Skip to content
English
Create a ticket View All Tickets
Create Ticket
  • There are no suggestions because the search field is empty.

The Great Rate Engine

Enter the Great Rate Engine…

A truly Great rate engine will have all of the qualities of the initially mentioned “good rate engine”…

But should have these additional features:

Modular architecture – can be de-coupled from the TMS and operate as a standalone
Libraries – data variables, workflow, and calculation code that a non-technical Logistics User can understand:
  • Algorithm Library the primary component of the rate engine that calculates any variable input; algorithm can be User manually configured, or pre-set in the Knowledge models below (everything from Freight cost to Carbon Reduction Program calculations to complex freight broker agent Commissions)
  • Knowledgebase Library – works in conjunction with the Algorithm library; this library house all the analytics & decision-aided support Models; examples include:
  • setting up a Scorecard & ranking system for carrier performance
  • setup a Price-tolerance Ratio that considers your margin, the carriers willing rate, and the clients acceptable Cost
  • preferred (primary & secondary) carriers to use but rank using, in conjunction, their Risk Mitigation scores, their Performance scores, and the Price-tolerance ratio
  • Carbon Reduction Program (CRP) or other supply chain sustainability initiatives can be “scored” make decisions…not just for reducing cost, but for minimizing the ecological impact (CRP models can score & track how each consolidation plan or reduction of dwell time due to better planning & appointment scheduling can reduce CO2 impact, etc.

 

  • Qualifiers Library – any variable that can be used by the Algorithm library; this can include standard units such as the dimension & capacity of a 24’ box truck, or can be a User configured variable…for instance, a fresh produce distributor can configure his commodities & commodity mix restrictions, the reefer equipment needed, the temperature ranges for shipping, the handling unit requirements & size, and the lead-time requirements (deliver on x date)
  • API Library – if the rate engine is the “heart” of the TMS, the API library is its “vascular” system; these rate engine API will work within the TMS as well as enable rate engine communication outside of the TMS (i.e. get the R&L cost for Secure Access accessorial, etc.)
  • Agreements Library – contract formats & templates (Client cost, Commissions, Vendor Compliance chargebacks, etc.)
  • Freight Costs & Billing (including Accessorial & FSC)
  • Manage Service (Gainshare savings, freight spend, freight settlement services, etc.)
  • Scorecard rating for carriers (low score for on-time delivery), vendors (high score for route guide compliance), facilities (low score due to 3 hr. average detention time), etc.
  • Commission – can be setup for internal agents, contractors, and co-broker partners
  • Compliance & Chargeback – vender compliance in the retail industry is a necessary consideration to help calculate any PO invoice reduction due to chargebacks
  • Benchmark agreements can be setup for continuous measuring & monitoring for such programs as Gainshare manage service, freight as a % of Sales or freight as a % of shipping product value ceilings, etc.

Carrier Library (accounts) – pre-loaded, major carriers

Enhanced API rating – for efficient use of carrier webservices rating by added qualifiers & filters (i.e. only use Estes for distances over 375 miles…or never use R&L for Client A, etc.)

More importantly, all the above needs to be offered as Micro-services. These micro-services must work within as well as external to the TMS platform. For instance, if the Sales & Marketing department want to plan & price a product launch (obtain delivery cost, transit times & packaging restrictions)…micro-service components of the rate engine & API will be exposed or integrated into their core system so they can take this into consideration for their product pricing.

The benefit of having a rating engine with microservices will allow specific real time integration & interface with varying systems…some examples include:

· CRM systems to obtain carrier & vendor scorecard rating

· Finance system for accrued expense, budget, & pricing

· WMS for optimal pallet build based or order consolidation based on shipment schedule & routes

· HR systems to calculate Commission payout and revenue targets for employees & contractors

· Procurement (PO) systems to rate compliance Chargebacks

· other TMS – enable an outside TMS or multiple TMS sharing the same rate engine

 

Automation & AI – the Smart Rate Engine

Using the Algorithm & Knowledge Base libraries, Users can build an operational model that will initially filter down the choices to the top 5…and prioritize these top five to meet the programs objects.

Examples of the models that can be configured & built to provide prioritized responses that will 1) enable decision support to the User, 2) foster results-driven automation, and 3) build the foundation to enable AI & true machine learning:

§ Predictive modeling

§ Predictive pricing

§ Predictive leads to Prescriptive (what if leads to implementing a executable program with benchmarks)

§ Price tolerance ratios (PTR) & thresholds (measured value : expected value, etc.)

§ Cost based optimization

§ Max Margin optimization – make the most profit while maintaining route guide compliance, client expectation & requirements, and optimal carrier performance

§ Differential pricing

§ Network density

§ Lane price predictability (specifically for the shipper…not just the market)

§ Promised volumes targets

§ Rate Confidence Index

 

For example:

Rate shopping a specific Client shipment will usually return all the carrier rates that provide service to that lane…then the system will usually sort this in ascending order starting with Least Cost Carrier; this may result in 50 or more carrier rates.

But if the smart rate engine can also take into consideration:

a. Client’s route guide, preferred carrier

b. Client’s must ship or delver-by requirements

c. Carriers transit times

d. Carrier scorecard for on-time performance

e. Carriers damage claims ratio

f. Carrier promised-volumes

g. Lanes historical pricing (and specifically to this carrier)

h. Consider seasonality price fulgurations

i. Profit Margin targets and tolerances

j. Current market price / index

1) Decision-aided support - the resulting rate response can quickly reduce the 50 responses, down to the top 5 that should be selected…and in many cases…the least cost carrier is not always on this short list.

2) Automation - If the rate engine is monitoring the selection process, then we can also establish an automation model based on client, lane, commodity, selected carrier, acceptable price range, performance, and restriction…so that next time this shipment needs rating, the smart rate engine will just return the top 2 qualified rates, and eventually can automatically select the best scoring rate, once the User “trusts” the system.

3) Machine-learning – monitoring the shipment frequency & cycle, client requirements, carriers performance, freight profile, User response to the results...the rate engine can begin selecting the single, “true” best rate result. And we can add one more rating algorithm, the Rate Confidence Index, which a user can now set to “systems self-select” if the Rate Confidence Index has a min score of X

 

A report by The Visual Capitalist, shows that corporate Planners go through the same manual calculation steps every month for their S&OP; and procurement folks repeat approximately 40% of the same tasks each month, year over year. The same is true for Logistics & Transportation. The problem here is that most supply chain processes do not include analytics that learn and retain the knowledge and experiences of the past.

 

Finally, it’s not enough just to have a smart rating engine, but it needs to also be:

  • Easy to implement
  • Simple to maintain

Because of the micro-services architecture, any single rate or model can be taken off-line without interfering with critical operations.

Most importantly, the rate engine must have an easy-to-use Workflow tool that will allow a simple Business Process person, to quickly build a rating model by “pulling” together all the necessary micro-services pieces, then testing this model before “publishing” it to the live system.