|
Top-level
 While obviously not very scientific given that it mixes changes in OSM mapping with changes in the real world, I think this gives a fair impression of how much work is involved in keeping POIs up to date. I THOUGHT I had been doing that here, but my recent survey showed I had missed a lot of changes. However, it's possible that this particular street has been exceptionally unstable. I am continuing to survey a couple of streets directly north of here, so I will have a better idea of that soon. 16 comments
 @alan These are critical metrics in order to understand what is involved in maintaining #OpenStreetMap data. In the 2013-9 period we could manage maintenance through 1 hour mapping events before our pub meeting roughly ensuring systematic annual coverage of city centre POIs. It's harder to cover suburban POis in the same way. We could do with an explicit way of capturing "I've seen this POI, and it's still there" without bumping the version with a check_date tag a la @streetcomplete. 2/3 @SK53 Yes I can imagine that would be a problem for scaling up an analysis like this. My sample here is small enough, and I know the area well enough, that I was able to look at the changes individually and see that none of them were due to remote mapping. There were some manual typo fixes but that's inevitable, there is no way to check the name of a shop especially if it's not a brand (and most of these shops are not). @SK53 I think I have more or less reconciled myself to check_date, Everydoor also uses it. I had my doubts originally as it feels a little odd to have this kind of metadata in the tags. But I have found it has significant practical benefits when trying to systematically survey an area. I now realise that in the past I had been repeatedly visually confirming some shops while ignoring others. Even the side of the street I tend to walk down turned out to be more biased than I thought! @SK53 @streetcomplete Yes, I hope that once I get the current data (for an area a bit larger than I have shown here) into a fairly complete position I will be able to better monitor the rate of change in the future. Here the 2020 data was not as complete as I originally thought. One other complication I found was deciding which "amenity=" values to include to capture e.g. bars and restaurants but not drown the analysis in benches and parking spaces. One trend that has struck me (not just on this street but in this general area of the city) is the growth in the number of hairdressers, barbers, and beauty salons mapped since 2020. Again, this might be partly because I previously overlooked some of these (they are often small and easy to miss when they are closed). But I think many are genuinely new. Yesterday I noticed one of them had helpfully confirmed this by displaying "Est. 2023" over the door! This is getting absurd, this evening I found 5 previously unmapped hairdressers / barbers / beauty salons in about 200m of street. This is my life now, walking very slowly with Everydoor open on my mobile, peering up at the shop signs for hairdressers. I'm struggling to remember why I ever thought "map every POI in my barrio" was a good idea. Some hard numbers to back up my growing impression that "there are a lot of hairdressers around here". There are now 180 shops mapped in my area of interest (defined simply as objects with a shop tag in OSM). Of those, 38 are hairdressers, barbers, or beauty salons - about 21%. To save you all rushing off to look at Taginfo: the worldwide equivalent is a little under 8%. (Green markers are hairdressers, grey markers are all other types of shop). @vespucci_editor Yes I think almost all in my sample are sole traders too, and squeezed into physically very small premises in many cases. One thing I've also realised, and that is especially relevant to hairdressers, is that there are far more businesses on small side streets than I used to think. It's partly because I tend to pass by at night when all shops on a side street may be shuttered, whereas shops on main thoroughfares are more likely to be open. @alan @vespucci_editor this was the pattern for Europe 9 years ago. Methodology: http://sk53-osm.blogspot.com/2015/05/retail-outlets-on-openstreetmap.html?m=1 @alan @vespucci_editor Also back in 2018 mapping of hairdressers in the UK 🇬🇧 was, if anything accelerating: http://sk53-osm.blogspot.com/2018/07/can-we-identify-completeness-of.html. Second only to convenience stores then, but fast catching up now (26k vs. 30k). #OpenStreetMap #RetailMapping #hairdressers #AccumulationCurves. @alan Looks like we only have ~50% mapped in the UK 🇬🇧. Rule of thumb from Nottingham figures suggest 1 hairdresser per thousand people, which gives an estimate of roughly 60k (assuming under 5s don't use them), which is close to figures here https://www.nhbf.co.uk/documents/industry-statistics-2023-sheet/1a5-industry-statistics-sheet-2023.pdf @alan the other approach is bavk-of-the-envelope. If we assume 60 million customers getting hair done 4 times a year and taking 20 minutes, we get 80 million person hours or roughly 50,000 FTEs. Obviously very conservative values, but given most salons are sole traders it gives a reasonable approximation. I'm now going to have to get my hair cut so I can interrogate my hairdresser about her customer mix. |
Gregory's Server
@alan That looks a high rate of change, although the Covid period has undoubtedly accelerated turnover of retail outlets. From work in Nottingham and Derby I think I estimated 20% turnover of businesses per year (2013-9) based on name and primary tag alone (would be significantly higher if all tags included). We used to retain some of the old names in tags (very useful for matching old photos, local memory, archived open data etc.), but a newer mapper 'cleaned' them up. 1/3