CASE STUDY subtext Studios Singapore

PLEASE ANSWER THE 2 QUESTIONS AT THE END OF THE CASE STUDY.

C.K. was clearly upset. Since he had founded subtext in the fast-growing South East Asian computer-generated imaging (CGI) market, three years ago, this was the first-time that he had needed to apologise to his clients. In fact, it had been more than an apology; he had agreed to reduce his fee, though he knew that didn't make up

for the delay. He admitted that, up to that point, he hadn't fully realised just how much risk there was, both reputation and financial, in failing to meet schedule dates. It wasn't that either he or his team was unaware of the importance of reliability. On the contrary. 'Imagination', 'expertise' and 'reliability' all figured prominently in their promotional literature, mission statements, and so on. It was just that the 'imagination'

and 'expertise' parts had seemed to be the things that had been responsible for their success so far. Of course, it had been bad luck that, after more than a year of perfect reliability (not one late job), the two that had been late in the first quarter of this year had been particularly critical. 'They were both for new clients,' said CK, 'and neither of them indicated just how important the agreed delivery date was to them. We should have known, or found out, I admit. But it's always more difficult with new clients, because without a track record with them, you don't really like even to admit the possibility of

being late.'

The company

After studying computer science at the National University of Singapore, C.K. Ong had worked in CGI workshops in and around the Los Angeles area of California, after which he returned to Singapore to start subtext Studios. At the heart of the company were the three 'core' departments that dealt sequentially with each job taken on. These three departments were 'Pre-production', 'Production' and 'Post-production'.

Pre-production was concerned with taking and refining the brief as specified by the client, checking with and liaising with the client to iron out any ambiguities in the brief, story-boarding the sequences, and obtaining outline approval of the brief from the client. In addition, pre-production also acted as account liaison with the client and were also responsible for estimating the resources and timing for each job. They also had nominal

responsibility for monitoring the job through the remaining two stages, but generally they only did this if the client needed to be consulted during the production

and post-production processes.

Production involved the creation of the imagery itself. This could be a complex and time-consuming process involving the use of state-of-the-art workstations and CGI software. Around 80 per cent of all production work was carried out in-house, but for some jobs other specialist workshops were contracted. This was only

done for work that subtext either could not do, or would find difficult to do.

Post-production had two functions; the first was to integrate the visual image sequences produced by Production with other effects such as sound effects, music, voice overs, etc., the second was to cut, edit and generally produce the finished 'product' in the format required by the client.

Each of the three department employed teams of two people. Pre-production had two teams, Production three teams, and Post-production two teams. For Pre-production and Post production work, one team is always exclusively devoted to one job. 'We never allow either one team to be working on two jobs at the same time, or have both teams working on one job. It just doesn't work because of the confusion it creates. That doesn't apply to Production. Usually (but not always) the Production work can be parcelled up so that two or even all three of the teams could be working on different parts of it at the same time. Provided there is close coordination between the teams and

provided that they are all committed to pulling it together at the end, there should be a more or less inverse relationship between the number of bodies working on the job and the length of time it takes. In fact, with the infamous 'fifty-three slash F' job that's exactly what we had to do. However, not withstanding what I just said about shortening the time, we probably did lose some efficiency there by having all three teams working on it. Our teams generally work until the job is finished. That level of work is factored in to the time estimates we make for each stage of the process. And, although we can be a little inaccurate sometimes, it's because this type of thing is difficult to estimate' (C.K. Ong).

The fifty-three slash F job

The fifty-three slash F job, recently finished (late) and delivered to the client (dissatisfied) had been the source of much chaos, confusion and recrimination over the last two or three weeks. Although the job was only three days late, it had caused the client to postpone a presentation to its own client. Worse, subtext had given only five days' notice of late delivery, trying until the last to pull back onto schedule.

The full name of the job that had given it so much trouble commitments, the GDA set up an independent body to monitor the project. All potential contractors tendering for parts of the project were aware that a major underlying objective of the Games initiative was regeneration. The Games site was to be built on highly industrialised and contaminated land. The supermarket's new logistics boss was blunt in his assessment of its radical supply chain implementation. 'Our rivals have watched in utter disbelief', he said. 'Competitors looked on in amazement as we poured millions into implementing new IT systems and replaced 21 depots with a handful of giant automated "fulfilment factories." In hindsight, the heavy reliance on automation was a big mistake, especially for fast moving goods', said the company's CIO. 'When a conventional facility goes wrong, you have lots of options. You have flexibility to deal with issues. When an automated "fulfilment factory" goes wrong, frankly, you're buggered.' Most damning was the way that the supermarket pressed on with the implementation of the automated facilities before proving that the concept worked at the first major site. 'I'd have at least proved that one of them worked before building the other three', he said. 'Basically, the whole company was committed to doing too much, too fast, trying to implement a seven-year strategy in a three-year timescale.' 'It's impossible to overemphasise just how important this launch is to our future', said the CEO. 'We have been losing market share for seven quarters straight. However, we have very high hopes for the new XC10 unit.' And most of the firm's top management team agreed with her. Clearly the market had been maturing for some time now, and was undoubtedly getting more difficult. New product launches from competitors had been eroding market share. Yet competitors' products, at best, simply matched the firm's offerings in all benchmark tests. 'Unless someone comes up with a totally new technology, which is very unlikely, it will be a matter of making marginal improvements in product performance and combining this with well-targeted and coordinated marketing. Fortunately, we are good at both of these. We know this technology, and we know these markets. We are also clear what role the new XC10 should play. It needs to consolidate our market position as the leader in this field, reduce the slide in market share by half, and re-establish our customers' faith in us. Margins, at least in the short term, are less important.'

Selected further reading

Chapman, S.N. (2005) Fundamentals of Production Planning and Control, Pearson, Harlow.

A detailed textbook, intended for those studying the topic in depth.

Goldratt, E.Y. and Cox, J. (1984) The Goal, North River Press, Croton-on-Hudson, NY.

Don't read this if you like good novels, but do read it if you want an enjoyable way of understanding some of

the complexities of scheduling. It particularly applies to the drum, buffer, rope concept described in this chapter

and it also sets the scene for the discussion of OPT in Chapter14.

Kehoe, D.F. and Boughton N.J. (2001) New paradigms in planning and control across manufacturing

supply chains - The utilization of Internet technologies, International Journal of Operations & Production

Management, vol. 21, issue 5/6, 582-593.

Academic, but interesting.

Pinedo, M.L. (2016) Scheduling: Theory, Algorithms, and Systems, 5th edn, Springer, NY.

A very technical, but well-established text.

Vollmann, T.W., Berry, W.L., Whybark, D.C. and Jacobs, F.R. (2010) Manufacturing Planning and Control

Systems for Supply Chain Management: The Definitive Guide for Professionals, McGraw Hill Higher

Education, New York.

The latest version of the 'bible' of manufacturing planning and control.

Slack, Nigel, and Alistair Brandon-Jones. Operations Management, Pearson Education Limited, 2019. ProQuest Ebook Central,

http://ebookcentral.proquest.com/lib/massey/detail.action?docID=5751938.

Created from this year up to the current time (day 58; every working day was numbered throughout the year). Figure10.19 shows the schedule for this period. The job had been accepted on day 18 and had seemed relatively straightforward, although it was

always clear that it would be a long production job. It was also clear that time was always going to be tight. There were 32days in which to finish a job that was estimated to take 30days. 'In hindsight we underestimated how much having three teams working on the production stage of this job at one point or other would increase its complexity. OK, it was not an easy piece of CGI to carry off, but we probably would have been OK if we had organised the CGI stage better. It was also real bad luck that, in our efforts to deliver the fifty-three slash F job on time, we also disrupted the fifty-four slash D job that turned out to be the only other new client we have had this

year' (C.K. Ong). The job had proved difficult from the start. The pre production stage took longer than estimated, mainly because the client's creative team changed just before the start of subtext beginning the work. But it was the actual CGI itself that proved to be the major problem. Not only was the task intrinsically difficult, it was difficult to parcel it up into separate packages that could be coordinated for working on by the two teams allocated to the job. More seriously, it became apparent within two or three days of starting the production work that they would need the help of another studio for some of the effects. Although the other studio was a regular supplier at short notice, this time they were too busy with their own work to help out. Help eventually came from a specialist studio in Hong Kong. 'The subcontracting delay was clearly a problem, but it was only half way through the

production phase that we first realised just how much difficulty the fifty-three slash F job was in. It was at that stage that we devoted all our production resources to finishing it.

Unfortunately, even then, the job was late. The decision eventually to put all three teams on to the fifty-three slash F job was not easy because we knew that it would both disruptother jobs and potentially cause more coordination problems.'

'No way will be doing that again'

'No way will be doing that again', said CK to the core teams when they met to pick over what had gone wrong. 'We are desperately in need of a more professional approach to keeping track of our activities. There is no point in me tellingeveryone how good we are if we then let them down. The problem is that I don't want to encourage a "command and control" culture in the studio. We depend on all staff feeling that they have the freedom to explore seemingly crazy options that may just lead to something real special. We aren't a factory. But we do need to get a grip on our estimating so that

we have a better idea of how long each job really will take. After that each of the core departments can be responsible for their own planning.'

QUESTIONS

1 What went wrong with the fifty-three slash F job and how could the company avoid making the same mistakes again?

2 What would you suggest that subtext does to tighten up its planning and control procedures?