New Zealand is a nation of small businesses. Out of a total of nearly 300,000 local enterprises less than 3,000, or a mere one percent, employ more than 50 people, according to the latest figures available from Statistics New Zealand (see figure 1).

Some of the country’s smaller companies have good reasons for staying small. If you are catering for a niche market then obviously your growth potential is limited by the size of that market. Certain industries also favour smaller players - advertising agencies don’t benefit from the same economies of scale as telcos or airlines for example – and many firms are doing quite nicely as micro enterprises. According to The National Bank’s Small Business Monitor the net profit margin for firms of between 1-5 employees is double the average for other business sizes (see figure 2).

February 2006

By Michael Foreman

But for most companies not only is growth desirable but it can be necessary for survival, especially when competitive pressures demand increased efficiency. Manufacturers can often realise dramatic reductions in their per-unit production costs by increasing their sales volume, just as retailers can extract better prices from their suppliers when their purchasing volume grows. Growth can also help businesses retain employees, either by virtue of the opportunities it creates for rewarding talent or simply by making a company a more interesting place in which to work.

Horses for courses
In his book, Mastering the Rockefeller Habits, US-based ‘growth guru’ Verne Harnish describes a series of evolutionary stages in the growth of a company from a small firm to a large company, and the barriers which must be overcome in order to leap from one stage to another. Harnish argues that the essential ingredient that a company needs to grow is leadership, of which the two key elements are the ability to delegate and the ability to predict. The ability to predict becomes progressively more important as the company increases in size and it is the overriding factor in determining the type of technology that is required at any stage.

“When you go from two employees to ten, you need better phones systems and more structured space. If your company goes to 50 employees, you still need space and phones, but suddenly you need an accounting system that shows more precisely whether, which, and how projects are actually making money.”

From 50 employees up, which under US conditions equates with a revenue of US $10 million to US $50 million, Harnish says that typically all the information systems need to be upgraded. Above US $50 million they will need to be revamped again, as the organisation tries to tie all systems to one database of customers and employees.

According to National Bank statistics, the companies in New Zealand which are showing the fastest change in their financial ratios – usually evidence of growth – are in the 21-50 employee range with a turnover of between $4 million to $10 million. A 50 employee company in New Zealand will have an average turnover of around $10 million, and this is generally regarded locally as the point where a business makes the transition from being a medium sized company to a large one.

When a company grows beyond this point, Harnish says the organisation is expected to have enough experience and position in the marketplace that it can accurately predict profitability. Profitability has of course been important all along the way but at this stage it becomes more critical to predict it more precisely, since small percentage swings either way can represent millions of dollars.

“The fundamental journey of a growing business is to create a predictable engine for generating wealth as it creates products and services that satisfy customer needs and creates an environment that attracts top talent,” Harnish concludes.

The importance of predictability received high profile media attention lately when Carter Holt Harvey’s independent directors were forced to make an embarrassing about turn and support Graeme Hart’s takeover bid following a big profit downgrade. The independent directors, led by Kerry McDonald, said: “The quality of management’s forecasting makes it difficult to reliably predict long-term cashflows and valuation; limits the ability to separate short-term cyclical movements from longterm structural market developments; and limits the ability to separate short-term execution and timing issues relating to productivity improvements from the longer-term benefits.”

Delivering predictability
The generic word which best describes the type of technology which can deliver the level of predictability Harnish calls for is a business intelligence (BI) system, but according to local software vendors we spoke to, such systems are playing a wider role at all stages of company growth. By revealing patterns that would otherwise lie hidden in data BI systems can help managers make the right decisions that lead to growth as well as make sense of the complexity or sheer scale of an operation once that growth has been achieved.

Wakenshaw says he came across an example of this when he was involved in an early application of BI at the Land Transport Safety Authority and he decided to “play around” with some crash analysis figures.

“The more data you have to analyse, the more useful BI systems tend to be,” says Wakenshaw. “Banks, publishers, or grocery chains who have large customer databases or sales data going back several years will find these systems most useful of all.”

“A grocery chain could see what’s selling and where and compare that with a year or two back, they could look at what happened to a single product in mid-November, and then look at the same period for the past few years.”

Wakenshaw says such analysis might show a retailer for example that sales of certain cuts of beef have declined due to a trend by consumers to reduce the fat in their diet.

“It’s the little trends like this that a BI system can reveal. They would be hard to see otherwise if you were selling 20 tons of beef a week. You can see your data more clearly. That in essence is what BI systems are all about.”

RTI provides two-day BI training courses which basics of how to organise the data by clicking and pointing on data sets to formulate simple or complex ad hoc queries and reports.

“It’s not too hard to get your head around, all you really need is a basic understanding of the business,” he says.

While BI systems may not always produce life saving results, Wakenshaw says software like the Hummingbird BI query and reporting application can be very useful in the day to day running of a company by helping users to ask questions and visualise the answers for more effective decision-making.

“For example it’s easy to set up a digital dashboard - press ‘F2’ and you get a report on how the company performed yesterday. It might tell you that 3% of your workers reported sick and you could compare that with the normal trend for a weekday or you could drill down and find out exactly who was off sick and how many days they had missed on other occasions.”

Keeping on track

For example where a conventional ERP report might show sales by customer, a good BI tool will allow the user to see sales by location, by product, or by groups of products, groups of stores, or almost any other combination. (see figure 3 for example)

“You can drill down to the smallest possible unit, and you can retrieve product or customer information,” says Biggs. “Now the sales manager can see where past sales have been made and where they are tracking.”

Biggs says SQL Analysis Services, which is available as a free add on to Microsoft’s SQL Server, is a very effective tool for BI purposes, and for smaller companies it is probably a less expensive option than a standalone business intelligence package. When Complete Solutions configured and installed SQL Analysis Services at a local food manufacturer the company benefited greatly from being able to rapidly and interactively analyse its sales figures.

The system soon revealed for example that demand for some of the food company’s products was seasonal, while other products were more popular in some locations than others.

“Before we installed the system one of the company’s biggest problems was shortages of product,” says Biggs. “In the cuisine business tastes can change very rapidly and so the more information you have about product sales the more prepared you are to avoid stock outs and short supply.”

In the year since the BI system was installed the food company’s sales have grown by more than 30%.

“We can’t take all the credit for this but we think we have contributed,” says Biggs.

Scaling the heights

One of the main BI products Cortell supports is the TM1 application developed by US-based Applix, and Cortell’s local client list include some large household names such as New Zealand Post, ASB Bank, and Telecom New Zealand.

While most organisations initially deploy BI systems for financial applications such as forecasting and budgeting, Hill says an increasing number of companies are using BI for non-financial applications such as production planning.

Hill says TM1 is based on OLAP (On-Line Analytical Processing) technology which allows users to quickly view and understand large sets of complex data. At Telecom for example, TM1 is being used to analyse 30 million transactions a week, a process which takes about 15 minutes. At the other end of the scale, Hill says Cortell has successfully installed a BI system for a much smaller customer which runs in conjunction with a MYOB accounts system.

BI for the masses
Chris Mitchell, a director of Targit NZ, says a drawback of some BI packages is that they are too sophisticated and that people need to have a high degree of business analysis skills to use them. But Targit has been designed from the ground up as an end user tool which can easily be deployed out to middle management.

“The philosophy behind it is to provide BI for the masses. It’s a tool that can be used by 90% of the organisation rather than 20%,” he says. Mitchell says the package includes the ability to restrict the information managers can see to what is relevant to their individual roles and parameters can be easily set centrally by a point and click process. These parameters can then be used in traffic light-type displays which show unacceptable profitability for example as red while desired levels or outcomes are shown as green.

In the US where Targit was developed, the package has proved to be popular in the retail industry but early local take up has been in industries like utilities, insurance and distribution. Mitchell says Northpower has used the package very effectively to manage electricity network maintenance (see Northpower case study).

Mitchell says Targit will work with virtually any relational database including Oracle and DB2 although the package has been designed around Microsoft’s .NET architecture. “If you are a Microsoft or Navision or Axapta user we have developed five pre-built business analytics and it’s just plug and play.

For more information visit the Business Intelligence Research Pavilion