Inventory Import Algorithm
DealerTrend has been managing automotive inventory since 2001. In that time we have evolved our ability to send and receive data. A big part of what we do happens with our data import algorithm. Below are the steps of that algorithm broken out to help you understand what is happening under the hood.
Import Watcher
This part of the algorithm test the the general state of the file to be imported and determines if the import should even start. There is quite a bit more happening under the hood, but at its core this part of the algorithm is testing these two criteria.
- Has the file changed? If the file hasn't changed at all it won't be imported.
- Is there anything "abnormal" about the feed? This works as a safety net for when feeds come in with bad data.
Translation
When receiving data from multiple sources it is important to translate into a single common format. The DealerTrend translation process has evolved over the years to cover just about every possible scenario.
Data Scrubbing
Vehicle scrubbing is a very important piece to the DealerTrend algorithm. Our goal is to help you market your vehicles online with as much quality information as possible. Since a VIN only has a finite amount of information available, our system looks at all the data available to us about a single vehicle in aggregate. Then based on rules that we have built over the years we determine the exact match vehicle and apply the appropriate features and equipment to it.
One key point that you should be aware of.
1) Our system won't "guess": If left with 2 or more possible exact matches we error on the side of safety. If you find that you have vehicles that are missing information you can enhance the data by adding things like 'trim', model code, etc.
Data mapping and filtering
Since no two dealers are the same we will review each new inventory feed to map the appropriate fields and add filters as needed.
Scripts can be set up to modify the incoming inventory. Common use cases:
- mapping price fields
- removing vehicles (filters can be by New, Used, Price, Make, etc.)
- removing the first photo
Merging multiple data sources
It is very common for dealerships to have more than one data source. Example 1: Used cars from a data collection company and New cars from the Dealer Management System (DMS). Example 2: Inventory from the DMS and prices from a third party vendor.
When these situations occur rules need to be applied to make sure that everything syncs up as expected. Below you will get an idea of how we treat feeds
Single Feed
- All data gets added.
Multiple Feeds (description below is for two feeds, but we can handle more if necessary (more than two is not recommended))
Primary vs. Secondary
When there are two feeds one feed becomes the 'primary' and the other feed is the 'secondary'. When data populates VMS it initially fills in the fields with the 'primary' data source, and then ANY field that did not populate from 'primary' will be populated by 'secondary' (if it exists). Note: ONLY one feed can be 'primary'.
| Feed1 | Price1 | Price1 |
| Feed2 | Price2 | Price2 |
| Primary? | Feed1 | Feed2 |
| Result | Price1 | Price2 |
Additive vs. Complimentary
Depending on which feed is 'primary' will determine which vehicles show up, based on how the 'secondary' feed is merged. The two choices are 'Additive' and 'Complimentary'.
- Additive - adds data from BOTH feeds.
- Complimentary - adds data if the vehicle exists in a existing feed.
For the table below vehicle 'X' is in the following data feeds, and 'A' is 'additive', 'C' is complimentary.
| Feed 1 | X | X | X | X | ||
| Feed 2 | X | X | X | X | ||
| Mode | A | C | A | C | A | C |
| Result | X | X | X | X | X |