Seeing the big picture: don’t let the data deceive you


Data are useful. But don’t let them deceive you.

I was thinking about data-driven approaches while having a conversation with Greg Sherwood on Thursday, when I popped into Handford Wines. This shop has an amazing selection of wines. So many really interesting bottles. Sometimes it even seems like too many. If you got the accountants in, they might suggest streamlining the business, focusing on the wines that sell best, and jettisoning some harder-to-sell lines, reducing the amount of capital tied up in the business.

Or think of a major agency business, importing scores of producers into the UK. As well as some gems – real blue chip wineries – they represent lots of smaller, somewhat niche estates. The bean counters might like to strip this business back to just the most profitable wineries, and then run a leaner, more profitable business: the data show that most of the profit is generated by, say, a quarter of the portfolio, so it makes sense to let the rest go.

Or let’s talk media. These days it’s possible to track reader behaviour. I know the sorts of articles and Instagram posts that will get me the most views and interaction. Any consultant would tell me to focus on these, and do less of the geeky, niche stuff.

But in all three cases, the data lead the businesses into danger. For Handford, stripping back the range would erode consumer trust. People shop there, I suspect, because they have such a broad range of interesting bottles, spanning a range of styles – first growths to Ganevat. The diversity engenders trust, and even the bottles that don’t sell well could be playing a role in that range.

Take a successful agency business and cut out the more niche producers, and you will see that even if people weren’t buying a lot of these wines, their presence in the portfolio likely served a purpose. Would the new business be leaner and more efficient? Perhaps. But it might also lose customers for whom the breadth of the portfolio mattered, even if they weren’t buying a lot of the fringe wines.

And as for media, it’s a similar story. If I focused more on what gets the big viewer numbers, would my work be better? It would be less interesting and less authentic, and it would quickly resemble other media outlets all focusing on the middle of the normal distribution, rather than working the tails. Data can drive people into the same competitive space. Diversity suffers.

[Picture: time for reflection, Vancouver Island]

Reminder: this blog is irregularly updated, but there are daily updates on the main wineanorak site.

wine journalist and flavour obsessive

Leave a Reply

Back To Top