The Details You Type Twice: A One-Week Audit of Duplicated Admin

Every small firm has a version of the same habit. A client's address is typed into the quoting tool, then again into the invoice, then again into the courier label. A VAT number is lifted from an email and pasted into an accounts package, then checked against a spreadsheet, then entered a third time when the purchase order is raised. Individually, each act takes seconds. Collectively, they consume a slice of the week that nobody has ever measured, and they introduce a category of error that is difficult to trace and expensive to correct.

The difficulty is that repetition of this kind is invisible in the way that a missed deadline or an unpaid invoice is not. It does not announce itself. It sits inside otherwise productive working hours, disguised as diligence. The remedy, therefore, begins not with a tool or a template but with a record.

Begin With Evidence Rather Than Instinct

It is tempting to skip the measurement stage. Most owner-managers can name the processes they find tedious, and the temptation is to fix those first. Tedium, however, is a poor guide to cost. A task that feels irritating may occur twice a month; a task that feels unremarkable may occur forty times a week. Without a log, the firm reforms the wrong thing and concludes that administration is simply resistant to improvement.

A single working week is usually sufficient to establish the pattern. Choose a week that is representative rather than exceptional — avoid the week before a VAT return is due, and avoid a week in which half the team is on leave. The aim is not precision to the minute but a reliable ranking of offenders.

The Log Itself: Four Columns and a Willing Team

A shared spreadsheet is adequate. Paper on the desk of each participant is often better, because it is faster to reach and less likely to be forgotten. Whichever medium is chosen, four columns carry the whole exercise:

• The detail: what was copied. Client address, VAT number, job reference, bank details, contact name, delivery instructions.
• From and to: the source system and the destination system. This is the column that reveals whether the problem is a missing integration or simply a missing template.
• The count: a tally mark each time it happens. Tallies are preferable to a running total of minutes, because people estimate time badly and count occurrences accurately.
• The risk note: a brief flag where the copied detail is one that would be costly to get wrong. Bank details and VAT numbers warrant a flag; a postcode rarely does.

Ask participants to record entries as they occur, not at the end of the day. Recollection compresses a week into a vague impression within hours. It is also worth stating plainly at the outset that the exercise is not a performance review. If staff believe the log will be used to judge their speed, they will under-report, and the audit will produce a comfortable fiction.

Reading the Tally: Where the Week Actually Goes

At the close of the week, the data is rarely surprising in substance but frequently surprising in scale. Sort the entries by count and by risk separately. A detail copied sixty times carries a different kind of cost from a detail copied six times but flagged on every occasion.

Three patterns tend to emerge.

First, a small number of details account for the majority of the copying. Typically two or three fields — usually the client address, the job or order reference, and the bank details — dominate the log. These are the candidates worth attention.

Second, the copying clusters around a handful of handover points: quote to invoice, invoice to payment run, order to delivery note. The number of systems involved is often smaller than assumed; the gaps between them are the problem.

Third, much of the repetition happens at the boundary of a role rather than inside it. One person holds the information, another needs it, and the transfer is done by retyping rather than by reference. That is an access question dressed as a typing question.

Choosing the Route: Template, Delegate or Automate

Once the offenders are ranked, each one deserves a deliberate decision rather than a general intention to be more efficient. Three routes are usually available.

Templating is the cheapest and fastest. A saved reply, a client record with fixed fields, a standard job sheet, or a browser's autofill function will remove a substantial share of retyping at negligible cost. It suits details that are stable and used by one or two people.

Delegation suits high-frequency, low-judgement copying that must still be checked. A virtual assistant can maintain the transfer between systems provided there is a written procedure describing exactly which field goes where and what to do when a detail is missing. Delegation without that document simply relocates the error.

Automation, in the sense of a direct link between two systems, is the most expensive route and should be reserved for the largest offenders. It is worth costing honestly: setup fees, subscription increases and the time required to test the connection can easily exceed the value of the typing saved if the volume is modest. Where the log shows a detail copied four hundred times a year, the case is straightforward. Where it shows forty, a template will usually do.

The Risk Column Deserves Separate Attention

Cost is only half the picture. Repeated manual entry of bank details is a recognised fraud vector: an intercepted email, a plausible-looking instruction, and the payment goes astray. Any detail flagged in the risk column should be reviewed on risk grounds even if the frequency is low. Reducing the number of times a bank detail is copied reduces the number of occasions on which it can be copied incorrectly or from the wrong source.

Setting a Review Date That Will Hold

An audit of this kind fails in a predictable way. The findings are discussed, two templates are created, and by February the old habits have returned. The countermeasure is a dated review with a named owner and a defined test.

Choose a date approximately ninety days out — for a firm running this exercise in November 2026, the second week of February 2027 is a sensible target, and it avoids the year-end period. Record three things in the diary entry itself: who owns the review, which offenders were selected for action, and what success looks like. Success should be measurable against the original log, which means repeating the tally for a second week and comparing counts. A reduction of half on the top three offenders is a realistic and worthwhile outcome; the elimination of all repetition is not.

Finally, circulate the outcome. A short note recording what was changed, what it saved, and what was deliberately left alone gives the exercise a conclusion and prevents it from being re-run from scratch next year by someone starting again with a blank spreadsheet.

The detail you type twice is rarely anyone's fault. It is usually the residue of a process that grew without review. A week of tally marks is a modest investment, and it converts a vague sense of wasted time into a list of decisions that can actually be made.