Picking a winner

Schedule reliability indexes becoming important parameter in ocean carrier selection.

   February is always a crucial month for transpacific shippers, coming as it does near the beginning of ocean carrier service contract negotiating season.
   This year, the stakes have been raised with systemic operational issues and the threat of work stoppages at U.S. container ports, which makes picking the right carriers, and choosing the right allocations for each of those carriers, all the more vital for shippers.
   One element of those carrier selection and allocation decisions is service reliability—that is, the extent to which carriers arrive at the port of destination on time relative to their published schedules.
   In the past, schedule reliability information was more interesting than informative, akin to a top 10 list as opposed to a practical decision-making criterion.
   But things have changed in recent years, thanks to the proliferation of two things: more detailed and available schedule reliability indexes; and shipment management systems that allow shippers and logistics service providers to account for the data from those indexes in more actionable ways.
   In short, shippers and logistics service providers (LSPs) can use schedule reliability data to decide not just what carrier to use, but also which port pair and broader market have more reliable ocean service. And they can use that information in a more data-driven way, rather than by intuition about the way a carrier is performing.
   This development has deep supply chain implications, even all the way back to sourcing, as it could allow a shipper to pick a manufacturer in a region that provides more reliable ocean service options than one they may currently be using. So understanding the power of schedule reliability is not just about picking the best carrier on an existing trade lane, but about deciding whether a shipper should even be involved in that trade lane at all.
   To be fair, the use of schedule reliability data as a true objective metric in ocean procurement decisions is still in its infancy. While shippers talk about wanting better schedule reliability from their carriers, on-time performance still lags far behind rates, actual transit times, and capacity guarantees as a determining factor in carrier selection.
   But the point is there are existing and emerging tools that allow shippers to make reliability a higher priority in their decision-making matrix, should they choose to do so.
   Examining carrier performance at this time of year is a logical step—this is the time when shippers decide if their existing carrier’s service matches its rates – but these tools are best utilized on an ongoing basis, not as a one-off.

   “Traditionally, many shippers have analyzed their carriers’ performance for the year to prepare for the next year’s contract negotiations,” Jim Wu, business development manager for the shipment management software provider CargoSmart, said during an American Shipper webinar on schedule reliability in December. “However, measuring schedule reliability is not just for negotiating contracts with carriers. Shippers can use the information all year round to measure performance more frequently, negotiate updated rates into your contract, and to plan alternative routings, or partner with new carriers to meet new customer demand. Even to measure carriers during unexpected events, such as port strikes.”
   Philip Damas, director of Drewry Supply Chain Advisors, said during the same webinar that incorporating information about schedule reliability into carrier selection and ocean procurement is part of a broader industry development.
   “Things are becoming much more data-driven now for shippers and logistics managers,” he said. “You do your procurement policy based on data, and your carrier selection based on data, not just gut feel or relationships.”
   Shippers and LSPs need to strive to have a better understanding of how rates impact their business versus reliability.
   “It’s important to identify the tradeoff between cost and reliability in monetary and service terms,” Damas said. “So there’s nothing wrong with using low-cost and unreliable carriers. It can probably work for nearly every shipper. But you should not use low-reliability carriers for 90 percent of your freight. You have to reduce that, or limit that, to a particular percentage which works for you, and allocate the rest to reliable carriers.”
   Damas added the next step will be for shippers to integrate schedule reliability metrics in a truly objective way into their procurement process.
   “Shippers are starting to use these tools, mainly because they’ve been bitten a couple times, and have become wiser,” he said. “But still today, when I speak to shippers, they tell me their top two criteria are freight rates and schedule reliability, but in practice, they don’t use schedule reliability with the same level of depth and analysis as you would expect.
   “So we don’t yet see a real focus and a real use of analysis of schedule reliability for carrier procurement decisions. I think you see logistics managers who say they will only work with reliable carriers, but you don’t see them do this in a scientific way. We’re not yet at the point where there’s a proper measurement of the tradeoff between reliability and cost,” Damas explained.

Maturity Curve.   What Damas suggests is that there’s an evolution shippers go through in understanding and integrating the information that schedule reliability indexes provide.
   “There is a maturity curve that shippers go through that might be broken down into five stages,” said Virginia Thompson, senior director of global transportation and trade compliance for Crate & Barrel (and a member of American Shipper’s editorial board). “The first stage is, we don’t think about this at all or try to understand how reliable our providers are. The second stage is, we look at their published reliability stats (self-reported by the carrier, of course) and may consider that in making decisions.
   “The third stage is, we get regular reporting from the carrier on how reliable they have been on our shipments, because frankly, averages across all their customers may not be reflective of what our shipments will feel like. This is still self-reported, but one doesn’t generally have reason not to believe the data. The fourth stage is, we have some sort of internal database, maybe in Excel or Access or some other tool populated manually by us, our broker, or whoever, that gives us some data to report on, but we know it’s probably not 100 percent accurate.
   “Finally, the fifth stage is, we have implemented a robust international transportation management system with EDI (electronic-data-interchange) feeds from the carriers, and awesome reporting abilities, so that we have reliable data that we can report on regularly to measure carrier reliability,” she said.
   Indeed, most shippers at or near stage five would likely understand well the benefits of carrier schedule reliability on their procurement activities, and would be close to knowing how to integrate that information into actual decision-making (if they aren’t doing so already).
   Wu said there are a couple dynamics that can unlock the power of schedule reliability information.
   “Reliability info is no longer a single-usage report,” he said. “It’s an analytical tool that can be used throughout the year.”
   This gets at the way reliability indexes have evolved, from quarterly published reports that felt more anecdotal than practically useful to the present, where information is updated on a constant basis.
   The second dynamic, as Wu described it, is that shippers can consider reliability along parameters that go beyond the global carrier level. Again, previous incarnations of reliability indexes tended to focus on the reliability of a carrier relative to other carriers on a global basis. It was nice to know, but of limited use to a shipper using a particular port pair.

Beyond Carrier Vs. Carrier.   Now, however, shippers can use third-party indexes to measure reliability by carrier, port pair, trade lane, or market. Wu suggested shippers use an aggregated methodology for comparing schedule reliability rather than just looking at a single carrier’s schedule reliability in isolation.
   “At the carrier level, you can of course compare your current carriers against the reliability of the current market, or against the most reliable carrier,” he said. “You can also review how reliable your carriers have been over the past quarter, the past year, or even historically. Check to see if there are patterns emerging, and which carriers really are the most reliable.
   “At the trade level, you can see if the trade you’re using is the most reliable and efficient. Is there a better route, with fewer schedule changes, and a better selection of reliable carriers?” he said.
   CargoSmart tracked 1.3 million vessel schedules from 20 carriers between July and September 2014. Those 1.3 million schedules had 22.5 million changes, or an average of 17.3 changes per service. That average fluctuates, of course, by carrier, but it also fluctuates by origin and destination port.
   CargoSmart statistics show that schedules from Shanghai and Singapore suffered the most changes in Asia, while those at Norfolk and New York/New Jersey were affected the most in North America.
   So how does this information affect ocean freight procurement? As an example, Wu said a shipper might look to an all-water route from Asia as it might be more reliable, even if transit times are longer, allowing you to provide a more accurate delivery time than what you are currently shipping on.
   That flexibility becomes even more pronounced if you look at schedule reliability at the market level.
   “As you look at manufacturers and partners, the reliability of getting your cargo out is just as important as analyzing the carriers’ rates,” Wu said. “It’s time to take into account if a port of origin, such as the Port of Singapore, is still a viable port for you, with so many schedule changes. Or, if a new potential manufacturer you may be looking at, moving through (the Port of) Surabiya, is a more viable option because it’s just more reliable as a port.
   “If you need to ship from Asia to Oakland, traditionally, a shipper might have looked simply at the best performing carriers overall, or to Oakland. But you could, in fact, compare Singapore and Shanghai as origins, compare which lane has the most reliable schedules, which ports or markets from where you might ship have the most reliable schedules, and which carriers are then most reliable in those markets,” Wu said.

How Important Is It?   But there are other questions: Is reliability really that important, or have shippers been conditioned to expect vessels to be delayed? What’s more, is port-to-port reliability the key component in overall shipment reliability? 
   “Many shippers actually build a buffer time so that when a carrier says, ‘I will arrive on a Monday,’ in the shippers’ internal systems, they actually expect the ship to arrive on a Tuesday or Wednesday to avoid bad surprises,” Damas said. “The impact is that you can miss delivery windows, which is really bad news in terms of operational performance if you’re a logistics manager. Even worse, you can run out of inventory. This is one of the worst crimes in logistics and supply chain management. There’s also a cost attached to this.”
   He said a major structural delay, like the current U.S. West Coast ports situation, can severely hamper schedule reliability and throw a supply chain out of whack.
   Crate & Barrel’s Thompson agreed.
   “We have some sort of understanding of the cost of holding inventory,” she said. “We use that metric when weighing decisions like how often to purchase replenishment of non-seasonal items. Larger purchase orders that ship less frequently can ship more cost effectively often, but then you are holding the inventory longer. So we use that cost when, for example, analyzing the (return on investment) on the Amber Road tool (Crate & Barrel uses Amber Road for global transportation management and visibility). Because we know that we do pad transit times for variability and any days that we could take out of that padding would save us money.”
   Thompson, however, downplayed the impact of schedule reliability, especially in relation to the importance of actual transit times.
   “I’m not convinced that there is enough variability in carrier reliability that I would take out padding based on a carrier shift,” she said. “There are so many other things that go into it, that I don’t feel carrier reliability is that impactful. That’s not the same thing, though, as carrier transit times.
   “If a given carrier has one transship on a given lane and another one has two, we are absolutely going to adjust our lead times if we switch carriers. As far as expedited shipments, I don’t feel that when we expedite it’s due to unreliable transit times from a given carrier. It’s going to be because the shipment is just needed more urgently than standard ocean can get it here,” she said.
   Thompson added that liner carrier reliability is really only a piece of a larger puzzle for which shippers must plan.
   “How important reliability is depends on your commodity or business,” she said. “Crate and Barrel, for example, is probably on the lower end of that scale, because roughly half the products we sell are ongoing, so they don’t have the same urgency as an item that is highly seasonal and will only be in the stores for a few weeks or months. Missing even a few days of an anticipated selling season on a short-term product like that is expensive.”
   Thompson also pointed to manufacturers that have a greater need for dependable ETAs to supply just-in-time inventory planning for assembly lines.
   “But a carrier’s reliability is only part of what drives consistent transit times,” she said. “There are so many parties involved (the manufacturer, overseas trucker, U.S. trucker, railroads, customs brokers and, of course the U.S. and foreign governments) that can impact the transit time of an ocean shipment, that it’s very hard to be truly consistent over time, even if the carrier is perfect.
   “We consider (schedule reliability) a little, but not too much,” she said. “It would never override the importance of rates or other service issues, like getting capacity, proactive alerting to issues, or accuracy in billing.”

Reliability Defined.   One of the sticking points when it comes to measuring carrier reliability is that there is no one way of doing it. Each index, whether it be a third-party index like that provided by CargoSmart or Drewry, or an internal measurement provided by a specific carrier, differs at least slightly from others.
   The general consensus is that on-time means the ship has arrived within a 24-hour window of its published ETA, but even that can be fuzzy. Is that 24-hour window based on a specific time of arrival, or a calendar day in general? If it’s the latter, the fuzziness only grows.
   For example, a ship due to arrive March 1 at 8 a.m. might actually arrive March 1 at 11 p.m. By the calendar day metric, that vessel would be considered on-time. But if that vessel was due to arrive March at 11 p.m. and arrived instead March 2 at 1 a.m., it might be considered a day late, despite being only two hours late, instead of 15 in the first scenario.
   Most modern metrics better account for this scenario, but really, this speaks to the importance of a shipper tailoring its tolerance for delays to its supply chain.
   “What is the threshold for being late?” Wu said. “It’s critical to drill down into the details to make sure the reliability details match the performance that you want to measure, by port, port pairs, and/or timeframe. As more data is coming into play, you should be able to try to select your own tolerance. The general consensus is the 24-hour window, but you may think that’s too much. And that can be done, and it should be considered.”
   Damas said Drewry, which uses CargoSmart solutions to power its reliability index, has switched its methodology to more accurately portray what is considered an on-time arrival.
   “We previously said a ship was on time if it arrived the day it was supposed to, or a day earlier,” he said. “With CargoSmart we’ve moved to a stricter or closer tolerance. It has to be within 24 hours of the time the ship was expected. Most shippers accept that, but they will also build one day of buffer time in their internal systems.
   “Import retailers often have a three-hour window for delivery into their (distribution center). That’s the toughest criteria I’ve seen in the industry,” he said. “If you compare that with the tolerance of the vessels, where only 60 percent arrive on the day, clearly there’s a big gap between what’s happening at the port and what’s expected at the DC.”
   The 60 percent figure refers to Drewry’s measure of on-time arrivals over the eight years it has tracked reliability.
   “The carrier industry does not have a good track record on punctuality and reliability,” he said. “The long-term average is 60 to 65 percent on-time, meaning arriving on the day promised. There is a big range between the most reliable carrier—currently 75 percent on time—and the least reliable—currently 46 percent.”
   Damas explained the carriers can be broken into three tiers of reliability performance—those with an average of 64-75 percent on time, those who are 53-64 percent on-time, and those who are 46-53 percent on-time.
   “Whenever possible, go for the top tier,” he suggested, even if that top tier still means that one-quarter to one-third of sailings won’t be on time. “These are not minor differences in reliability.”
   The difference in on-time arrival performance between Shanghai and New York for 15 different services ranged from 100 percent at the best, and 20 percent at the worst, he said during the webinar.
   “Some people say there are no differences between the carriers, or that it’s a commodity business, but it’s not,” he said. “There are big differences between different carriers, or even different loops, or even between carriers in the same alliance.”

Big Ships And Alliances.   Wu said schedule reliability as a defining metric will only grow in importance as shippers grapple with the effect of 10,000-TEU and larger ships, and new carrier alliance configurations.
   “One area where we may see changes is on the north-south trade lanes,” he said, noting the major east-west alliances may move to those trades next. “So as the alliances form, some of the services may have the same name, but the reliability of the service can change depending on the operator. Hopefully, over time, it will lead to better performance, as carriers optimize their services.”
   As for the impact of big ships on schedule reliability, Wu said the biggest ships in operations today are missing their arrival windows more frequently.
   “Longer turn-times in loading and unloading are adding on to the schedules,” he said. “And not all of the carriers are keeping up to date on that, especially if there are alliances in play.”

Recommendations.   Damas said shippers looking to account for reliability in a systematic way have a couple different avenues.
   “There are two main sources—your own internal measurement, through an ERP or a shipment execution system,” he said. “In some cases a 3PL can run this type of shipment execution system for the shippers. The advantage is you can measure any kind of KPI – not just arrival at the port, but arrival at the inland ramp for intermodal, or arrival at DC. So you can get many good KPIs with a good internal system. The disadvantage is you only measure the carriers you are currently using.
   “The alternative, or complementary source, is that you use an independent monitor of the entire market, such as CargoSmart or Drewry, and there you can compare the reliability of every single carrier in the world nearly. The disadvantage is that you can only access public information, primarily arrival in port, or departures in port. But there’s not enough public information today for arrival at the inland ramp. In other words, it’s a broader carrier base, but less detail in the KPIs,” Damas said.
   More broadly, Damas suggested each shipper look at the composition of its cargo and decide what percentage requires greater carrier reliability.
   “If you have a relatively broad business, you have to divide and segment your shipments between those that are time-sensitive and those that are not,” he said. “And based on each one, you will have to plan your policy and carrier selection accordingly.
   “You have to plan for the risk of late deliveries because they will happen. You have to have some sort of buffer inventory, you have to have some sort of contingency plan, like being willing to air freight, or adding some buffer time in your transit times.
   “It’s important to identify reliable carriers, both current and potential, before your freight tender. So you have to be aware of which are the good carriers you have—presumably you know that already—and know that there are some other carriers out there with a high, proven level of reliability, then you should invite them for your tender for time-sensitive products,” Damas said.

This article was published in the February 2015 issue of American Shipper.