I want to admit something publicly… something that I have been guilty of in the past.
Ok, here goes… I am guilty of… giving a ballpark estimate for a large project with a 20% variance in cost.
“Why’s that bad” you ask?
“Surely that’s great from the client’s perspective”
“it sounds like you know what you’re talking about and can narrow down any uncertainties before you kick off”
Well, yes and no, and I suspect that any PM, producer, or anyone involved in estimation has been guilty of this sin at one time or another in their careers to make it sound like they’re confident in their estimates.
The fact is, for any project, you can’t know everything at the start. Especially when you’re dealing with creative work. Technologies change, Platforms change, People’s minds change, that’s just the nature of what we do and how life goes on, and it would be crazy if we didn’t change if it meant doing something better. For the project in question
- I didn’t know for sure who the team working on it was going to be
- The scope, whilst detailed in parts, was loose at best in the unexplored areas of the brief
- The technical platform was not well-defined, and there were a large number of other uncertainties and assumptions that my fellow estimators had recognised.
- There were several factors that could not be accounted for (sickness, late feedback, unexpected bugs)
The reality of an estimate is to truly gauge the cost of doing something. Nobody on the client side, the agency side, or the production house side wants to be exposed to the risk of a project failing on time, cost or quality, which it will be if you underestimate the task in hand – everyone’s disappointed then. But then everyone wants “Value for money”.
But we have to kickstart a project into inception somewhere, and that somewhere is usually a place after concepting called budget signoff.
Now for my younger years, I was always a believer that a good estimate was one with a small variance.
“between 24-26k”
“this is coming in at 220-240k”
Several coworkers, all with more commercial experience than myself, albeit not on the estimating front, all told me that a good estimate removed the margin of error between the initial figure and the final cost, that made sense. Then later on, some other experienced folk – again with little experience of making estimates – told me that the top end of the ballpark should never be exceeded in the final cost, again, this made sense, but was sometimes tricky if you couldn’t ascertain every factor. Finally, at the same time as the last piece of advice, I was informed that this top-end figure could not be significantly higher than the low end of the spectrum. Anything else would be commercial suicide when it came to selling in the costs because it would look like you didn’t know what you’re talking about.
On the face of it, that seems to make sense, but when you delve deeper, it’s simply not true, in fact I would go so far as to say that it’s an outright lie in some cases. The variance of a good ballpark estimate should reflect the degree of unknown factors and use the experience of the team estimating and completing the work to remove as much error as possible. An experienced team will know that every project where serious collaborative endeavour goes on will throw out lots of leftfield problems and obstacles that simple cannot be rationally accounted for at the start.
In fact a good ballpark estimate should be wide because it highlights the risks and unknowns faced in a project. It’s only after you start to test through experimentation, prototyping and iterations that you dig out the real facts, get rid of assumptions and see how something behaves in captivity and then the wild.
What’s crucial in estimating is planning the most important areas of a project to be able to say with a high degree of certainty (some say 90% sure) that those things will be done. Passion, experience and good collaboration should then ensure that you get as much in the time as possible.
Suburban Superbia
Disclaimer: Arcade Fire are my favourite bestest best band ever. So I’m likely to be slightly biased about this!
Have you had a go at this yet? Go on give it a go… I really enjoyed it and found the experience quite moving. Given that I had grown up in the relatively protected ‘burbs of Newcastle, the overlaid streetview images fitted very nicely with the sentiment of the song and the album. Not sure how well it would work if you were brought up in Streatham though, or at the other end of the scale, a luxury pile in Monaco.
The whole experience has a feel of a ‘final project at university’ or something (in a very good way), where they’ve just decided to sod the taboo of having multiple pop up windows running at the same time. Ok so it struggled with runtime a bit and made my laptop slow down to a snails pace if I tried to do anything else – but it’s an immersive experience and one that warrants a bit of artistic license.
The things I liked about it;
1. The use of multiple windows
2. The integration of streetview and google maps
3. The ‘unexpected’ 3d animation at the end
4. The text animation/effects on the ‘Postcard’ section
5. The fact that they’re using up loads of your processor power but just don’t care that much – even putting a processor disclaimer warning for you to shut down other programmes. ‘Ok so it might be inconvenient to watch this but hey – who cares!’
It’s also probably one of the most high profile HTML 5 experinces out there at the moment, passing the test as an interactive experience, although I think I’ll reserve judgement on the technology for a while yet! Would love to see more HTML5 experiences like this – send them to me if you have any…