The bottleneck is your best person

01 The hour in which the answer is worth something

A distributor’s customer asks a question that sounds simple. Is this component compatible with that model. Is there an equivalent in stock. What is the delivery time for the non-standard variant. The question arrives by email at 9:40, and it is not a request for information: it is the visible part of a purchase that is already in motion, probably at more than one supplier.

How much time there is to answer has been measured. In a Harvard Business Review study, James Oldroyd, Kristina McElheran and David Elkington audited 2,241 American companies by sending each a test enquiry and timing the reply. 37% answered within the hour, 16% between one and 24 hours, 24% took longer than a day and 23% never answered at all. Among those who did reply within thirty days, the average first response took 42 hours.

The companion study, run on 1.25 million leads across 29 B2C and 13 B2B companies, measured what that delay costs. Contacting the customer within the first hour made a real conversation with a decision maker nearly seven times more likely than contacting them one hour later, and more than sixty times more likely than waiting a day. The research is from 2011, and it shows: it describes a world of web forms. What it measures, though, is not a level but a relationship, between how fast you answer and whether you get to talk to the person who decides. That relationship has not become more forgiving since.

It matters more now because the contact window has narrowed at the other end too. Gartner’s research, summarised by Brent Adamson in HBR, found that B2B customers dedicate only 17% of their purchase process to talking to potential suppliers. Everything else happens without you in the room. The few exchanges you do get are therefore worth disproportionately more, and each one that goes cold is not one lost email: it is a fraction of the only access you had.

02 The queue nobody measures

Here is the part the sales literature tends to skip. In a technical distributor, the delay is rarely a motivation problem. Nobody is sitting on the enquiry. The salesperson read it at 9:41 and cannot answer it, because answering requires knowing something specific about a catalogue with thousands of references.

So what actually happens is a queue, and it looks like this:

1. 09:41 The salesperson reads the question and knows roughly where the answer might be. Not well enough to send it to a customer.
2. 09:45 They look for it themselves: the technical datasheet, the shared folder, a similar past order. Twenty minutes, no certainty.
3. 10:05 They ask the colleague who knows. That colleague is on the phone with another customer.
4. 11:30 The colleague answers, from memory, correctly. Or asks for the model number again, because the question arrived stripped of half its context.
5. 12:10 The answer goes out to the customer. Two and a half hours after it arrived, on a good day, and only because nobody was on holiday.

Two and a half hours is a fast day, and it has already spent the entire window in which the answer was worth the most. Nothing in that sequence was anybody’s fault, and no CRM would have shortened it, because the CRM knows the enquiry exists. It does not know whether that component is compatible with that model.

03 What it costs the person everyone asks

Every company of this kind has one. Sometimes two. The technical salesperson with fifteen years on the catalogue, who knows the equivalences, remembers why that customer refuses that brand, and can tell in three seconds that the code in the email is a discontinued version.

That person is the company’s most valuable asset and, at the same time, the reason the queue exists. And the cost is paid twice. Once by the colleagues waiting. Once by them, because a person who answers thirty internal questions a day is a person who is not selling, not visiting customers, not training the new hire: the three things they would be best in the building at doing.

The background noise is already documented. Our study on the cost of knowledge that can’t be found used the IDC and McKinsey Global Institute measure: white-collar workers spend 19% of their working time looking for information their company already owns. Gartner’s 2023 survey of digital workers found that 47% struggle to find the information their job requires, spread across an average of eleven applications, up from six in 2019. In a distributor, that ordinary friction has a specific shape: it converges on the desk of whoever is best.

There is also a quieter risk. A company whose technical knowledge runs through one person has a single point of failure that does not appear on any risk register. It fails on the day that person is at a trade fair, or ill, or has finally accepted the offer from the competitor who has been calling for two years.

04 And the new hire takes a year

The obvious answer is to hire another one. In Italy, the arithmetic makes that harder every year.

Excelsior, the Unioncamere and Ministry of Labour forecasting system, projects that “commerce and tourism” will be the supply chain with the largest hiring need in the country between 2025 and 2029: between 574,000 and 702,000 people, about 18% of the national total. It is also the chain with the largest replacement need: 544,000 workers to be replaced, again 18% of the national figure. Sales staff are the single largest professional group by requirement, at 237,000 to 272,000 people.

And the supply is not there. The same report expects a shortfall of 14,000 to 17,000 young people every year in the administrative, secretarial and sales-services educational track alone. So the market will not hand a distributor a replacement for the person who knows the catalogue. It will hand them, at best, someone capable of learning it.

Which returns the problem to where it started. If learning the catalogue means a year of standing next to the person who already knows it, then hiring does not remove the bottleneck: for the first year it makes it worse, because now the expert answers the customers’ questions and the new colleague’s questions.

05 It is not a CRM problem, and not a catalogue problem

The instinct is to buy something. Usually a better CRM, sometimes a new product information system. Both are useful and neither addresses this, because the failure is not in storing the information. It is in retrieving it under time pressure, in a form someone can send to a customer.

Consider what the company already has written down, in almost every case:

• Manufacturer datasheets, thousands of PDFs, often scanned, in a folder organised by supplier rather than by question.
• Compatibility and equivalence notes, frequently in an email thread or a spreadsheet maintained by one person.
• Past quotations and orders, which contain the single most useful thing in the business: what was actually proposed to a customer with the same problem, and whether it worked.
• Correspondence with manufacturers, where the exceptions live: the substitution that is allowed, the tolerance that is not.

All of it is written. None of it is answerable. The gap between those two words is the whole problem, and it is why the queue forms in front of the one person who has effectively memorised the index.

06 What actually shortens the queue

The goal is not to replace the expert. It is to make the first hop unnecessary: the salesperson gets a usable answer without escalating, and escalates only the genuinely hard cases, which is what the expert should be spending the day on anyway. Four requirements, and the order matters:

1. Answers must cite the document and the page. This is first, not last. A salesperson will not forward a compatibility claim to a customer on the strength of an answer they cannot verify, and they are right not to. Without the source, the tool gets used once and abandoned.
2. It has to read what you actually have. Scanned datasheets, price lists, past quotations, email threads: the material is messy by nature, and anything that requires tidying it up first will never start.
3. It has to say when it does not know. In technical distribution a confident wrong answer is worse than no answer, because it reaches the customer and then comes back as a return.
4. What gets corrected has to stay corrected. When the expert fixes an answer, the correction must become part of the shared memory for everyone, immediately. That is the mechanism by which fifteen years of catalogue knowledge stops being biography and becomes company property.

The measurable effect is not “the team is more productive”. It is narrower and easier to check: how many questions reach the expert this month compared to last, and how long the customer waits for the first useful answer. Both numbers are visible from the first week, and both are worth more than they look, because they operate on the 17% of the buying process where the supplier is still in the room.

07 Methodological note

All figures come from primary sources, read in the original. The HBR article and the Excelsior report were read directly from their PDFs.

• Response times: Oldroyd, McElheran and Elkington, HBR, March 2011. The audit covers 2,241 US companies; the seven-times figure comes from a separate study by the same authors on 1.25 million leads across 29 B2C and 13 B2B companies. A qualified lead is defined by the authors as “a meaningful conversation with a key decision maker”. The study is fifteen years old and is presented as such.
• 17% of the buying process: Brent Adamson, Sensemaking for Sales, HBR, January-February 2022, reporting Gartner research.
• Italian labour figures: Excelsior, medium-term forecast 2025-2029. “Commerce and tourism” is a supply chain that includes retail and hospitality, so it is wider than technical wholesale distribution: it is used here to size the pressure on commercial roles, not as a measure of the wholesale sector.
• The timeline in chapter two is an illustration, not a measurement. There is no published data on internal escalation times in Italian distributors, and we have not invented any: the individual figures in this article are all sourced, the sequence is not.

One figure we deliberately left out. The widely quoted claim that 78% of buyers purchase from the supplier who responds first traces back to a survey with no published report and no methodology: every citation leads to another blog. It is a good line and we have no way to verify it, so it does not appear here.

Sources

1. James B. Oldroyd, Kristina McElheran, David Elkington, The Short Life of Online Sales Leads, Harvard Business Review, March 2011.
2. Brent Adamson, Sensemaking for Sales, Harvard Business Review, January-February 2022 (Gartner research).
3. Unioncamere, Ministry of Labour and Social Policies, Excelsior Information System, Previsioni dei fabbisogni occupazionali e professionali in Italia a medio termine (2025-2029), 29 August 2025.
4. Gartner, Survey on Digital Workers, May 2023 (4,861 full-time workers, organizations with over 100 employees).
5. IDC, McKinsey Global Institute, The Social Economy, 2012.
6. ekory, The cost of knowledge that can’t be found, July 2026.