If you have any other gems like that one, I’d love to hear them!
Hi Nick. A few thoughts:
- I imported my whole cellar list and the dialogue box showed that all bottles had been successfully imported. However in my “Collection” section, there are only 113 bottles.
- I use offsite storage and store in boxes to maximise space. It doesn’t look like I can store multiple types of wine in a single slot (e.g. 6 different wines in a 6-pack box in a slot).
- I echo the comment someone made before re multiple locations (e.g lockers) with different layouts. Some of them are u-shaped and others are just cubes.
- A reason I put a lot of effort into maintaining the metadata for my wine list is so that I can use a visualisation tool (Power BI in my case) to traverse my collection easily. Your “Analytics” page goes part of the way there and has visual graphs of some of the key attributes like Country, Region or Wine Type. I would find it useful to have a page which has all of the useful attributes in charts which enables me to progressively filter down. For example, click on “2021”, then on “Barolo” then on “Castiglione Falletto” to filter down to the wines I have that meet that criteria.
Cheers, Will
I just spent some time looking through the app and I like the way to organized. Curious how it will stack up to Celler Tracker.
I do have one suggestion, which is to have a way to code bottles that are partially consumed, say using a Coravin. A few times I’ve pulled a bottle out of the cellar to discover that I’d Coravined it months ago and had forgotten.
Apologies if this is already an function and I somehow missed it.
Good luck with the app!
@Noah_C-Thanks for the feeback. That’s an interesting idea! I’ll add it to the roadmap.
Hey Noah, thanks so much for taking the time to explore the app and share your thoughts!
You’ll be glad to know that’s been built — and we just shipped it today. Each individual bottle now has a “Partial Access” log where you can record every time you use a Coravin (or any other device — Durand, recork, whatever you use). It captures the date, how much you poured, and any notes. The bottle stays in your cellar as “In Cellar” but shows a small access count so you always know it’s been accessed.
You can find it in the Bottles tab on any wine detail — click the bottle to select it and the Partial Access button is right there.
Really appreciate the feedback, and great minds think alike — that exact scenario (pulling a bottle and realising you’d forgotten you Coravined it) was exactly what we were thinking about when we built it.
More good stuff on the way!
1 Like
Hi Will, really appreciate the detailed feedback — this is exactly the kind of input that helps us build something genuinely great.
On the import count — that’s a bug we’ve now fixed. The import was silently failing partway through on larger collections and still showing success. It should now correctly import all bottles and surface a clear error with a rollback option if anything goes wrong. Worth trying a fresh import and let me know if the count still doesn’t match.
The offsite storage / box scenario is a great one. The current model assumes each slot holds one wine, which clearly doesn’t work for box storage. Multiple wines per slot is on the roadmap — I want to make sure we design it properly rather than bolt it on, so it’s a near-term priority rather than immediate, but it’s firmly on the list.
Multiple locations with different layouts (u-shaped, cube lockers etc.) — also on the roadmap. You can already create multiple racks with different row/column configurations, but the layout shapes you’re describing need more work. Noted.
The progressive drill-down analytics is a really compelling idea. What you’re describing — click a vintage, then an appellation, then a producer and have the charts update — is essentially a faceted explorer, and it’s much more powerful than static charts. That’s going on the roadmap as a proper Analytics v2 feature. Really well articulated.
Thanks again Will — you’re clearly a power user and this kind of input is gold.
@nicksumerfelt,
I’m impressed at how many changes you’ve implemented on the fly!
Here’s an issue I just ran across. I clicked on “past peak”, and it brought up the list of wines. However, when I clicked on the wine, this error message resulted:
Cheers,
Warren
Ha! Thanks! We write lots of comments in the code, makes patching much easier.
I think we found the bug, have a look when you get a chance and please let me know if you run into any other issues.
Thank you for the feedback!
I too am impressed at how quickly you can make features..
question on the approach - are there plans to implement a community at all? i think one of the most valuable feature of cellar tracker is the database of users/wines/notes etc..
@Mark_Y We currently have a Community feature but a lot of what we want it to hasn’t been implemented yet. I’ll be sure to follow up when Community is closer to complete.
Thank you,
Nick
1 Like
All good now. I’m going to spend some time on the website within the next few days. I’ll flag any issues and add suggestions if indicated.
Thanks, @Warren_Taranow. Looking forward hearing your thoughts.
Hi Nick - Good looking site so far!
I signed up for a Beta account and tried to import a .CSV generated from my Cellar Tracker account that has 3500+ bottles. The import keeps failing completely (i.e. no bottles entered). While I can’t see exactly why, I suspect it has something to do with my exported Location and Bin info.
My cellar has a combination of grid storage and bins. Because of CT’s limited location information (only Location and Bin), all of my wine is designated as “Bin” storage. Location designates the rack, and Bin designates either the actual bin or a column on the grid. (Even where I have grid racking, I have divided each column on the grid into a Bin, figuring I’m not compulsive enough to keep each column properly ordered anyway.)
So my problem is that each of my “Bins” has anywhere between 12 and 25 mixed bottles. When I try to edit the Cellar Map, it does not allow me to specify a Bin size of more than 8 bottles. And although I’m not able to test it, is it true from your comments above that I would not be able to put more than 1 unique wine in a Bin? This seems very limiting.
Maybe I’m misunderstanding how the Cellar Map is supposed to work. Let me know if I can provide more details.
EDIT: Edited to say that I just tried importing again and skipping the Location mapping altogether. The import still failed completely - no bottles entered.
@s_kukshtel_I’m very sorry about that. That was a huge bug! We made a schema change that blew up our database rules. The issue has been corrected, you shouldn’t have any further trouble importing your inventory.
Thank you for the heads up on that!
Sorry for the confusion on both counts.
The 8-bottle bin size capimplemnted in testing and not updated— it’s removed in the update going out now, so your 12–25 bottle bins will work correctly. And yes, multiple different wines absolutely can share the same bin. A bin is a named compartment, not a single-wine slot. Each wine is tracked individually within it, and tapping a bin on the map shows you exactly which wines are inside.
Your CT Location and Bin data won’t cause any import problems. Location maps to the rack name and each unique Bin value becomes a separate slot within that rack — whether it’s an actual diamond bin or a grid column you’re treating as a loose pile, it works exactly as you’d expect.
I really appreciate your patience.
Hi Nick,
Thanks for creating this. I have a few user interface recommendations:
- In the Collections tab, in the table format, it would be nice if I could sort by columns by clicking the title of the column instead of using the sort by box. In the grid format, there seems to be a lot of white space, where each grid is a bit too big.
- For the region, it often lists Unknown. For some cases, it seems like the system knows where the bottle is? I have a bottle of Left Foot Charley Blaufränkisch Old Mission Peninsula, which it identifies as Old Mission Peninsula, USA, but still labels Unknown. There are some other cases where there is no location to identify. I think it would be best if it labels it as fine as it can identify. For example, if it is a Vin de France, France would be a better location than unknown.
- The aging curve in Analytics is not showing me any bottles, and the same for drink now and peaking soon. Do I have to do this manually?
- For the vintage charts, the color for in your cellar is really close to the other colors. The app uses color for score and text for whether its in its drinking window. Would it be possible to make this more visual? For example, maybe color for score and shade for drinking window? If not, I still find whether its in the drinking window to be better for color.
Please let me know if I can help in any way. I’ll send more comments as I use the site more.
Best,
Daniel
Hi Daniel,
Thank you — this is exactly the kind of feedback that makes the product better. Good news: we’ve addressed all four points.
-
Column sorting is now live in the table view — click any column header to sort. We’ve also tightened up the grid cards to reduce the whitespace.
-
The region display now shows the most specific location we can identify rather than falling back to Unknown. If we know it’s Old Mission Peninsula, that’s what you’ll see.
-
The aging curve, Drink Now, and Peaking Soon views in Analytics pull from your drink window data. These populate automatically once drink windows are set — the registry does this for well-known wines, and we’re continually expanding coverage. If you have bottles that aren’t showing up, opening the wine detail will trigger a lookup.
-
The vintage charts now use color for drinking window status and shade for score, which should make both dimensions much easier to read at a glance.
Really appreciate you taking the time to write this up. Looking forward to hearing what else you notice.
Best,
Nick
Nick, the scores-in-cells and drinkability text are a huge improvement, thanks for running with those.
The new color scheme isn’t working for me though. It’s using four hues (green, yellow, red, blue) for drinkability, with shade varying by score within each. For someone with red-green deficiency, green, red, and yellow collapse together, so when I scan the grid I can’t tell if I’m looking at a great vintage or a weak one. I end up reading every text label individually, which means the color is adding noise rather than doing work.
I dug into the research on this and the interesting finding is that multi-hue schemes aren’t just a problem for colorblind users; they’re measurably worse for everyone. The brain doesn’t intuitively order hues. Colin Ware, a perception researcher at UNH, puts it bluntly: “If you give people the colors red, blue, green and yellow, they will not know which order to put them in.” What the brain does read naturally is luminance: light vs. dark. That’s why the standard recommendation from WCAG, Tableau, Datylon, and Atlassian all converge on the same thing: for ranked data, use a single hue varying only in lightness.
Easier to just show it. Here’s the chart with a single-hue green ramp for score, keeping your drinkability text labels as-is:
Here’s the interactive version: https://swsvz6.csb.app/ (blue variant here: https://x383yt.csb.app/)
Trying to show two things with color at once is making both harder to read for everyone, and impossible for me. Score is the natural fit for color since it’s continuous and rankable; drinkability already works well as text. I know I keep coming back to this, but the latest change moved away from what was working for me. The text labels are great and I rely on them entirely now; the color just needs to lean on lightness instead of hue to do its job.
1 Like
Hi Luke,
I’ve been hoping someone would provide some feedback about the colors. I was thinking the same thing but I’m struggling to find the proper gradient.
Thanks for the input!
Nick
That makes sense, and I can imagine it is hard. I’m not a design guru by any means and would probably dive into some of those linked articles/studies if I were going to select one for a dark background theme. Or, maybe ask Claude, chatGPT and Gemini to come up with recommendations based on a sample of what you are working with, and see how they compare to one another.