The problem with programs like Accountant is that they are written by accountants, not by practising business managers. If this sounds revolutionary (or just plain daft), let me explain:

Accounting performs two functions for a company. The first is to satisfy those interfering outsiders: the auditors, the Inland Revenue, the VAT man and the shareholders, all of whom, quite properly, want assurances that the business is being honestly run and profits are being declared and not put straight into the boss's pocket. "Audit accounting" tries to ensure that every invoice or cash payment and every nut and bolt, taxi fare or telephone call they represent, is recorded and included in the overall year-end accounts, so that an auditor can certify them as honest. This is the task that Accountant has obviously been designed to do and it does it pretty well.

I am the last to decry this process which safeguards everyone: customers, shareholders, creditors and employees. Unfortunately, it tends to overshadow the other function of accounting, which is to tell the manager responsible for a business operation exactly what is happening, why it is happening and what is going to happen tomorrow and next year. This process is called "management accounting" and neither Accountant, nor dozens of similar programs, does it.

This is no place for a lecture on the topic so I will confine myself to three illustrations:

Firstly, Accountant does not allocate costs and revenues to "profit-centres"; it conventionally groups costs by "account", such as "stationery", "telephone", "depreciation" or "consumables" (which is important and essential), and then groups accounts by "group", in order to combine them when presenting annual profit-and-loss or balance sheet. Thus it offers no guide to the profitability of one division of product or project; essential for the manager to know which to expand, which to close, where to direct his advertising budget, his capital investment or his time. It does take one first and most valuable step: it offers a facility for constructing a budget and compares "actuals" with it. It starts to take a second by allocating a box for "type" of each transaction, but then draws hastily back, saying "for reference on printouts only".

Likewise with customer accounts: the best reason for keeping them is not to please the auditors but to make sure your customers pay and pay promptly. Accountant duly produces an "account" of debtors but does not break it down into individual customers. As one cannot ring up a customer and say "I am owed 10,000 by all my customers put together, so pay up", this is pretty inadequate information!

Thirdly, and perhaps worst of all, the dreaded words "cash-flow", which brought Freddie Laker and a million lesser tycoons to their knees are nowhere mentioned. This is the most vital function of accountancy; without it the program should be renamed BookKeeper, else Rolls Royce died in vain!

So much for tears over what Accountant will not do; now consider what it does. It is a formidable program with 45 separate parts and 20 different files. To make it easy to change from a manual to computerised system, it follows ordinary manual procedures for entering sales, purchases and cash transactions in day books and analyses them into the separate accounts of a "nominal ledger". This makes it possible to transfer only part of the system, for example the cash book, at one time.

The budgeting process is particularly simply presented, there is a choice of no budget, or of breaking the year's budget into quarters, calendar or lunar (four week) months and of spreading the budget equally across the year or variably. Likewise the selection of the nominal ledger codes has enormous flexibility, coupled with strong support of suggested codes broken down in great detail. Most users will need to make only minor changes to these and there is a sensible procedure at the start for printing out those that are selected as a permanent memorandum for the clerk who has to type in the data.

The main anxiety expressed by my own accountant when I discussed computerising my accounts, was that one can end up with worse confusion than with a manual system. To avoid this, Accountant insists on printing out "audit trails" (details of all the transactions that have been entered), with their postings to the nominal ledger, to the debtor control account and to the VAT accounts. It will produce a completely adequate VAT return, query any account and its transactions and produce nominal ledger trial balances, either at the year end or at the end of the accounting period.

A variance report analyses the actual expenditure against budget for each account. It insists that all this be carried out before the month-end routine which, in order to save space on floppy disc, clears all individual transactions from the files and only carries forward month-end balances. It warns you to take a back-up copy of the data disc before doing this and to have produced all the necessary reports and audit trails.

A model accounting system is provided with some clear explanations of how the results of Accountant can be used at the year end to produce profit-and-loss accounts and balance sheets. A full range of balance sheet accounts for reserves, suspense balances and capital account is provided and there is provision for special cases such as partnerships. If other outputs are required, these can be produced (by those sufficiently skilful), in unlimited variety, by use of Nucleus, with which Accountant is fully integrated.

There are some detailed criticisms. Most arise because accountants are obsessed with accuracy, whereas what businessmen are looking for is less work and more speed:

1. I can pardon American spelling of "disks" but not "enter" for the "return" key.
2. All single-key responses have to be followed with a "return"
3. It leaves the operator to calculate VAT quantities.
4. It uses a "batch" entry system; you have to calculate manually the total value of all the transactions in the batch.
5. There is no facility for the "VATless" user to avoid all references to VAT in the program; instead he or she has to make countless nil entries

I have tried many book-keeping and accounting programs designed for the BBC B on its own and found most of them hopeless. I am impressed with Accountant's lucid hand book, the clarity of its screen explanations and the comprehensive range of its operations. A potential user could get the following advantages:

1. A clean, crisp set of books with full audit trails, that will reduce the amount of work that his accountant has to do at the end of the financial year and thereby reduce his accountancy costs.
2. Complete VAT returns at year-end or more frequently if required.
3. The rapid provision of comparisons of actual expenditure and sales against budget, soon after the month end. Normally the work required to do this manually delays it beyond the time when effective action can be taken.
4. Rapid production of trail balance so that any inquest can be carried out while details are fresh in people's memories.
5. Rapid solution of queries.

Despite my earlier reservations, there are substantial benefits. The price that will be paid to achieve them is that:

1. It will undoubtedly take longer to post transactions using Accountant than simply to write them into day-books.
2. If the office is in a muddle when Accountant is first used the process will be either uncomfortable or totally disastrous and time will be wasted to produce poor results at the end. However, the improvement of office discipline and tidiness is in itself a great advantage, not likely to be achieved by any other method!
3. Regularity of book-keeping becomes essential. My lazy habit of letting six months' invoices and receipts accumulate and then shutting myself away for three bad-tempered days would not work at all well under the Accountant regime.

To the businessman willing to adapt in this way, I would offer the following advice:

1. Play around with Accountant for two or three days. Practise entering samples of your own transactions and practice the month-end routine before you delegate anything.
2. If as a result you are confident your staff are able to use Accountant, ask your own accountant to examine these tentative efforts, study the hand-book and confirm that there are no obvious pitfalls.
3. For at least the first month (and many people would urge for the first year!) continue keeping books by hand, for safety's sake. It is highly probable that data will be lost most during the first few weeks.
4. Buy enough (good duality) floppy discs so as to make two backups of every data disc, each day
5. Spring-clean the book-keeper's office and organize sufficient filing trays, lever-arch files and other aids to being systematic. Arrange for the cleaners to dust daily and ensure windows are not left open so papers blow around.

If these careful procedures are carried out I see no reason why Accountant should not produce an excellent set of books. It will not save money, indeed may actually increase book-keeping costs. However it will improve control which, if the business has potential for profitability, will soon recover these costs. However it does not provide an alternative to effective management accounting and careful budgeting and control of cash-flow.