Ask Ratbert

[Ratbert]

@Ratbert My dealer said one thing that might be holding up my Jan JLUR order is that I ordered a manual transmission. Can you see anything on manual orders?

They're being built and the backlog between start of production and getting a window sticker is dropping, but that that time period is most likely due to other parts.

Instances of 6-Speed Manual Transmission over time:

day	build sheets	window stickers	backlog
2022-03-19	2,291​	2,155​	136​
2022-03-20	36​	25​	147​
2022-03-21	43​	41​	149​
2022-03-22	38​	35​	152​
2022-03-23	71​	1​	222​
2022-03-24	6​	0​	228​
2022-03-25	28​	0​	256​
2022-03-26	0​	47​	209​
2022-03-27	0​	137​	72​
2022-03-28	38​	42​	68​
2022-03-29	26​	28​	66​
2022-03-30	36​	27​	75​
2022-03-31	14​	17​	72​
2022-04-01	52​	41​	83​
2022-04-02	0​	19​	64​
2022-04-03	0​	11​	53​
2022-04-04	74​	40​	87​
2022-04-05	31​	10​	108​
2022-04-06	37​	25​	120​
2022-04-07	25​	25​	120​
2022-04-08	19​	26​	113​
2022-04-10	48​	71​	90​
2022-04-11	0​	0​	90​
2022-04-12	8​	29​	69​
2022-04-13	27​	6​	90​
2022-04-14	48​	63​	75​
2022-04-15	11​	17​	69​
2022-04-16	7​	19​	57​
2022-04-17	0​	6​	51​
2022-04-18	6​	4​	53​
2022-04-19	11​	17​	47​

Sponsored

 

[OfftheJeepN]

You're right. It makes much more sense when there's context.

month	# sequence numbers	# build sheets	% haven't started build	% w/build sheets	# window stickers	% w/window stickers	% of incomplete build sheets w
Oct '21	12,399​	10,116​	18.4​	81.6​	9,779​	78.9​	3.3​
Nov '21	13,099​	10,162​	22.4​	77.6​	9,844​	75.2​	3.1​
Dec '21	32,583​	18,651​	42.8​	57.2​	18,409​	56.5​	1.3​
Jan '22	30,742​	16,588​	46.0​	54.0​	16,202​	52.7​	2.3​
Feb '22	24,976​	8,677​	65.3​	34.7​	7,927​	31.7​	8.6​
Mar '22	22,898​	1,210​	94.7​	5.3​	798​	3.5​	34.0​
Apr '22	???​	10​	???​	???​	7​	???​	30.0​

Once again @Ratbert in his spare time is able to give us more information than anyone full time at Jeep, their dealership, or JC. Keep up the great work, it is really appreciated !!!

 

[dragoneggs]

Once again @Ratbert in his spare time is able to give us more information than anyone full time at Jeep, their dealership, or JC. Keep up the great work, it is really appreciated !!!

Ha ha... yeah thanks so much Ratbert, good stuff!

Looks like I might not be in a 1% club if I digested the data John provided, correctly. Recently took delivery on a 2dr Snazzy Rubicon. Dec 18th order and dealer received exactly 3 months later.

 

[DfromBama]

I'm trying to figure out what the April stats imply. Will be adding numbers to this soon.

Edit: see my subsequent post which has additional columns.

month	% w/build sheets	% w/window stickers	% of incomplete build sheets
Oct '21	81.6​	78.9​	3.3​
Nov '21	77.6​	75.2​	3.1​
Dec '21	57.2​	56.5​	1.3​
Jan '22	54.0​	52.7​	2.3​
Feb '22	34.7​	31.7​	8.6​
Mar '22	5.3​	3.5​	34.0​
Apr '22	​	​	30.0​

Edit: see my subsequent post where I've added lots of details to this.

Thank you so much!! Slowly but surely they’re getting built. You should start your own tracking page and charge for membership... I’d rather get my info from you than JC any day!

 

[bjm00se]

@Ratbert Hmmmm. There's an anomaly here I can't figure out. The "% w/Build Sheets" of Oct, Nov, Dec orders with build sheets (also window stickers) appears to have gone *down* between the 4/11 query and the 4/20 query. I don't see how this could be possible. They don't "unbuild" Jeeps.

( This query output was posted on 4/11/22 )

month	% w/build sheets	% w/window stickers
Oct '21	91.1​	85.2​
Nov '21	83.4​	78.3​
Dec '21	84.1​	60.2​
Jan '22	71.5​	54.5​
Feb '22	38.6​	30.4​
Mar '22	3.5​	2.0​

( This query output was posted on 4/20/22 )

month	# sequence numbers	# build sheets	% haven't started build	% w/build sheets	# window stickers	% w/window stickers	% in backlog
Oct '21	12,399	10,116	18.4	81.6	9,779	78.9	3.3
Nov '21	13,099	10,162	22.4	77.6	9,844	75.2	3.1
Dec '21	32,583	18,651	42.8	57.2	18,409	56.5	1.3
Jan '22	30,742	16,588	46.0	54.0	16,202	52.7	2.3
Feb '22	24,976	8,677	65.3	34.7	7,927	31.7	8.6
Mar '22	22,898	1,210	94.7	5.3	798	3.5	34.0
Apr '22	???	10	???	???	7	???	30.0

I was kinda assuming you were building a database of build-sheets and stickers that had previously been found, and then you re-ran the query each day against the VINS that hadn't yet been located.

If you're feeding all the VINS as query parameters each time, the output's going to be much more subject to timeouts and transient data errors where sheets that exist simply can't be located by a system that's probably not all that reliable to begin with?

At any rate - I'm impressed with the effort! There's a lot of code and admin effort here you've slung by evenings-and-weekends for-the-hell-of-it coding.

 

[TriumphJeepster]

I am too stupid to understand, but it's awesome to have this level of insight.

 

[boomersooner]

I appear to be in the 42% of December orders without a build sheet. Tomorrow will be 4 months since order.

 

[longbucks08]

@Ratbert Hmmmm. There's an anomaly here I can't figure out. The "% w/Build Sheets" of Oct, Nov, Dec orders with build sheets (also window stickers) appears to have gone *down* between the 4/11 query and the 4/20 query. I don't see how this could be possible. They don't "unbuild" Jeeps.





I was kinda assuming you were building a database of build-sheets and stickers that had previously been found, and then you re-ran the query each day against the VINS that hadn't yet been located.

If you're feeding all the VINS as query parameters each time, the output's going to be much more subject to timeouts and transient data errors where sheets that exist simply can't be located by a system that's probably not all that reliable to begin with?

At any rate - I'm impressed with the effort! There's a lot of code and admin effort here you've slung by evenings-and-weekends for-the-hell-of-it coding.

I can't remember the exact details, but I thought Ratbert said he had to exclude some of the data because it was flagged for export or something like that.

Also, Ratbert, didn't you say something about running 2021 data and finding out that a certain % of sequence numbers were never built? Do you remember what that percentage was? I wonder if that will be roughly the same for MY 2022 builds.

Finally, since you have to assume everyone looking at these threads have custom builds, is there any information in the build sheets or window stickers that would give us an idea of what % of being built is custom rides compare to the total population?

Thanks as always @Ratbert!

 

[MoHeTow!]

This is a pretty cool thread! Have you run the numbers for days per build by model? Just to see if certain models sit and wait longer than others.

 

[Ratbert]

@Ratbert Hmmmm. There's an anomaly here I can't figure out. The "% w/Build Sheets" of Oct, Nov, Dec orders with build sheets (also window stickers) appears to have gone *down* between the 4/11 query and the 4/20 query. I don't see how this could be possible. They don't "unbuild" Jeeps.





I was kinda assuming you were building a database of build-sheets and stickers that had previously been found, and then you re-ran the query each day against the VINS that hadn't yet been located.

If you're feeding all the VINS as query parameters each time, the output's going to be much more subject to timeouts and transient data errors where sheets that exist simply can't be located by a system that's probably not all that reliable to begin with?

At any rate - I'm impressed with the effort! There's a lot of code and admin effort here you've slung by evenings-and-weekends for-the-hell-of-it coding.

That's an acute observation. I started reporting only US dometic stats in an attempt to represent meaningful relationships between build sheets and window stickers. By excluding non-domestic builds I have a high degree of certainty that a missing window sticker is missing due to not existing versus being explicitly hidden or in other ways not existing on the sites that we know about.

That means that the relationships between numbers of build sheets and window stickers are now much more accurate. And, as you've pointed out, the relationship between US orders and the overall range of sequence numbers for a given month is now off.

I'll need to figure out a way to cleanly represent both sets of data points in a table like this without it all becoming overwhelming.

I'm also trying to figure out what to call the delta been build sheets and window stickers. I'm thinking that "being built" might best describe it.

 

[Deleted member 59498]

Asking Ratbet,

Will the big Comet found hit earth?

 

[bjm00se]

simplest solution I that occurs to me off the cuff would be to change the number of sequence numbers to:

Number of Sequence Numbers(*) - and apply the same filter.

EDIT - wait. That idea is stupid. You need a build sheet before you can decide whether to filter it. I'm an idiot. Never mind.

IF you could do that - then the numerator and denominator for the percentage are using the same units.

Lifting myself by my bootstraps here....

 

[Ratbert]

This is a pretty cool thread! Have you run the numbers for days per build by model? Just to see if certain models sit and wait longer than others.

Not really, but that's a good idea. My assumption is that it's specific parts that are holding things up. If that's a valid assumption then higher end models should spend more time between build sheet and window sticker. I'll see if I that theory pans out tomorrow.

Edit: gathering those details is significantly more challenging than I had initially envisioned, so it'll have to wait until I build those types of queries into my codebase. This isn't a simple / obvious database query.

 

[bjm00se]

At any rate, to me, a big part of the value of this representation is how it changes over time, so running that same view (using the now updated algorithm, assuming it doesn't change again) will show how much progress they're making on burning down the backlog of unbuilt orders in the older months.

 

[Ratbert]

I was hoping by my birthday 5/5 but don't look good ? ?

I've been trying to come up with an approach to render multiple parts in a way that lets you see which might be causing delays. Thanks for giving me something concrete to experiment with. I'll post the results if/when I get it working. It'll likely need to be in HTML since the table formatting options available here are extremely primitive.

Sponsored