The most important decision an Agency FD will make

job costing

Every Agency FD makes dozens of decisions daily.  Today will be no different. Some are trivial, some more important. Cumulatively they add up to the impact an FD has on the performance of their Agency.<
Often it is about deciding what to spend your time on. Maybe cash-flow is pressing or there is a pitch cost to review . There is no shortage but there is one decision which dwarfs all of the others.

There is one call that is down to the FD alone and it will impact their working life and the agency profoundly. It will affect what information is available, in what format and how quickly it is available. It will decide ultimately whether you can do a good job or not.

This decision is which job costing system to choose. Less time will be spent by most Agency MDs about this hugely important strategic decision than about which stationary to order. Partly because they have plenty of other things to worry about but mainly because it falls so squarely outside their remit they will leave the decision to the FD.

There should be little mystery about job costing systems but there are some poor ones around which can be hugely expensive in both time and money. The wrong one can also hamstring you for a long time during which your decision making will be based, at best, on incomplete information.

The good news is that it is relatively straightforward to eliminate the bad ones and find yourself with one which is fit for purpose.

The first point to make is that this decision must be about job costing rather than agency workflow. Because time sheets are a fundamental component of any job costing system it is tempting for both agencies and the developers to add on a resource planning module. Don’t. Never do it. At least not until you have the systems and resource in place to do it properly. The level of complexity this introduces into the programming is such that it can overwhelm the simplicity of what a good job costing system should be about. In addition I have never seen a system that is as good as a traffic manager.

I’ve mentioned simplicity. Job costing systems should be based on a very, very simple principle. It is all about comparing actual time and costs to the estimate for the job. This can be sliced by client, by job, by function but everything is about actual versus estimate. Job costing modules which want to be resource planners add complexity which makes this simple objective harder to achieve.

Another complexity which, whilst tempting, should be avoided is the urge to divide projects into ever smaller categories so that you can analyse the revenue by type of activity. Think carefully about what your agency does and I’m sure there will be a number of broad categories such as on and offline which would be good to know. Beneath that it is useful to know how online breaks down into, say, website versus email or how media breaks down into radio versus press. Beyond that though you will need to devote a large part of your time making sure every job is correctly categorised. Compare that to the benefit of analysing your revenue to that degree.

Another key question is whether to get a stand alone job costing system or an integrated solution. I have to confess to a personal preference here as I would always, always go for an integrated package. Having to integrate your job costing system with another accounts system has always struck me as an unnecessary complication. I’ve also found integrated systems to be more robust and efficient – I want both the job and nominal updated when something is posted, I don’t want to have to wait until the batch file is processed to know what the new balance is. I also believe that having to worry about the integrity of the nominal ledger is a very useful discipline for programmers of job costing systems.

I do sometimes wonder with some job packages whether an accountant has ever been involved in the development process. Too often stand alone job packages have rules that no experienced FD would insist on. My favourite rant at the moment is about the insistence that invoices can only be entered if there is a matching purchase order. Great theory and easy to programme as you only have one process to follow but why, oh why do you have to raise a PO when you already have an invoice which has been signed off? Lazy programming and divorced from reality. Rant over.

Another key issue for me is reporting. Any system is pretty much useless unless you can get information out of it quickly and easily. Again it’s a personal preference but I’m very wary of a sales pitch which says you can get whatever information you want, however you want it. Flexibity is great but I don’t want to write reports; I want standard reports showing me unbilled expenses or actual time versus budgeted time for a specific range of clients or jobs. It’s also comforting to know that whoever has developed the system either knows or has listened to people who know about the information required to manage an Agency.

A good job costing system will cover 90% of what you need with standard reports which only need you to select the data range. A good job costing system provider should work with you to deliver the final 10% for you.

So, true to an earlier article, here is my checklist for choosing a new job costing system.

1) Keep it simple and focus on the core functionality. Good looking systems which act as a resource planner should come with a health warning.

2) Can you try it in action before you buy it? Preferably with a play system or a site visit to a happy client. Try before you buy it.

3) Be wary of systems with too many parameters which don’t reflect the way you work.

4) Define the information you need on a regular basis and ensure these reports are available.

5) Read the small print of the contract. I recently came across a contract from a software provider which said “All warranties implied or expressly imported by statute including, without limitation the implied warranties of merchantability, fitness for a particular purpose, title, non infringement and quality of service are hereby disclaimed”. Naturally that didn’t inspire confidence.

6) Only integrate with a separate accounts package if there is a clear cost or functional benefit from doing so.

7) Think about any related process or resource implications at the same time. Is it going to increase the decrease the processing time? Unless the additional time is needed to improve processes rather than manage the system is it really the right system?

There are good and bad systems out there and I’ve worked with both. If your Agency is going to grow and become more profitable it will be a lot easier if you have a decent system in place to help manage that growth. If you need some help finding and installing the right system for you email me on