Auto-Select Deep Dive: Maximize Pricing Selection

Auto-select is one of the most-used buttons in CalcuQuote, and one of the least understood. If your first pass is not pricing 80 percent of your lines, the config is the problem, not the data. This session demystifies what those preset buttons actually do, the difference between filters and preferences, how the Jaro Winkler cross-score handles MPN spelling variation, and how to audit your setup on the review screen. Plus quick tips on merging AML tabs upfront, because cleaner BOM input is the upstream half of the auto-select story.

Headshot Remington Leland_250x250px
Remi Leland
Account Manager, CalcuQuote
Headshot Mike Turner_250x250px
Mike Turner
Product Manager, CalcuQuote

Remi Leland and Mike Turner break down how auto-select actually works under the hood. The focus is on understanding configuration versus data quality, how filters and preferences influence results, and how to audit and improve your setup so first-pass pricing coverage improves without manual rework.

Request a Demo

Key Takeaways

Filters narrow the pool, preferences break ties

Know the difference and you stop accidentally filtering out 60% of your BOM.

  • A filter excludes any line that does not meet its criteria

  • Require custom reels and a part that cannot come on a custom reel is out

  • Preferences only matter when multiple options satisfy the filters

Know what each preset button actually does

The two quick buttons are configurable, not magic.

  • If you have been pressing them for years without auditing the config, the criteria may not match how you buy today
  • Audit annually, or any time your sourcing strategy changes
  • A 10-minute review prevents a quarter of bad pricing

Three common scenarios drive most setups

Pick the scenario you actually run before you start tuning.

  • Need it fast
  • Need it cheap with blanket orders, where excess is fine
  • Need it cheap with no excess
  • Each maps to a different filter-and-preference combination

Cross-score uses Jaro-Winkler, not AI

The cross-score filter is a string-comparison algorithm, not a model.

  • An MPN with a dash and an MPN without one score "high" but not "exact"
  • CalcuQuote’s top recommendation: run with cross-score greater than or equal to high, which allows for packaging variation
  • Run exact only if you need strict matching

Three passes, configurable, audit via the review screen

Tight on pass one, looser on pass two, fallback on pass three.

  • You set the rules for each of the three passes
  • Healthy result: 80%+ of your lines priced after the first pass
  • Below that, your filters are likely too tight or too loose
  • The Review tab’s unpriced and cross-match buckets show what auto-select missed and why

AML on a separate tab? Merge it first, then auto-select runs clean

A simple upstream cleanup that gives auto-select the data it needs from pass one.

  • When the BOM structure and AML list arrive on separate Excel tabs, get them into one workbook
  • Find a common identifier (customer part number, internal part number)
  • Use the merge spreadsheet option in the top right of import
  • One template setup per customer means subsequent BOMs follow the same join