Hi Alistair -
I agree that it may be all for nought, however what got me on to this in the first place was seeing a very strong correlation between outdoor ambient temperature and a consistent ideal boiler set point across all of our shops on a given day (although this has become a little less clear upon further review!). This has lead me to believe that there may potentially be a limited number of variables that, if everything is is kept as constant as possible, I may be able to define, track and use to predict close-to ideal temps instead of reacting to them. At the risk of opening up a completely unrelated discussion on the realities of the labour market and the need to develop systems instead of relying entirely on human engagement, I simply don't think I have that option given the size and nature of our organization. Maybe in time.
Don't get me wrong though - I am a huge advocate for engagement and have witnessed what it can do for coffee and for our organization. In fact, my ultimate goal is to to identify a number of key variables that help us predict an ideal set point in the morning and have our staff check on these first thing - much like a baker would check on the humidity, tap water temperature, etc, before they start making bread in the morning. I think this would do wonders for their level of engagement, with resulting positive impacts on the attention and care going into coffee. In fact, there has already been a very positive impact just from me emailing ideal temps and having staff check the results.
I do see your point though, and perhaps it would be time better spent training enough espresso tasters within the company that know how to relate taste balance and flavour properties to necessary changes in brew temperature. This isn't at all unreasonable now that I consider it, but ultimately I'm a "systems" kind of guy and I'm much more inclined towards the approach of giving all 130+ of our baristas control over the coffee through well defined systems. If in the long term I end up with no meaningful results, then I'll have to rethink my approach! In the mean time this is a fun and very much enlightening experience to go through.
Andy, you make a great point - I hadn't thought of that variable. I've been checking the grinder's ambient temp with a thermocouple meter on every test. I said "more or less" before I started looking at this situation in detail, and there was a degree of ignorance in that statement I must admit!
There is in fact a variation due to the door being open on a cold day or a struggling A/C on a warm day. Looking back at the data I've gathered so far, there's no direct correlation between the fluctuations in ambient temperature I am seeing and the ideal set point, although it certainly may be influencing things. Just to pull data from one batch of coffee, I've got some shots being pulled at the same ideal boiler set point when the ambient temperature is 19C and 22C. What's both puzzling and encouraging is that I'm finding the same set point across all of our locations that, on a given day, all have minor variations in ambient temperature between them.
Mike - you're right, the variables are definitely overwhelming! It's quite fun =)