Thursday 6 March 2014

What closes a brewery? It's probably the quality of its beer

After my two posts on the thorny subject of the reasons breweries close, the subject has been taxing my mind. I have come to realise breweries stop operating for a range of reasons, from changes in owners’ personal circumstances right through to the premises they occupy being sold. Undoubtedly, the breweries that are most interesting to me are those that closed because they were, according to their owners, unprofitable. But why were they not making money? Are there too many new breweries being set up, the result being there are not enough ale-loving consumers to sustain them all? Or were the closed breweries’ beers undrinkable to the point that no discerning pub-goer would touch them with a barge pole?

I resolved to answer this ‘market’ vs. ‘quality’ question. It would be impractical for me to measure if there are too many breweries in the UK for the market to sustain them all; this would require time, resources and data sets which I have no chance of acquiring any time soon. So, I decided to attack the ‘quality’ issue, to determine whether the breweries that closed citing alleged unprofitability did so because they produced beer that was simply not good enough.

On a night out, when I decide to log and rate the beer I am drinking, my girlfriend finds it amusing. I am, there is no doubt, a bit of a ‘beer spotter.’ I log my beers on Untapped website, but the dominant player in the ‘beer logging’ world is ratebeer.com. On it, users can give beers a score on a scale of one to five, and the beers’ average score is then calculated. As more drinkers rate the beers, these average scores are updated accordingly. It is these scores that I have used to measure the quality of breweries’ beers, give each company what I am calling an ‘overall beer rating’, then determine whether beer quality is a factor in their survival.

The fairest way to calculate the overall beer ratings was to use weighted averages (see end note for a description of my calculations – I won’t explain it here, that would be dull). I have calculated these for all the breweries that I know have closed in the last three years because they were allegedly unprofitable, as well as those that have been taken over by new owners. For comparison, I have also determined the overall beer rating for an similar number of producers that have opened in the last three years and are still operating. These were mostly chosen at random; however, I have also calculated the rating of three breweries which have national recognition as producers of excellent beer.

The results are shown in the graph below. Breweries that have closed citing unprofitability are in crimson, those that have been taken over are in orange, while breweries that are still operating are in blue. I have anonymised the breweries to avoid understandable embarrassment (and perhaps some self-congratulatory back-slapping).
Closed: Crimson         Taken Over: Orange        Still Open: Blue
While it is important not to overstate the value of these results – after all I have only worked out forty-one breweries’ overall beer ratings – they clearly suggest that there is a correlation between the quality of beer a brewery produces and its survival. If we consider the breweries in the sample that have the twenty-one lowest overall beer ratings (2.653 to 2.941), only five of these are still open. Nine that closed down citing unprofitability were in this end of the graph, while seven that were taken over are here too. Consequently, it can be very easily suggested that these breweries ceased trading because their beer was not of a high enough or consistent quality to appeal to drinkers.

Conversely, if we look at the twenty breweries with the highest overall beer ratings, the situation is reversed (2.941 to 3.550). Only three of the breweries that closed down citing unprofitability are in this end of the graph, as is one brewery which was taken over. Because they are here, rather than at the other end of the graph, it is plausible to suggest factors aside from their beer quality caused their downfall. For example, they might have expanded their operations too quickly and were subject to unsustainable debt repayments. Notably, sixteen of the breweries in this end of the graphic are still open, with twelve having ratings of greater than 3.00. This heavily implies that those breweries that are producing more impressive and appealing beers are more likely to stay open.

While these findings are very tentative, I feel that this research has started to reveal something important. It strongly suggests that the quality of a brewery’s beer is a major determinant of its survival. Of course, in some cases other factors will have played a role in a brewery's closure, but, simply put, the market seems to be working. Those companies that fail to provide drinkers with quality products are seemingly doomed to fail, while others that excel, innovate and possibly take risks, are more likely to be successful. In future posts I will consider these points in more detail, as well as reassess them.

All comments and ideas are, as usual, very welcome.

End note on weighted averages

The weighted averages (breweries’ overall beer rating) were calculated by multiplying each beer’s average rating by the number of ratings it had received, adding all the beer’s scores together, and then dividing this by the total number of ratings all the beers had been given. The table above is an example. The benefit of this approach was that if the majority of a brewery’s beer received a low rating, but they had one short-lived product that was rated as exceptional, both these things would be reflected in the overall score the brewery received. A short example will explain this. If a brewery produced in a five year period two beers which had an average rating of 2.33 and 4.27 respectively, a basic calculation of the brewery’s overall beer rating would be as follows:

(2.33 + 4.27)/2 = 3.30

But consider if the first beer was rated 150 times, and therefore was theoretically produced for a long period, while the second beer had only received five ratings, and thus had a short production run. The consequence of this would be that the overall beer rating of 3.30 would not be an accurate reflection of the quality of the majority of beer the brewery turned out in its existence. The weighted average solves this problem, as it takes into account the extent to which each beer was produced. The calculation is as follows:

((2.33 x 150) + (4.27 x 5))/(150+5) =
(349.50+21.35)/155=
370.85/155 = 2.39

Thus, the weighted overall beer rating for the brewery is 2.39, which is much closer to the score the majority of its beer received, but does take into consideration the one time the company brewed a superlative beer.