Digitising wholesale: what the studies really say about time saved

At a trade-show stand, a boutique buyer confirms her selection, signs the order form and hurries to her next appointment. For her, the order is placed. For the brand, the work is just starting: that form has to be re-keyed, checked, merged with the other reps’ orders, then turned into quantities for the factory.

“How much time will we save?” The question comes before any decision to digitise order-taking, and it too often gets figures with no traceable origin. Academic research, consulting surveys, public evaluations: no study looks specifically at fashion wholesale, but each sheds light on part of the problem. This article goes through what they measure, what they leave out, and how to put a number on your own gain without relying on a generic figure. Every study is dated, so you can judge how fresh it is.

Where does the time go when orders are taken on paper or in Excel?

The hours are lost after the appointment, along the path that takes the order from the stand or showroom to the factory. The comparison of paper order forms, Excel or a B2B platform describes each of these steps. Seen in terms of time spent, they boil down to six items:

• Double entry: the paper form or the rep’s file is re-keyed into the central spreadsheet, then often a third time into the ERP or invoicing software.
• Checks: reviewing references, colours, sizes, country prices and currency, and verifying the minimum order and the discount granted.
• Corrections: tracking down the source of a wrong total, a swapped size or a missing colour, then calling the customer back to confirm.
• Consolidation: merging files from several reps and reconciling versions at the end of the show.
• Follow-ups: unsigned order forms, payment terms to confirm, deposits to chase, buyer questions about their order.
• Factory re-keying: rebuilding quantities by style, size and colour for each manufacturer.

None of these items appears on the trade-show schedule. They spread over the following days and weeks, between the sales rep, sales administration and often the founder. Scattered like this, they almost always escape any costing.

Spreadsheet errors: 5.2% of formulas affected in audits

Raymond Panko, a researcher at the University of Hawaii, devoted much of his work to spreadsheet errors. In his review “Spreadsheet Errors: What We Know. What We Think We Can Do”, presented in 2000 at the first EuSpRIG (European Spreadsheet Risks Interest Group) symposium and posted on arXiv in 2008, he compiles audits of real-world files carried out in the 1990s. The studies that measure errors formula by formula find them in 5.2% of the formulas examined.

The researcher places this result within the wider research on human error. Simple but non-trivial cognitive actions, such as typing, calculating or writing an instruction, show a base error rate generally between 1% and 5%. A spreadsheet adds up human actions, and each of them carries that rate. Software has changed since then; the hand doing the typing much less so.

What this rate means for a wholesale order

A wholesale order is not a single line. It is a grid of quantities by style, colour and size, with prices by country and discounts. Take a hypothetical order of 8 styles, each in 3 colours and 6 sizes: 144 cells filled in by hand. At a 1% error rate, you would expect more than one error; at 5%, about 7.

Across the hundreds of cells in a selling campaign, errors are therefore almost certain. The real question is when they will be found: at the check, or at delivery, when the boutique unpacks a size it never ordered. Size distribution follows its own rules, covered in the guide to size curves and prepacks, those pre-assorted packs sold as a single carton.

Sales automation: a third of tasks, one company in four

In its article “Sales automation: The key to boosting revenue and reducing costs”, published in May 2020, McKinsey draws on research by the McKinsey Global Institute. The firm estimates that about a third of sales and sales-operations tasks can easily be automated with the technology of the time.

The same article measures the gap between that potential and practice. Only one company in four had then automated at least one sales process. Early adopters reported efficiency gains of 10 to 15%, with potential sales uplift of up to 10%. These figures predate the rise of generative AI: the potential has probably grown since, and so has adoption, but no comparable measurement has been published.

In a wholesale brand, the tasks in question are easy to spot: order entry, checking terms, sending order summaries, follow-ups, preparing files for the factory. They tie up sales reps and sales administration during the selling weeks, the very period when every hour spent on a file is an hour missing with customers.

B2B buyers: McKinsey’s “rule of thirds”

McKinsey’s B2B Pulse Survey polls B2B decision makers every year. Its 2024 synthesis, “Five fundamental truths: How B2B winners keep growing”, is based on nearly 4,000 decision makers in 13 countries. It describes a “rule of thirds”: buyers’ interactions are split roughly equally between in-person contact, remote exchanges with a sales rep and digital self-service.

This split marks a change in habits. Before 2020, about half of respondents still preferred in-person contact to research and evaluate suppliers. Meeting face to face still carries weight when a relationship begins, though: 41% of respondents favour in-person contact with a new supplier, and 40% for a first purchase.

For a brand, the division of roles follows. Trade shows and showrooms remain the places where accounts are opened. In-season reorders, changes and follow-up increasingly go through digital channels. A portal where buyers place and track their own orders keeps the sales rep in the loop and frees them from low-value tasks. The article on digital wholesale trends analyses this shift.

Invoices and purchase orders: the cost of manual processing

Hasselt University: over €7,000 in potential savings per year

On the invoicing side, a Belgian academic study provides precise figures. Published in 2016 in the International Journal of Digital Accounting Research by Poel, Marneffe and Vanlaer (Hasselt University), “Assessing the electronic invoicing potential for private sector firms in Belgium” is based on a survey of 683 companies.

The authors put the potential savings from digital invoicing at over €7,000 per year for an average small company in the sample. These small companies say they are willing to invest €2,380 on average, nearly three times less than the estimated gain. At national level, using 2014 data, the total cost of invoicing for Belgian private companies is estimated at €3.47 billion (0.96% of GDP). It could fall to €1.46 billion (0.38% of GDP) if all invoices were digital, a difference of about €2 billion.

APQC: from $14 to over $54 to process a purchase order

On the ordering side, APQC, the US benchmarking reference body, published its benchmarks on the cost of processing a purchase order in April 2026. The most efficient quarter of organisations spends about $14 per order; the least efficient quarter, more than $54. This gap of nearly one to four comes mainly from the way the work is organised.

The figure concerns the buyer side, meaning your boutique and department-store customers. A supplier that simplifies how they place orders saves them time, and that saving becomes a sales argument to put forward in appointments.

Digitising wholesale: how do you calculate your own gain?

Studies provide orders of magnitude. They say nothing about your collection, your team or your customers. The method takes one season:

1. Time about ten recent orders, step by step: entry, checks, consolidation, sending the summary, follow-ups, preparing the factory file.
2. Count a season’s errors: lines corrected after sending, credit notes, exchanges or disputes caused by an entry error, and the time spent handling them.
3. Multiply the average time per order by your annual order volume.
4. Value those hours at the fully loaded hourly cost of the people involved: sales reps, sales administration, founder.
5. After switching tools, repeat exactly the same measurement over a full season and compare.

The protocol matters as much as the result. If the “after” measurement covers different steps or different people from the “before” one, the comparison no longer proves anything.

Illustrative example (assumptions, not a study)

Take a fictitious brand that handles 600 orders a year over two seasons. The assumptions used for the example are as follows: 40 minutes of post-appointment work per order on paper and Excel (re-keying, checks, follow-ups), 15 minutes once the order is entered directly into a tool, 5% of orders with an error needing one hour to fix, and a fully loaded hourly cost of €40.

• Time saved per order: 40 − 15 = 25 minutes; over the year: 25 × 600 = 15,000 minutes, or 250 hours.
• Errors: 5% of 600 = 30 orders × 1 hour = 30 hours, assumed to be avoided for simplicity.
• Total: 250 + 30 = 280 hours, or 8 weeks of 35 hours; valued at €40, 280 × 40 = €11,200 per year.

Spread over the brand’s two seasons, the gain comes to 140 hours per selling campaign, concentrated in the weeks when the team is busiest. These figures only hold for their assumptions. Replace them with your own measurements, and add what the example leaves out: the cost of the tool, training, and the sales made possible by the hours freed up.

The limits of studies on time saved

Each of the studies cited has its blind spots. Before quoting a figure in an investment case, keep five caveats in mind:

• None of them covers fashion wholesale: they measure spreadsheets, sales forces or invoices in general.
• Contexts differ: large companies, countries, years of data collection; a 1990s spreadsheet audit or a figure from 2014 or 2020 describes a different state of technology.
• Some figures are potentials, not observed gains: “automatable” does not mean “automated”.
• Projections often overestimate: according to the European Commission’s REFIT Scoreboard, which draws on the 2020 evaluation of the second invoicing directive (2010/45/EU), the expected savings from e-invoicing had been estimated in 2010 at up to €18 billion, assuming all invoices moved to electronic format, against €1.04 billion actually achieved over 2014–2017. What was achieved is less than 6% of the high estimate.
• The gain depends on usage: a poorly adopted tool, bypassed by a parallel Excel file, saves nothing.

Measure first, choose the tool second

Studies help you ask the right questions; they do not put a number on your return on investment. The order of steps follows from that. Start by timing about ten orders from the current campaign and counting last season’s errors: that baseline is yours, and no one can dispute it.

Then compare the available approaches with ALTORE versus the other solutions on the market, and run them through the guide to choosing B2B order-taking software. Once the tool is in place, repeat the same measurement over a full season. That is the only figure that will stand up in front of your business partner or your banker.