU measure that covered the UK was repealed in 2016¹⁷, and we found no equivalent measure for these products in force in Great Britain in published notices. China, for its part, extended its own anti-dumping measures on fasteners from the EU and UK for five years in 2022¹⁸. Rules set in Brussels or Beijing can shift supply routes and prices that a 12-person firm in the Midlands has to absorb.

04 · How orders actually arrive

Buyers have moved to ten channels, but the order still lands as an email that someone has to read, interpret and retype

McKinsey’s 2024 B2B Pulse found buyers use an average of ten channels across their journey, up from five in 2016, and that buying preferences split into thirds: one third in person, one third remote, one third digital self-serve, across industries and company sizes⁴. Among suppliers that offer e-commerce, it now brings in more than a third of revenue⁴, and McKinsey reports online ordering as high as 40% of orders in Western Europe¹⁹.

Two other findings matter more for a small distributor. Gartner’s 2026 survey found 67% of B2B buyers prefer a rep-free experience²⁰. And when McKinsey asked decision makers how they most liked to order, emailing a sales rep came top at 39%⁵. For an owner-run distributor without a trading portal, those two facts meet in the shared inbox.

What “an order” looks like at a fastener counter

• PDF purchase orders

  Generated by the customer’s own system, in their layout, with their part codes. Every line has to be matched to yours.

• Part lists and RFQs

  Spreadsheets or email bodies listing size, length, grade, finish and quantity, often “same as last time but in M6 and M10”.

• Kanban and bin top-ups

  Scanned bin labels, counts phoned in after a site visit, or a photo of empty bins on a production line.

• Phone and counter

  Still the channel of choice for urgent work. The order is captured on paper or from memory and keyed in later.

“Someone losing half the morning retyping emailed orders and price requests.”

05 · Pain one: rekeying orders and RFQs

An order that arrives by email or PDF costs around three times as much to process as a digital one

APQC’s benchmarks are drawn from larger organisations, so the dollar figures will not map neatly onto a 12-person firm. The ratio is what matters: orders through traditional channels (email, fax, phone) cost roughly three times as much to process as digital ones, whether or not the organisation is disciplined about accuracy⁶. APQC also notes that fax and email orders traditionally carry higher error rates and slower fulfilment⁶.

Fasteners make the traditional channel especially slow. A typical part list describes each line by thread, length, head type, drive, grade, material and finish. Each attribute has to be read, matched to a stock code and checked against the customer’s price. The time is driven by lines, not orders.

Illustrative model Assumptions, not data from any business.

• 613 hrsa year spent retyping what customers have already typed, in the mid scenario. That is 37% of a full-time person, before any errors are fixed.

Small keying error rates compound into a meaningful share of orders that need fixing

In a controlled study of manual data entry, Barchard and Pace found that 0.95% of values entered once by hand were wrong. Visual checking reduced that only to 0.82%, which was not significantly better. Double entry, where two people key the same data and differences are flagged, cut errors to 0.027%⁷.

The participants were students, not order processors, so the absolute rate should not be read as a distributor benchmark. What carries over is the arithmetic. Errors are per value, and a fastener order contains a lot of values: a 12-line order with size, length, grade and finish per line is 48 separate entries before quantity and price.

The strings are also unusually similar. M8x30 and M8x35, A2 and A4, 8.8 and 10.9, bright zinc and hot-dip galvanised: each pair differs by one or two characters and each mistake produces a part that looks right in the box.

1. M8x35 keyed for M8x30
2. Wrong part picked and packed
3. Delivered, found on site
4. Collection, re-delivery, credit note
5. Line stopped, trust dented

• 1.6% → 0.4%Order entry error rate at MEDRAD before and after it focused on first-time data accuracy, one of the cases APQC reports⁶.
• 98 vs 86Median perfect-order score for organisations that measure and reward first-time accuracy versus those that do not⁶. What gets measured gets fixed.

06 · Pain two: the regulars who go quiet

The most expensive event in a distribution business generates no complaint, no cancellation and no credit note

It is a regular customer who simply orders less.

• 25 to 95%Rise in profits from a 5% increase in customer retention, in Bain research by Frederick Reichheld⁸.
• 5 to 25×How much more it costs to acquire a new customer than to keep an existing one, depending on study and industry⁸.
• 96%of customers who had a high-effort service interaction became more disloyal, against 9% after a low-effort one²¹.

These figures come from cross-industry research, not fastener distribution, and the Bain range is wide because it depends on the business. But the direction is consistent: keeping an account is far cheaper than replacing it, and effort, such as chasing a quote or sending back a wrong part, is what drives customers away. McKinsey found more than half of B2B buyers are likely to switch suppliers if they do not get a smooth experience across channels⁴.

Why fastener accounts drift without anyone deciding to leave

A fastener customer rarely sacks a supplier. A new buyer joins and brings their own contacts. A national group offers to take over the bins. A competitor answers an RFQ an hour faster, and then again the following month. Each event is small. The pattern shows up only as a gap in the order history, and only if someone is looking at each account against its own rhythm.

1. Week 0 · order
2. Week 3 · order
3. Week 6 · order
4. Week 9 · missed
5. Week 12 · missed
6. Week 15 · missed
7. Week 18 · flagged

Illustrative model How a three-weekly account goes quiet. Each missed order looks like nothing on its own; the gap is only seen if accounts are reviewed against their own rhythm.

Illustrative.

Standard reports average the warning away

Accounting and ERP reports are built around periods: sales by month, by customer, by product. They answer “what did we sell?” They do not answer “who should have ordered by now and has not?”, which needs each account compared with its own normal rhythm. With prices rising, a total can grow even while the number of active regulars falls: metal product prices rose 47% between 2019 and September 2026¹, which flatters any sales line measured in pounds.

• No owner

  The office owns orders, the reps own relationships, the owner owns the numbers. The overdue-reorder list falls between all three.

• No time

  Salesforce found sales reps spend only 30% of their week selling²². Building a call list from exports is exactly the non-selling work that crowds it out.

• No trigger

  A missed order does not create a document, an alert or a task. Nothing lands in anyone’s inbox until the account has already gone.

“Fed up with regular customers going quiet before anyone notices.”

07 · Supporting pains: goods in, money in

Two quieter drains sit either side of every sale

Goods in: more suppliers, more open purchase orders

Because most product is imported and sourced at different price points, a fastener distributor typically runs open orders with several importers and mills at once. Each purchase order needs confirming, a delivery date needs checking, and a late consignment needs passing on to the customer waiting for it. None of this is captured in a sales report, and it usually lives in one person’s inbox.

Money in: late payment is the norm, not the exception

• 63%of small businesses spend time chasing overdue payments, at an estimated cost of up to £5,200 a year²⁴.
• 80%of small businesses still rely on manual bank transfers to get paid; 24% use Direct Debit²⁴.
• £11bna year: the cost of late payment to the UK economy, which the government links to 38 business closures every day²⁵.

For a distributor on 30-day terms, an overdue trade account is a double risk: cash is tied up, and an account that has stopped paying is often an account that is about to stop ordering. The same lack of a trigger that hides a quiet regular also hides a slow payer until month end.

08 · Why the usual fixes fall short

A new ERP, another hire, a web shop or working harder each solve something real, but none closes the gap

• Why a new ERP does not fix this

  An ERP records what happened; it does not read a PDF in the inbox or notice an account that has gone quiet. Gartner expects more than 70% of recently implemented ERP initiatives to fall short of their original business case, with as many as 25% failing catastrophically⁹. Panorama’s 2024 survey found 32.8% of projects ran over budget and 31.3% finished late²⁶.

• Why hiring does not fix this

  Another hire adds capacity to retype, not a way to stop. It is also harder and dearer: 71% of firms trying to recruit report difficulties¹⁰, the wage floor is £12.71 an hour², and employer National Insurance is now 15% from the first £5,000³. And a new person takes months to learn which M10 a customer means.

• Why a web shop or portal does not fix this

  Digital ordering works where customers want it, and suppliers that offer e-commerce earn over a third of revenue through it⁴. But preferences split into thirds⁴, and email remained the most preferred way to order through a rep⁵. A long-standing account that sends a PDF from its own purchasing system will keep sending it.

• Why working harder does not fix this

  Checking your own keying barely helps: visual checking produced error rates that were not significantly better than no checking at all⁷. And every hour a rep spends building a call list is an hour out of the 30% of the week that reps actually spend selling²².

09 · Diagnostic framework and questions

Most distributors sit at stage one or two on at least one dimension, and the weakest dimension sets the size of the leak

The order-to-reorder maturity grid describes how a distributor handles the full loop from a customer’s order to their next one. Read across each row and mark the cell that best describes how things work on a normal Tuesday, not on a good day. Stages are cumulative: stage three assumes stage two is in place.

• How to read your result

  Two or more marks in stage one usually means the costs modelled in section 10 apply in full. A business mostly at stage three still tends to leak through whichever single row is weakest, because work piles up where the process is manual.

• What moves a business up a stage

  Moving up is rarely about buying a bigger system. It is usually about making one thing explicit: who owns the overdue-reorder list, what counts as a complete order, how often open purchase orders are checked, and when an account is escalated.

Ten questions that reveal, in under ten minutes, where the loop is leaking

Answer honestly for the last four weeks. Score 0 for the first option, 1 for the second and 2 for the third.

• 0 to 7Most of the loop runs on memory and goodwill. The worksheet in section 11 will likely show a large number.
• 8 to 14The basics are there but one or two rows of the maturity grid carry most of the cost. Find them first.
• 15 to 20A disciplined operation. The remaining gains are in speed and in catching drift earlier.

10 · Illustrative cost model and sensitivity

For a typical 12-person distributor, the cost of doing nothing runs from about £16,000 to over £110,000 a year, with a mid case near £44,000

Illustrative model The model below is not based on any real business. It combines published figures with assumptions for a business with around 12 staff: an owner, two or three in the office, two or three reps, warehouse and drivers. Three scenarios show how sensitive the answer is to the inputs.

Working year: 230 days or 46 weeks. One full-time equivalent is taken as 1,650 hours. Error and attrition rates are assumptions chosen to be conservative; they are not published benchmarks for this sector.

What the model leaves out: quotes lost because they went out late, customer line stoppages caused by wrong parts, supplier chasing and late payment, and the owner’s own time.

• 41%of the mid-case total is margin lost on regulars who drift away. Retyping is the visible cost; attrition is the expensive one.

11 · Worksheet

The cost of doing nothing, in your own numbers

Replace each illustrative number with your own. The illustrative column repeats the mid scenario from section 10. Illustrative only, not data from any real business.

Harder to count, but worth writing down

• Wrong-part returns from mis-keyed lines (M8x30 for M8x35, A2 for A4), per month.
• RFQs that went back the next day or later, per week.
• Kanban or VMI sites where top-ups have slowed and nobody has asked why.
• Hours a week your reps spend in spreadsheets working out who to ring.
• Open supplier POs nobody has checked this week.

12 · What leading distributors do differently

The distributors that hold their regulars treat the order-to-reorder loop as a process with owners and measures

Practices drawn from the benchmarks and research cited in this paper, in the language of a fastener business. None depends on a particular product.

• 1. They measure first-time accuracy

  APQC found organisations that measure, track and reward first-time data accuracy are 2.8 times more likely to rate their order management as extremely effective at improving fill rates and on-time delivery⁶. Count wrong-part returns every month and review the causes.

• 2. They run one intake queue with a clock on it

  Every emailed order, part list and RFQ is logged on arrival and has a target time to confirm. The queue is visible, so nothing waits in one person’s inbox.

• 3. They hold each customer’s codes and rules once

  Customer part codes, agreed prices, pack sizes and delivery rules are recorded against the account, not in someone’s head, so a new starter can key a regular’s order correctly on day one.

• 4. They watch every account against its own rhythm

  Rather than reading monthly totals, they keep a list of regulars who are overdue against their normal reorder interval, and a named person works it every week. Kanban sites are watched on usage, not visit dates.

• 5. They protect rep selling time

  With reps spending only 30% of their week selling²², the call list is prepared for them, not by them.

• 6. They sell the 85%, and run it like they mean it

  If the part is only around 15% of the customer’s total cost¹⁴, the rest is speed, accuracy and certainty of supply. Open purchase orders are checked weekly and customers are told about delays before they ask.

The test of all six is the same: on any given morning, could someone who is not the owner say which orders are waiting, which regulars are overdue, and which supplier deliveries are late?