Are we back to the ‘70s yet?

No, this isn’t a blog about energy crises, supply chain issues, spiralling cost of living or overworked medical services even though these topics do seem to be somewhat familiar.   This is about the increasing trend towards adopting standardised software solutions to run businesses using outsourced platforms – and the required managerial disciplines – as an enabling process.

In preparing for the launch of a market wide survey of attitudes towards Cloud Migration undertaken by Intelligent Linking, K2 Consulting Partners have looked to see how the current trends fit into a historical perspective. 

Some of us, entering the IT industry in the late 1970s as freshly minted graduates with irrelevant degrees but passing logic and algorithm aptitude tests, took up roles as ‘trainee programmers’ – this being before the days when coding skills were passed on though DNA from the mother to prepare the child for a career with Google, Facebook or gaming software companies.   The ‘old hands’ would explain that the systems then being run were brought in house from bureau computer processing companies, usually applications such as payroll and accounting.   Newer systems had been developed, e.g. ‘Stores and Inventory’ using small armies of new minted graduates and these now needed to be maintained.  In fact, everything needed to be maintained, from payroll systems still processing Pounds, Shillings and Pence (8 years after decimalisation) to an Engineers Stores system which somehow couldn’t keep an accurate number of items in stock.   (The storemen were keeping some items on the side in case someone needed them?!?!?).   It soon became apparent that the earlier annual influxes of graduates had responded to apparently legitimate requests for enhancements – most of which could no longer be traced back to the original requestor – and a lot of the systems were now developing unplanned functionality seemingly all on their own.

Since this was ‘before the Dawn of History’ – to misquote Spinal Tap – and was about as far as any of us could remember, my colleagues and I in K2CP used this as the start point for our discussions.  If you are Grace Hopper, Tommy Flowers or Countess Ada Lovelace you will be able to go back further than this.

So here are our recollections of the main stages in the use of business computing over the past 40 years or so.

Dawn of history – standardised bureau processing for systems where functionality was mandated by standards or legal requirements, e.g. payroll and accounting.   The software was standard – and run by someone else – and the only variation between customers was the data.

Move In house – development of individually developed systems, strung together in ‘batch runs’ to process everything from sales orders to manufacturing planning and purchase orders with centralised HR and Accounting using ‘packaged software’.  It wouldn’t always work if new graduates had recently been hired.

Adoption of integrated software solutions (Part 1 – Business Process Re-engineering) – after careful analysis of the As-Is solution (usually consuming 80% of the budget) a detailed To-Be solution was developed (consuming an additional 80% of the budget) indicating all the additional customisations required to ensure that the IT solution would deliver ‘competitive advantage’.   A further 160% of the original budget was then used to implement the solution but, interestingly, not all of the original ‘absolutely essential’ custom components were implemented.  This was usually because either (a) no-one could track down the original requester and rationale – see ‘before the Dawn of History’ above, or, (b) everyone was keen not to ask for more money for something which was already 220% over budget.

Adoption of integrated software solutions (Part 2 – Package Enabled Re-engineering) – people figured out that since the software was going to work the way the software worked, that the first 160% was a bit of a waste and that it would be better to focus on just over-running the budget by 60% and re-engineering the business to work the way the software worked.  This had the added advantage that people could ask for more customisations to really get a ‘best of breed’ solution using different suppliers’ solutions and still get nowhere near the earlier budget overruns.

Adoption of industry standard processes – software companies used industry standard processes to show that their systems achieved the same core business results as a so-called optimal solution e.g. purchase orders were issued, invoices received and paid, employees were hired and paid, etc. and made the case that it was cheaper in the long run to use the standard solution.   Having sold the licences and implemented the solution ‘in house’, software company revenue targets were achieved and this allowed a new set of graduates – more expensive ones with STEM degrees, supplied from large consulting firms – to get to work on developing customisations.

Move to the Cloud – the adoption of standard business process solved part of the problem and allowed companies to focus on quality of product development, customer service and innovation.   In retrospect, this standard-process phase of business computing now also looks like an enabler for the early Cloud proposals that companies should no longer have to run their own systems.   If companies want standard solutions for standard business processes achieving standard business outcomes, then this can be run by third parties for a monthly or annual fee, thus avoiding the need for capital investment and the continual recruitment of scarce IT skills.   The downside?  A business must actually take the standard solution and put in place the disciplines required to manage the improvements / upgrades pushed out by the service supplier.   This is true for Infrastructure as a Service, Platform as a Service and Software as a Service.   Each of the above has increasing benefits for companies who can discipline themselves to take the standard route.

So, are we back to ‘70s yet?  No.  The range, sophistication, and scope of software solutions from the main vendors is far beyond the simple bureau processing of the past.  If we are willing to take the best of both, namely, the willingness to adopt standard solutions in a disciplined way, coupled with the significantly enhanced functionality now on offer we might now have finally reached a fit-for-purpose solution.   In this solution, software companies make the investments in developing and improving functionality (they will need graduates…) but the end business can focus on the things they want to be good at.

As a postscript, Thomas J. Watson (IBM Chairman and CEO 1914-1956) once said: “I think there is a world market for maybe five computers.”  He has often been quoted as an example of how even the most powerful and successful can occasionally be wrong.   However, if you replaced the word ‘computers’ with ‘Cloud Services suppliers’ he may well have been 70 years ahead of everyone.