Smart Tender Automation: An introductionPosted December 14, 2017
What is Tendering?
Truck tendering decides for each shipment which carriers get offers to transport it, in what sequence, at what price, and how to handle rejection or non-response. It creates new cost-to-serve options (procurement) and is an execution step for each shipment (dispatch). When tendering fails the results are delayed shipments, unnecessary cost, or both.
In the most naïve case, tendering is done by following a set of rules, predefined for a long period of time. This can be done manually or via automation software. Most companies quickly outgrow transport demand that allows them to make manual decisions. Thus, Tender Automation - the software approach – is the focus of the remainder of this article.
A base-level implementation of Tender Automation is a tendering engine that simply offered a shipment to the lowest contracted carrier (or carriers) and then waited for their response before moving to the next.
The problem with such static, order-by-order approaches is that they ignore evolving market dynamics, the aggregate transport demand, and carrier relationships.
Adding Intelligence to Automation
Smart tendering solutions thus take three key factors into account:
- Intelligence about the likely carrier reaction
- Assumptions about the state of the market’s supply and demand
- Optimising over aggregate transport need instead of a single shipment
Know Your Carrier
Tenders use a valuable but overlooked resource: time. Any tender that holds up other tenders is using up time you will never get back. Therefore, incorporating a model of carrier behaviour makes sense.
The simplest version of this intelligence is not to hold up other tenders waiting for a likely non-responder. More advanced versions find pricing sensitivities on aspects like dock door slots, cargo types, or time of the day.
As in poker, smart tendering is about sensitivity to tells. In tendering those tells come from aspects like when a carrier typically reviews tenders, how fast they respond, in what order they respond, and so forth.
Machine learning in this area is much better than laboriously setting up rule trees and then keep them updated as the business environment shifts.
Know the Market
Tenders need to be attractive to the carrier, and to assess attractiveness one must understand the market in terms of supply and demand.
A tender that is likely to be accepted immediately in a low season would be under-priced in high season. Holidays, weekends, carrier bankruptcies, and the general flux of the trucking market are other examples when prices change.
Under-priced tenders waste valuable dispatching time because they will never be accepted. Overpriced tenders waste valuable transport budget. Either way, you’d rather the Tender Automation took this factor into account. To do so, the Tender Automation needs spot market data and strategies for setting price relative to it.
Be Long-Term Greedy
Goldman Sachs is known for coining the phrase “Long-Term Greed”. The idea is that greed is problematic when the horizon is too short, and there will be negative consequences later.
Naïve Tender Automation without considering the very next shipment is short-term greed taken to the extreme. For example, consider a carrier that costs 15% below other options, but has only one truck. Which job should be tendered to them first? The answer requires considering all other tenders, but also the long-term implications of such a decision.
Tendering Automation intelligence is about modelling how carriers would respond to various tendering patterns and finding the one that minimises the cost to serve for all transport jobs, not just the next on the list.
Without being smarter about decision making in light of available information, Tender Automation provides preciously little value to organisations. Applying the same rules to all shipments, year around is a recipe for suboptimal results. Both financially and in terms of customer service.
In my next blog article, I have a look at how leading TMS providers approach the issue of Tender Automation.