A few years ago, the main themes at Manhattan’s annual user conference were focused on their platform journey — 2010:Platform Thinking — >2011:Platform Activation — >2012:Platform Payoff. The investments that Manhattan has been making in their architectural underpinnings over the past decade or so laid the foundation for the launch this year of Manhattan Active™ Solutions, comprised of Manhattan Active Inventory, Manhattan Active Omni, and Manhattan Active Supply Chain applications. Some of the key attributes of the Manhattan Active Solutions are:
- “Always current, seamlessly interconnected, continuously adaptive, run anywhere”—That’s how Manhattan describes the architectural model and rapid release cadence of the Manhattan Active Solutions. This means much of their portfolio1 has been architected to run as a single-instance, SaaS solution, but with the option for the customer to run it on-premises if desired, with mechanisms to continually keep their on-premises code current on the latest version. Customers on either SaaS or the on-premises version can run with ‘zero downtime’ for upgrades. That is, server components from version X and version X+1 can run together simultaneously, which allows for uninterrupted upgrades.
- Single view across channels, locations, solutions—All modules across the portfolio share a single data base and data model, ensuring that everyone is working off the same single version of the truth. This is critical in an omnichannel world, to ensure reliable order promising, and alignment of execution.
- Extensible—Manhattan is moving to a component-based microservice architecture, currently comprised of over 5,700 RESTful APIs,2 enabling customers and partners to build extensions that are ‘future-proof,’ in the sense that they are guaranteed to survive upgrades.3
- Hyper-scalable—As transactional volumes spike (say during a big promotion or holiday event), the platform detects the need to scale itself up and spins up more resources, without the need for human intervention. It de-provisions those same resources once they are no longer needed.
Manhattan Active Inventory: Advanced Planning
The modules comprising Manhattan Active Inventory are Advanced Planning, Demand Forecasting and Replenishment, and S&OP. Manhattan’s Advanced Planning4 is used for retail financial planning, merchandising and assortment planning, and promotional planning, primarily for short lifecycle / fashion items.
Manhattan Active Inventory: Demand Forecasting and Inventory Optimization (DF&IO)
Demand forecasting and replenishment/inventory optimization are sold together as one capability. The system enables the user to remove anomalies, such as promotions or weather events, from historical data so that you have a clean unbiased perspective of demand to feed the forecasting engines. From this history, seasonal profiles representing normal demand (sans promotions or other events) can be created at the SKU level or higher up in the product hierarchy. DF&IO can do multi-echelon forecasting — starting with a Store-SKU level forecast, the system models upstream demand from there by factoring in the current inventory levels at the various DCs, inventory in transit, replenishment cycles from DC to store, lead times from vendor to DC, economic order quantities and other factors. It calculates the required DC service levels based on store-level service targets. In some cases, a 99% service level at the store may be achieved with a substantially lower service level at the DC.
Demand profiling can also be used to help predict the demand curves for new product introductions. Replenishment tools allow up to five ‘what if’ simulations to be run side-by-side, setting desired service levels by SKU by location (items might be considered an A item at one location and a C item at another).
Inventory Layer Analysis and Advanced Safety Stock Simulator
Inventory Layer Analysis shows the different ‘layers’ or purposes of each slice of inventory. Examples of inventory layers contributing to the total inventory held at a store include minimum presentation stock, safety stock, cycle stock (accounting for daily or weekly or X replenishment cycle, and economic order quantity), SKU requirements (case size or minimum order requirement from vendor), forward buy (when excess inventory was bought to take advantage of discount), and so forth. This provides transparency into why you are holding the inventory that you have at each location.
The advanced safety stock simulator lets you try out different service levels for different scenarios. Then, the system can make recommendations about replenishment, as well as opportunity buys, and transferring inventory between locations. For example, opportunity buys can be evaluated, weighing the extra carry costs vs. the savings. Inventory frontloading (buying ahead of demand to stock up for the peak season) can be evaluated, weighing the tradeoff of capacity constraints against carrying costs. The system reviews every SKU at every location every day, what is received and what is shipped or sold, to make suggestions of orders to be placed. Replenishment orders can be set up for auto approval once users have become highly confident in the system’s recommendations.
Each buyer’s inventory-related performance can be tracked and auto-emailed to them and their managers, showing service levels achieved, overstocks, and other metrics, benchmarked against the performance of their peers.
S&OP
Manhattan just introduced their new S&OP Workbench,5 which allows modifying the plans to respond to changes in demand, supply, and other factors such as an unplanned promotion or the business reaching its credit limit (necessitating a temporary pullback in spending). The Workbench allows various scenarios and approaches to be simulated, and view the impact on sales, inventory, and service levels before committing to the new plan. It allows the multiple stakeholders (marketing, logistics, store sales, procurement, and so forth) to view, collaborate, and suggest changes. Once a new plan is agreed on, it can be pushed into DF&IO (Demand Forecasting and Inventory Optimization) with a click.
Manhattan Active-Omni
Omnichannel is not merely about selling through multiple channels. It is when the retailer provides the consumer with a unified and seamless experience regardless of the channel. It requires the retailer to have a unified view across channels, which is nearly impossible when they have separate siloed systems for POS, clienteling, ecommerce, contact center, order management, store operations, store fulfillment, inventory management, pricing, promotion, and so forth. Manhattan Active Omni brings all of those together into one unified cloud-native application, with a single shared data model. This provides a single, unified view of all transactions and inventory across all locations. This unified approach also speeds implementation. Once a customer is live on order management, they are already part way to implementing POS and other components, because the core application/shared components and live data are already there.
Adaptive Network Fulfillment
Adaptive Network Fulfillment, a state of the art order fulfillment optimization engine. It provides a single view of inventory across the network and compares various options for fulfilling orders to select the optimal location and method, meeting the retailer’s objectives. A retailer/distributor/manufacturer might start with simple objectives, such as minimizing estimated shipping costs, or the minimum number of shipments per order, or ship from the closest location. They might then integrate feeds from FedEx, UPS, and other carriers to get actual shipping costs in real-time. Then they can add factors and rules to fine tune the fulfillment decisions to their specific objectives and situation. They could start to factor in labor costs and availability. They could factor in rejection rates (the rate stores reject orders sent to them) and direct more orders to stores with a lower rejection rate, to minimize bounced orders and reduce fulfillment times. Factors could include:
- Store Operations — Fulfillment capacity, cut off times, labor costs (by location), past performance, rejection rates
- Inventory — Days of supply, distressed units
- Selling Price Differential — Selling price vs. local price, regional/seasonal markdowns
- Customer Satisfaction — Promised delivery dates, service levels, probability of fulfilment success
- Cost — Minimizing split shipments and cost per shipment
The tool converts each of these factors to a cost. In this example $1.84 is added to the cost of fulfilling from a store with a high rejection rate, and $1.05 is subtracted because that store has excess fulfillment capacity available.6 In this way, the system will select the lowest total cost fulfillment option, providing complete transparency into how it is making that judgement. The customer can thereby continue to tune and improve the model to more closely fit their objectives. The administrator/user can configure the weighting and maximums for these costs. Manhattan has been developing their DOM for over a decade, so this is a mature tool that incorporates lessons learned and subtleties like the ability to proactively use up inventory where it is likely to be marked down or balancing workloads across the network.
Manhattan Active Store Solutions
Retailers are being forced to reimagine and reinvent their stores. As they do, store associates are being asked to do more and more: pick, pack, and ship online orders, provide personalized service, help customers with combined online and in-store orders, receive shipments at the dock door, prepare order for customer pickups, check out and take payments at the cash wrap, schedule fitting room appointments — the list and variety of responsibilities just keeps growing. The traditional approach of having separate systems for each of these functions is increasingly untenable. Manhattan, whose origin was in the warehouse, has steadily built up their portfolio of in-store capabilities. With the introduction of Manhattan Active Store Solutions, they have combined those capabilities under a common set of components, providing the various functions the store associate needs, in a single integrated and extensible application model. Store associates can stay in the single app and transition easily between the tasks, then return to half-finished tasks, supporting the interrupt-driven nature of today’s store associates’ work day.
Manhattan already supported iOS and Android platforms and now has added Windows 10 and browser based versions,7 to leverage PCs and existing hardware, in particular for the fixed cash wrap (with its existing PC-based hardware, tied to a scanner, payment reader, and receipt printer). Any extensions added to the system are made available across all the platforms. The frontend store associates app can be updated independently of the backend server applications.
The application provides a single view of the customer and all their past and current activities — online, in-store, returns through any channel — all on the same screen. It also provides a single view of inventory, no matter where it is, so store associates are empowered to sell inventory anywhere across the network and have it delivered in whichever location and method the customer prefers. One area that Manhattan has done well for several years is in creating an intuitive UX for the store associate (see Manhattan’s Store Inventory & Fulfillment App). This was apparent in the Manhattan Active Store Solution for picking orders, which has guided flows, confirmations, a progress indicator (number of steps left), navigation, and confirmation of correct picking.
Manhattan Active Supply Chain: Manhattan Active WM
Manhattan Active Supply Chain consists of Manhattan Active TMS and Manhattan Active WM (Warehouse Management). Here we cover Manhattan Active WM. Manhattan WM is one of the most mature and sophisticated warehouse management systems in the market. Their customers have built many extensions and customizations, making upgrades more challenging. Customers typically upgrade to a newer version every three to five years, because of the effort involved. In between these upgrades, the customers are missing out on all of the new capabilities and innovations that Manhattan develops and releases every year.
In an effort to provide simple and immediate access to the new version capabilities, Manhattan has come up with an innovative solution that allows customers to upgrade much more smoothly and easily every year, and thereby keep up with the latest functional and technological advancements. Manhattan has an annual release cadence for WM, typically in the spring each year. Manhattan will work together with customers who subscribe to the Manhattan Active WM program to co-author automated tests are then executed monthly across the customer’s WM instance. These are used to test the customer’s extensions, unique data elements, and specific configured workflows. Automated ‘test harnesses’ simulate any external systems that the WM system interfaces to, such as ERP, material handling systems, voice subsystems, parcel systems, and so forth.8 This ensures that problems are discovered and fixed as soon as they arise, throughout the year, as Manhattan adds new capabilities, building towards their next major release. This makes it easy for the customer to move onto the latest release every year, thereby taking advantage of all the latest and greatest technology. Manhattan Active WM was announced at Momentum for immediate availability. I was told there is lot of interest in it.
Order Streaming and Employee Engagement
Manhattan integrated WMS with Labor Management several years ago. Over the course of the last decade, their retail and e-commerce customers are seeing higher volumes of smaller orders, more variable volume, more demanding SLAs9 (shrinking time periods), and greater uncertainty about labor and capacity demands. In short, it has become harder to predict the order flow and ability to meet the demand. In response, Manhattan is introducing Order Streaming, a waveless approach to balancing demand with warehouse resources — a ‘master orchestrator,’ removing the peaks and valleys to create a steady stream of orders in from the OMS and out to the carriers. It is a single solution to manage orders, inventory, automation, and the workforce; dynamically swapping and reallocating resources to higher priority orders as they arrive. With Order Streaming, orders can be inserted into the fulfillment process when the downstream resources (such as the packing station) become available. Users of Order Streaming are seeing a 15% increase in fulfillment capacity without adding any automation or labor management.
A couple of years ago, Manhattan introduced their WMS tablet application (DM Mobile) for managers to track the operation’s performance. Now they have introduced Employee Engagement, which provides a gamified approach for all the warehouse employees to see exactly how they are performing against their personal goals and how they compare to peers. Manhattan likened it to a fitness tracking application. Both Order Streaming and Employee Engagement are free for Manhattan WM customers.
Manhattan’s Footprint — from DC to Store
Manhattan is the largest standalone supply chain execution software company. They have had solid revenue growth and excellent profit growth over the past five years, and look set to continue that trend. In addition to a continued healthy WMS market, the big area of opportunity for them is the store. As described in Stores’ Identity Crisis: The Reimagining of Physical Stores, we are in the middle of an existential crises for bricks and mortar retailing that is forcing retailers to reinvent what their stores are and do. This provides Manhattan an opening to become the next generation ‘store operating system,’ powering POS, fulfillment, task management, and overall store operations. They are not the only one with this vision, but perhaps the only one with a market leading warehouse offering and DOM to go with the store operations functionality. Manhattan said they expect their investments in store systems to bear financial rewards over the next 18 to 24 month timeframe. There is a lot of pressure to increase the speed of innovation, and they have put in place the software testing and release infrastructure to respond to that. It is an exciting time to see who will win in this reimagined store — certainly Manhattan is a key contender.
_____________________________________
1 The one exception for now is Manhattan’s core WMS offering. It represents the most lines of code, the most customer customizations, and the biggest on-premises installed base for Manhattan. Therefore, it will take a bit longer to migrate their WMS to a true SaaS architecture, compared to the rest of the suite. However, they already are taking an interim step with the Manhattan Active WM offering, which is described further in this article. — Return to article text above
2 RESTful APIs are ‘Representational state transfer’ (REST) web services that expose microservices, enabling highly granular extensions and integration into an application. — Return to article text above
3 Provided development guidelines are followed. — Return to article text above
4 Not to be confused with Advanced Planning and Scheduling (APS). Here “Advanced Planning” refers to merchandise financial planning in fashion retail. — Return to article text above
5 S&OP Workbench requires the user already has DF&IO. — Return to article text above
6 The amount depends on the weighting of each. For example, suppose you elect to provide an incentive of 40% of shipping costs for shipping from stores that are below 30% utilization, to drive more fulfillment to those underutilized locations. In the case of a $10 shipping cost, that equates to a $4 incentive. However, that incentive will likely not be the only factor the retailer wants to consider. Suppose that store labor utilization factor is weighted at 50% of the incentive and several other factors comprise the remaining 50%. In that case, the incentive would be 50% of the $4 incentive, or a $2 incentive to use that store. The factors comprising the remaining weighting may add to or subtract from the incentive for fulfilling from that particular store. — Return to article text above
7 Manhattan has written these in AngularJS to get a consistent user experience across all of these platforms. — Return to article text above
8 For Manhattan Active WM, Manhattan has been able to leverage all the work they’ve done over the past several years developing automated regression testing . On that test bed, they run about 350,000 different test cases (across their entire product portfolio), twice a day. — Return to article text above
9 service-level agreement — Return to article text above
To view other articles from this issue of the brief, click here.