I have learnt the hard way that trying to outsource on the basis of “manage my mess for less” is a sure fire way of crashing and burning at worst and being more expensive at best. Anything that is crucial to your firm’s success, you do not outsource. In other words, only outsource which is a commodity and it is easy to switch suppliers such as perhaps storage management, electricity supplies, sewage, catering, cleaning, etc.
Lo and Behold, here’s a paper which provides some more data to back up the idea that outsourcing actually pushes up your costs. The data used is crucial and I quote:
ITOS and IT spending data were obtained from InformationWeek magazine, a weekly print magazine aimed at business technology professionals. Since 1991, InformationWeek has conducted annual surveys to gather current year IT budgets from a variety of the largest US public and private firms and government entities that use IT. It has supplemented this with firms demonstrating innovative use of IT. In 1999, InformationWeek began asking firms what percentage of their IT spending is outsourced. InformationWeek recently provided the additional firm-level data for the 1998 to 2005 time period for this study with strict disclosure restrictions on the authors precluding the sharing of specific firm responses. Data are used from respondents who provided both IT spending and ITOS information for one or more years between 1999 and 2005. Observations for non-public firms were eliminated and merged with corresponding Compustat financial data to calculate the various control variables shown to affect IT budget levels in Kobelsky et al. (2008b). This reduced the overall sample to 1959 firm-year observations for 647 firms in the period 1999 to 2005
The model comprises of variables:
• itb/sls = firm IT budget for each year per InformationWeek data divided by sales for that year (Compustat data12);
• itos dummy = 1 if IT outsourcing percentage per InformationWeek data is positive in Current Year, 0 if not;
• size = log of Current Year sales;
• ind_conc_ratio = four-firm concentration ratio for four-digit SIC;
• uncertainty = standard deviation of earnings before extraordinary items for previous 5 years scaled by sales;
• rel_divers = related diversification (within 2 digit SICs);
• unrel_divers = unrelated diversification (across 2 digit SICs);
• op_ros = operating return on sales, before depreciation (compustat data13/data12);
• debt_ratio = debt ratio (Compustat data9/data6);
• ave_sales_growth = average sales growth for last two years;
• automate = 1 for firms in automate industries, 0 otherwise;
• transform = 1 for firms in transform industries, 0 otherwise;
• hi_tech = 1 if high-tech firm, 0 otherwise;
• lo_tech = 1 if low-tech firm, 0 otherwise.
- • year = 1 for each year 2000-2005, 0 otherwise.
90% of the sample companies partake of outsourcing some or all of their IT activities. The authors find that while on a project level, they might see a reduction in the IT costs and spend, on an aggregate firm level, the IT spend actually goes up. Note that they do control for scope and volume changes by looking at the sales growth. Within two years of outsourcing, the IT cost level of firms who have outsourced is correspondingly higher than firms which have not outsourced. While the authors suggest that this is because of capabilities are enhanced, I have my doubts. One cannot improve IT capabilities in 2 years, it is simply not possible to evolve the business and IT side so quickly that a statistically significant improvement in productivity and quality can be observed. It is, in my opinion, clearly aimed at the fact that the business case is frankly wrongly specified and outsourcing doesnt really help as far as cost control is concerned.
Business cases are rarely expressed in terms of ratio’s, in other words, you will very rarely find that the managers concerned or the IT outsourcing firm are quoting you IT costs as a ratio to say the sales revenue or operating costs or profits of the firm. This is why I am very nervous whenever I hear that outsourcing is happening which is going to drive down costs.
There is a good argument to outsource to improve efficiencies, drive a centre of excellence, to improve productivity, but for cost purposes, the figures do not bear out the benefits.