[OEM-EMS-ODM] Reduce stock levels of electronic components by accessing distributor API

Managing manufacturer lead-times to reduce stock levels is not an exact science. Lead-times are inaccurate. Manufacturers favour one customer over another without considering their respective order creation dates. Many suppliers communicate their lead-times by technology and not by manufacturing part numbers. And of course, lead-time discrepancies exist in the same technology family.

Inaccurate lead-times lead to excess electronic component stock levels

That being said, you still have to update lead-times as often as possible. Any under-estimated lead-time can result in a shortage, and any shortage will increase your stock levels. Any over-estimated lead-time results in unnecessary order-change management, pull-in, pull-out or cancellation requests that can also lead to shortages or excess stock.

How accessing the distributor API can help reduce stock levels

An Application Program Interface (API) is technology which automatically connects machine to machine. In the past few years, electronic component distributors like Digi-Key, Arrow Electronics, Mouser, Rutronik and Farnell to name a few, have invested in this technology to provide access to the information you need on any electronic component. On-demand and in real-time, you can request the manufacturer lead-times per manufacturing part number (MPN).

Comparing ERP lead-times v Distributor lead-times.

Let’s assume that distributor manufacturer lead-times are updated more often than manufacturer lead-times in your ERP. Wouldn’t a comparison be interesting? To keep it simple for a blog post format, let’s compare only our customer’s ERP lead-times with the manufacturer lead-times we can access via the Digi-Key API. Filtering by live articles from the anonymized data of our customers, we have a representative set of 123k items with an ERP lead-time which also have a Digi-Key lead-time on the same MPN.

Below is a graph charting the result with the ERP lead-time on the Y axis and the Digi-Key manufacturer lead-time on the X axis. For example, the highlighted point below is an item with an ERP lead-time of 42 days BUT with a Digi-Key lead-time of 273 days, resulting in a shortage risk.

All red points are ERP lead-times underestimated by 30+ days. All purple points are ERP lead-times overestimated by 30+ days. All green points are ERP lead-times equal to the Digi-Key lead-time within ±30 days.

Of course this graph is difficult to read, so here is the (quite shocking) relative proportions of ERP lead-times:

• 9% are under-estimated by 30+ days
• 61% of the lead-times are within ±30 days
• 30% are over-estimated by 30+ days

Lead-time discrepancy per manufacturer

Some manufacturers have very good results with most of the ERP lead-times very close to Digi-Key lead-times. Here are the top 5 manufacturers with their percentage within ±30 days:

• Samtec 97%
• Harwin 96%
• Schaffner 93%
• Würth Elektronik 93%
• Broadcom 92%

And here are the worst in this analysis (many MLCC manufacturers here):

• United Chemi-Con 20%
• Walsin 22%
• ROHM 22%
• KOA 25%
• Samsung 28%

It isn’t surprising that lead-time which are not equal ±30 days are more likely to be over-estimated. For example, looking at Texas Instruments lead-times, they are within the consistent boundaries for 54%, under-estimated for 3% and… over-estimated for 43%

Are these discrepancies due to the distributors? Manufacturers? The fact is that sellers who are in contract have very little incentive to communicate accurate lead-times: lead-time drops impact the book-to-bill ratio, a metric which is watched closely by investors.

What is the impact?  Can we reduce stock levels?

While it is very difficult to calculate the costs of inaccurate lead-times on stock levels, it is possible to draw an interesting curve: what is the percentage of your bill-of-material value which is in firm orders x days prior to your production?

Take a look at the graph below. You can see that 51% of the value is in firm orders 105 days prior to production. As you will also notice, most ERP lead-times are in week multiples.

Now, let’s compare this to the curve that you would have by accessing the distributor API from Digi-Key. As you can see, this is now only 34% of the value in firm orders 105 days prior to production.