Project Management, G4S and the Olympics

There's been a lot of speculation around concerning the failure of G4S to recruit and train enough security staff in time for the London 2012 Olympics.

Buckles buckles, and poor Tommy Atkins picks up the pieces

According to G4S' CEO, Nick Buckles, they will end up losing between 30 and 50 million on this contract. G4S are going to pay for the extra soldiers who will be forced to give up their recovery time to fill the gap. While I would bet much on Buckles still being in a job in a month's time, surely the most serious outcome is that some of the troops in Afghanistan will be required to stay out there for longer to cover for troops that have been drafted in for the games.
Some commentators have expressed surprise that Buckles was only made aware that they weren't going to make it 9 days before the story broke. I'm afraid that, with over 20 years experience of delivering complex projects (and this was not a particularly complex project), I was not surprised at all.

2000 to 10000 in 9 months, a slow-motion train wreck

Now I haven't worked for G4S, but I can make a pretty educated guess as to what happened: G4S were originally contracted to supply 2000 security staff. This was upped to 10000 staff following a review in November 2011, leaving them with less than 9 months to find an extra 8000 staff. Just to be clear, this represents a 400% increase. Now I'm willing to bet that the Project Manager in charge, and the recruitment team knew that this couldn't be done as soon as they were told about the change. The project manager may have even told people higher up the food chain that it couldn't be done. They would then have come back to him/her and said that it had to be done. They may have even offered extra money and people to do it. The PM probably took on some extra staff to try and avert the impending train wreck.

 

Classic mistakes

[As an aside, in software development circles, this is known as a classic mistake, adding people to a late project. There's a whole book on it, The mythical man month, which I recommend to anyone who's interested in why software projects go wrong. ]
What I'm guessing happened now, is that the PM and the team struggled on, knowing that they were not going to make it, but equally knowing that no-one in senior management would be interested in hearing the bad news until it was obvious to everyone. There's a name for this in software development circles too: The Death March Project.
I think (though I don't know) that the Olympics security debacle was a symptom of a corporate culture at G4S that is all too common in UK business: The people at the top think that they can get things done by demanding it, they don't like to hear bad news, and moreover, dismiss those who say that a thing can't be done as whingers, even while they are relying those same people to do the thing that can't be done.

"Star Trek is not like real life" shock

If there's one thing that I've learned in 25 years, it's that if you're relying on a person to do a thing, and they tell you it will take three months, and you tell them that it has to be done in two, it will take three months - if you're lucky. Real life isn't like Star Trek.
Another thing that is common to these sorts of organisations is a Mickawberish belief that something will turn up, and if you are lucky, you will get away with it. Again, 25 years of experience tells me that you are never, ever lucky. In fact, if something only takes the amount of time and effort you thought it would, and not more, then you can count yourself as lucky.
I have two prescriptions: one for the CEO, and another for the Project Manager:

 

A better way for the chief

For the CEO - This, like many aspects of leadership, is simple. Simple but not easy. What Buckles' successor has to do is to change the culture of the organisation so that open communication of bad news is encouraged and acted upon. And in case you think I'm being hopelessly idealistic - I have actually done this, for a team of 30 people, and it does actually work. The way you start is, the next time that someone brings you bad news, just take the time to actually listen to their concerns, engage with them in a way that shows that you understand and negotiate some changes that will make a difference. Then you keep doing that every day. Within 6 months, you will really see the change.
Imagine if someone had come to Buckles in November 2011 and told him that there was no way that they could go from 2000 to 10000 in 9 months and instead of just telling them to make it happen, listened to their concerns and responded to them. They might have figured out that the best that they could do was 5000. They could have gone back to LOCOG with that figure, LOCOG would still have had to get the troops in but they would have had 9 months to plan it and it would certainly have kept G4S out of the headlines.

 

A better way for the Project Manager

Too many PMs, when faced with a project, and a management like this, just give up. They stop planning and they stop reporting. This is the wrong thing to do. When the project goes wrong it is then the PM's fault, fair and square. What you should do is keep the plan up to date and keep reporting openly and honestly. Here at Stonivale we have a simple way of reporting progress, using red, amber and green statuses that are clearly defined and consistently applied:

• GREEN - All tasks are on time and the project is on time
• AMBER - One or more tasks is forecast to finish late but the project is still forecast to finish on time
• RED - One or more tasks is late and the project is forecast to finish late.

We insist on weekly reporting, not monthly as is common (and was the case for the G4S project). Senior managers tend to sit up and take notice if they are getting a report week after week that is telling them their project is late.

Does any of the above look familiar? Need help with sorting your projects and programmes out? Contact Stonivale for a chat. We might be able to help.

Not just consultancy

I was musing to my wife the other day - trying to explain how I conceptualise my skills - what makes what I do special, and a bit different from other people who might have the same role. I explained that what I like to do is to chat with people, find out about the problems they are having, and then try to work out ways that they can solve the problems. Sometimes the solution might involve technology, sometimes training, sometimes changing processes, sometimes changing the way that they think about the problem.
She said to me, "That's just consultancy!"
Instinctively I shrink from describing what I do as consultancy, and, over the past few days I've been trying to figure out why.

Three problems I have with consultants

Consultants tell you what you want to hear 

 

I think that a lot of businesses bring in consultants to justify what the MD or CEO wants to do anyway. They come in, listen to the Boss, write it all down, add some charts, take the money and away we go.

Your problem is always solved by what the consultant has to sell

The second problem I have with consultants is that a lot of them come in to sell you their own solutions off the shelf. Whatever your problem is, the things that they just happen to sell are the solution. So the process improvement consultants will sell you CMM, or EFQM, or Scrum, the software vendors will sell you a new CRM or an ERP system (if you're really rich and unlucky) and the system integrators will sell you lots of different systems which will cost a fortune to glue together.

Consultants charge too much, with no guarantees

The third problem is that they're so expensive. A one-man band will cost you £350-£500 a day, a consultant from a small to medium-sized consultancy will cost £600-800 and if you go to the big beasts then you can expect to pay anything up to £1500 a day. And whats worst about this is the risk is all yours: once they have delivered the report, or the new project management templates, or the shiny new website they are out of the door and it's up to you to turn all that spend into something that actually benefits your business.

Another way

I'd like to propose another way. My way has two parts to it: a new way of working and a new way of getting paid.

A new way of working

It's very simple really: This is how it works:

1. I work with you to figure out what your real problems are;
2. I work with you to figure out what is the simplest and cheapest thing that can be done to solve that real problem;
3. I work with you until the problem is solved or until you don't need me any more.

 A new way of getting paid

Instead of doing half the solution, taking the money and leaving you to do the rest, the consultant should have an up front conversation about how much you expect to gain from solving the problem. Then we can talk about how much and when the consultant should be paid.

How much

There's very little point in paying £5k for a new website if it doesn't bring in any new business. As an aside, this conversation (known in Prince2 speak as the Business Case) is a key part of step one of my 3 step plan, above. Maybe you would want to pay £5,000 if you were going to get £50,000 of business from it. I don't know, but it's a conversation we should have

When

I am of the firm opinion that any decent consultant should have enough faith in his own expertise to be willing to be paid by results instead of getting paid whether or not he (or she) makes a difference. Of course there are limits - I wouldn't expect a consultant to shoulder the risk of implementing a new data centre (and I hope neither would you) but it should be part of the conversation.

Do you think this would work for you? If so contact me and we can have a chat.

Lifehacking - look back in exhaustion

I've just finished my short film - Lifehacking. It was a really good experience - hard but good. I've been making short films and cutting together bits of footage taken on holiday and on events for a couple of years but this is a different order of magnitude. Doing a whole film from beginning to end and making it as good as you can make it is a big step up. I've learned so much about sound (I need to get much better - or get someone else), setting up the camera (I feel really comfortable with all the manual modes now and I'm pretty sure I know how to get the best out of the camera in any given situation), and loads about colour grading and colour correction.
The way I put the film together was this:

• First I did a rough cut of each scene as a separate project in Vegas. For each one I would leave it about a week and then go back and see where I could cut more out and where there were transitions that seemed uncomfortable.
• I then went through and did the colour correction and grading. This took many more passes than it would now as I had to learn the effect I was trying for as well as how to do it. The Red Giant Colorista tutorials on Vimeo were really useful for this as Stu Maschwitz takes you through the look he is trying to achieve in real detail. Since the controls are a bit different in Vegas then I had to figure out how to get the same effect. Unfortunately Vegas' secondary colour correction masking tool doesn't seem to work quite as well as Colorista for selecting skin tones so I struggled a bit until I found this Vegas tutorial that taught me all about compositing. Then I was flying! I'm really tired of the modern style of colour grading where everything is a sludgy blue/green/grey and skin is a putty colour (See BBC Sherlock). I went for an old fashioned Technicolor look. I'm especially pleased with the night-time scenes on the sofa as I was able to darken the background and relight the face. Unless you know, you can't really tell.
• I then assembled a rough cut of the entire film and handed it off to my brother Matt to add the music. I then left it for two weeks so that I could come at it fresh.
• Two weeks later I rewatched the whole thing (it was getting really tedious by now and hard to stop my mind wandering) and also listened to it right through without watching. I made a set of notes which I have reproduced below. I then went right through the list, fixing everything I could. The final edit took a full day, including re-recording (ADR) the voice for the entire first scene.

My gear

• Canon 550D with 28-55mm kit lens and 50mm f1.8 - Brilliant. Like Crazy was made with a 7D which is basically this camera in a metal body. Now the 600D is out you can get this much cheaper: For £500 you get a camera that you can make a pro film on. The 50mm lens cost £90 new! Great for shooting indoor scenes if you live within its limitations (No IS - Can hunt for focus in low light - though I now use manual focus with the preview screen zoomed in)
• Zoom H1 Digital recorder. Lovely and small. Great with the Lavalliere as you can slip it in your pocket.
• Audiotechnica stereo mic. Nice neutral sound quality although the output level is a bit low.
• Lavalliere (tie clip) mike. Really handy for outdoor shots where you want to reduce background noise.
• Asus X53E Core i7 laptop - Great! Renders 720p HD straight out of the camera at a little better than real time and a really complicated scene with colour correction and compositing layers at about half speed. I no longer have to bother with transcoding. The only problem is the vertical viewing angle on the screen is very small (the side to side viewing angle is fine, strangely so two people can easily watch the same thing, as long as their heads are at the same height) so you have to make sure you're looking directly at it to make sure you're seeing the colours accurately. When I can afford it I will buy a decent HD monitor.
• Sony Vegas Platinum - I'm a real fan of this. Once you're used to the interface it's really fast to use and you can do anything on it that you can do with high-end pro packages. The only thing that I wish it had from the Pro version (which is 10 times the price) is scopes and the ability to split the preview screen to make it easier to grade across scenes.

I'm now on to my next project, an adaptation of a short story by Checkhov. I will definitely use actors for this one, and hopefully my brother will come in to do the sound.

These are the notes that I took on the rough cut of Lifehacking:

Notes on lifehacking
====================
Sound
=====
Scene 1
-------
- Phone ring too loud - Done
- Less bass on the vocals
- Replace "Bank Account"
- Check "Bank Account"
- Clunk after "Still a bit hot?"
- Clunk after "mystified"
- Vocals are at about -18dB
- Background at about -48dB

Scene 4
-------
- Equalise vocals between 2 takes
Scene 5
-------
- Phone ring too loud. Needs room reverb?
- Reduce top end on vocals
Scene 6
-------
- Phone ring - scene 5
- Equalise vocals between scene 5 and 6
- A couple of vocal glitches. Replace?
Cut scene
---------
-Replace sound with wildtrack without wind-noise
Scene 7
-------
- Vocals a little harsh on top/mid - equalize with scene 6
- Vocal should shift to centre for final headshot

Visuals
=======
- Check beginning credits
- Replace font on Voidstar
- Narrower, smaller sans serif font
Scene 1
-------
-Improve secondary colour correction when seated to darken the fence in the background and pop the face
-Close-up needs colour correcting so it matches rest of scene
Scene 3
-------
- Can cut to seated shot earlier
Scene 4
-------
- Redo colour correction on front shot
Scene 5
-------
- Redo colour correction to darken background while relighting face
- Skintone is good! Replicate in scene 4
Scene 6
-------
- Match skintone with scene 5 and between medium and close-ups

- End credits

Lessons learned from making Lifehacking

Yesterday Matt sent me over a new cut of my short film, Lifehacking, with some music that he had added. It sounds great!
I'm now going to leave it a week and then do the final edit.
In the meantime I have been thinking about the things I've learned from making the film. Here they are:

The Good

The camera (Canon 550D) is great and the 50mm, f1.8 lens is a must have at only £90. It means you can shoot in pretty low light conditions and still get a decent result (see the last scene for an example of this)
Towards the end I got into using the magnify function to check the focus - vital when you've got the aperture wide open.
Sony Vegas is a fantastic editor and my new Asus core i7 laptop just flies. I no longer even bother to transcode the video off the camera, it's fast enough without reformatting - as long as you remember to keep the power plugged in.
I found the tutorial videos on using Red Giant Colorista really useful in learning how to use the colour correction tools in Vegas. Though I found myself wishing I had the power mask feature most of the time!
Using the Zoom HA-1 with a Lavalliere worked very well for recording the outdoor scenes. I found it easy enough to sync the on-camera sound with the recorded sound.
Just going out and shooting, even in quite crowded areas, was a lot easier than I thought it would be. People naturally seem to stay out of your way. I asked before filming the cafe scene but they were more than happy to let me go ahead.


The Bad

The sound for the inside scenes is a bit reverby. I put the mic roughly where the camera was. I should have close miked all the scenes and then recorded some wild-track.
If you have an actor in his 40s, don't write a script that requires that he looks down a lot. Very unflattering.
Some of the inside scenes are not quite in focus. I ended up shooting at f2 or 2.2, rather than f1.8 but it's still very hard when you're trying to guess where the actor's face will be. This would have been so much easier if I hadn't been both the actor and the cameraman.
In one of the outside shots there is a wooden fence going straight through my head. I wish I had noticed this at the time of shooting.
It's best if your lead actor has either very long hair or very short hair, especially if you are going to shoot over an extended period of time. My hair was in between, and when I had it cut, I had to wait about 6 weeks for it to grow back to the right kind of length

Things I'd do differently next time

Get actors instead of doing it all myself. I did it this way out of frustration because I'd been speaking to an actor about a collaboration over the summer and for one reason or another it hadn't happened. Next time I will definitely stay behind the camera
I need to be much better at sound recording. For my next film I'm hoping that Matt will be able to help out with the sound.
Because the script allowed me to read off the laptop screen, I didn't learn all the lines. I think I got away with it, but I wouldn't do that again.
I would also like to get someone to do continuity. It was really hard to remember what position I was in and where everything was between shots. There were a couple of scenes where in one shot I had my arm up and the other I had my arm down, which really cut down the options when it came to the final edit.

First cut of Lifehacking in the bag!

I finally got my replacement laptop on Friday, after a false start on Tuesday when I got to the door of Comet at exactly 6 o'clock and it was already locked. The new one is an Asus X53E, same as the old one, but a Core i7 rather than an i5 as the i5 is now £900! (The trackpad button failed on the old one after less than a month - bizarre. Hope it's a one-off).
I managed to restore all my data, reinstalled Vegas (the on-line registration worked fine, so the rumours that Sony allow you to install Vegas 5 times must be true). The first time I re-rendered scene one I was a bit disappointed - it took longer than the i5. Turns out it was AVG doing its initial virus scan that was slowing it down. The next one I tried was very fast - A fairly complex scene with multiple levels of colour correction, and a scene that I had reversed, takes about 130% of the running time to render. How long would it have taken on my old pc? 10 hours?
Anyeay, Lifehacking consists of 7 scenes that I edited as separate Vegas projects so yesterday I put the finishing touches in to each scene (mainly adding the sound of the mobile phone ring tone) and produced a first cut of the whole thing. It's longer than I thought, nearly 15 minutes. Watching it through (which I promised myself I wouldn't do) it seems very slow to me, but then I've seen every second of it dozens of times.
I'm going to give a copy of it to my brother, Matt, to see if he can come up with any music for the intervals between the scenes and then I'm going to leave it for at least 2 weeks with the hope that coming to it with fresh eyes will help me spot the errors and have a clearer view of the overall flow.
I've learned masses from doing this, which I hope to make the subject of a future post.

Finished with the Premiership

I have been unfortunate enough to see Everton play twice in the last two weeks. I thought last week was bad enough but this week has convinced me that something must be done. And for me, that something is to stop watching Premiership football.
But why, I hear you cry? (Oh, suit yourselves). Well, I'll tell you: Last week Everton played against Man City plus the estimable Howard Webb. Webb seemed to think that the mere fact of an Everton player successfully challenging for the ball was evidence of a foul on his part. Man City are obviously so superior in every way that it is inconceivable to Mr Webb that any player from the lower ranks can dispossess one of the Etihad anointed without stepping over the line between fair play and foul. This week was the icing on the cake, with the cherry on top. Jack Rodwell sees the ball, Rodwell gets the ball, Suarez falls over Rodwell and referee Martin Atkinson sends Rodwell off! Apparently in today's Premier league it is a red card offence to get the ball if your opponent wants it at the same time (if your opponent plays for one of the top 5 clubs, that is). I presume that players from the other 15 clubs are now supposed to stand around like particularly expensive bollards, while the Galacticos amaze us with their artistry. It makes football about as meaningful as a competition as a Nike ad.
This might all seem like sour grapes but it's more than that. I freely admit that both Man City and Liverpool were better than Everton (although not that much better - it may be a partisan view but of the expensively assembled talent available to Liverpool only Suarez and Kuyt would walk into the Everton side, at least on Saturday's showing. I didn't even know Downing was on the pitch until he got substituted). It would be a funny old world if they weren't - Liverpool have spent over £100m since Dalglish returned while Moyes has spent a big fat zero. No, my central point is that if you interpret the rules of the game in a way that only the most prodigiously gifted are allowed to succeed then you snuff out any possibility of an upset, of the underdog succeeding through hard work, or better organisation, tactics or teamwork. You might as well play top trumps before the start of the match: Liverpool - £114m, Everton nowt. Liverpool win!
Although Webb and Atkinson were culpable in these two cases, they're not the ones to blame. The real culprits are the Premiership and most particularly Sky. They've turned the game I and millions of others love into a product. And d'you know? It's a product I'm no longer prepared to buy.
So farewell beloved Everton, I wish you all the best. There will always be a part of my heart that belongs to you and I'm sure I'll sneak a look at the results every now and then. From now on though, the only team I'll be watching is the mighty Carlisle United of Division One. Come on you Foxes!

Lessons learned from making Buses with Bowties

Buses with Bowties was the first proper go I had at making a film using my new sound recording equipment. I learned a few lessons along the way that I want to share here:

• Make a list of your kit. Print out the list, laminated and put it in the kit bag on the top. Before you go out on location, check the kit against the list. That way you avoid forgetting something really important (like your sound recorder!)
• The lavalliere microphone doesn't work very well plugged straight in to the camera. I think this is because of the lack of ambient noise from the mic. The camera turns its auto-gain right up and the speech is distorted because it can't react fast enough when you start talking.
• Get enough wild-track at each location to cover the entire shooting duration. Otherwise you will spend ages piecing together bits of sound to try and make a decent wildtrack in the edit
• Record all the bits that are supposed to sound the same with the same mic. So if you are doing something to camera with a lavalliere in one shot, use the lavalliere for all pieces to camera (unless you have an obvious mic in shot, as I did for one piece.
• The auto-focus in live view mode is rubbish for close-ups. It's better to switch to a picture mode, focus and then switch back
• Have a big pause at the start of each scene (at least 2 seconds) and hold for a similar time at the end of the scene. Makes editing a lot easier. 
• Consider getting a clapper board and saying "Scene 3, Take 4" at the start of each scene. This will make matching the audio and video a lot easier. If the clapper board is in view at the start of the shot, it will show up in the preview mode when you edit
•  Before each shot check the microphone is on, then the sound recorder, then hit record on the camera.
• At the end of each shot turn off the microphone! Sounds like a pain but otherwise you will forget to do it and will run the battery down on the mic (as I am typing this at work I am wondering whether I turned my lavalliere off!)

You can see the results here:

Buses with Bowties from Steve Green on Vimeo.