Each of the successful Networked SaaS1 platforms in the marketplace comes with their own heritage, strengths, weaknesses, and strategy for growth. Many have grown primarily through organic development of functionality. In contrast, E2open has (in addition to their organic growth) become an ‘acquisition machine,’ especially since they were taken private by Insight Venture Partners in 2015. Their strategy has been to methodically build out a continually more comprehensive, end-to-end, integrated suite of supply chain planning, execution and channel management solutions, on top of their interconnected, multi-party network. Significant acquisitions include:
- icon-scm (2013) – dded rapid response planning to E2open’s execution and visibility;
- SERUS (2014) — Visibility into semiconductor and contract manufacturing operations (WIP, yields, non-conformance) and coordination of ECO/BOM changes;
- Terra Technology (2016) — Demand sensing, multi-echelon inventory optimization, transportation forecasting for many of the world’s largest CPG firms;
- Orchestro (2016) — Multi-retailer demand signal repository, harmonizes and analyzes POS data; highly complementary to the Terra acquisition;
- Steelwedge (Q1 2017) — End-to-end S&OP, demand and supply planning, exception management;
- Zyme and Entomo (Q4 2017, Q1 2018) — Channel Data Management, providing channel visibility and demand shaping;
- Cloud Logistics (Q4 2018) — Transportation Management System, including capabilities to sell to small and medium businesses (SMB);
- INTTRA (Q4 2018) — Ocean shipping network, founded by major ocean carriers, providing ocean booking and visibility. About a quarter of global container trade is booked on INTTRA. (We will provide our take on the INTTRA acquisition in the next issue of the brief.)
E2open is maintaining and enhancing the best-of-breed applications they have acquired, even as they integrate them into the platform. Most customers thereby will start with a single application, to solve a specific problem, but with the understanding that they can do much more down the road. That vision is compelling to many companies, even if they are not ready to implement a broader end-to-end suite immediately.
Brands Are E2open’s Primary Customers
At the E2open Leadership Forum, we got to hear from E2open about their future plans, as well as hearing from a number of their customers about some of the innovative things they are doing. E2open’s customers are primarily major brand owners in CPG, high tech, industrial manufacturing, pharmaceutical, automotive, aerospace & defense and oil & gas. The diagram E2open uses to visualize the number of companies connected to their network (Figure 2) shows the brand owner in the center. Thus, E2open’s end-to-end functionality is largely focused on the brand owners, although they provide functionality for suppliers, distributors, retailers, and transportation carriers as well.
E2open’s Roadmap
E2open’s SVP of Product Management & Strategy, Pawan Joshi, presented their upcoming product roadmap, which continues to fill out the end-to-end functionality, as well as improve the integration of the pieces. Pawan said, “Our goal is to bring together all stakeholders from procurement, manufacturing, logistics and execution, demand, supply, inventory planning, and distribution channels, together on one platform.” He talked about the importance of “outside-in thinking,” where companies are driven by demand all the way from the end customer for their supply chain back through multiple tiers of suppliers. Pawan touched on several other themes, including AI-driven automation, goal-oriented orchestrated responses across the network, continuous planning and execution, convergence of planning and execution, and the modularity of their suite.
Pawan used E2open’s marchitecture2 diagram (Figure 2 below) to describe how they support this outside-in thinking. It shows the players in the physical supply chain arranged along the bottom, arranged from supply side on the right to demand side on the left (i.e. material and product flows from right to left). Above that is a ‘digital twin’ of the physical supply chain, represented in the three layers of E2open’s solution stack. E2Net is the foundational layer; an integration platform and connected network of trading partners, which includes all of the players participating in the end-to-end chain. In the middle are the applications. At the top is the common user experience/interface called Harmony.
Figure 2 provides a high-level visualization of how E2open integrates the applications it acquires. Each acquired application is integrated with the other modules in the suite via the E2Net layer on the bottom. Each applications’ user interface is “updated” (refreshed or rewritten, as needed, over time) using the ‘Harmony’ layer on top. This creates a consistent user experience across an integrated set of applications.
The Five Application Categories
Next Pawan described the current and planned functionality for the five application categories in the middle layer — Channel Shaping, Demand Sensing, Business Planning, Collaborative Manufacturing, and Supply Management. The categories are laid out to mirror the flow of data and product across the supply chain and through the enterprise, with demand and channel data driving the S&OP business planning processes. The output of the business planning drives manufacturing plans, which ultimately drives sourcing and supply management.
- Channel Shaping — This includes the ability to gather, cleanse, normalize, and analyze data, including point-of-sale data, to make it “decision-grade” across all channels — retail, distributor/reseller, and online.For distributor/reseller channels this function is typically known as Channel Data Management; and as Demand Signal Repository for retail distribution.This decision-grade data provides visibility to inventory and sell-through across a multi-tier distribution channel, provides incentives and rebate management to stimulate and shape demand and detects and prevents out-of-stocks at the retail shelf. Most of this functionality came from their Zyme, Entomo and Orchestro acquisitions.
- Demand Sensing — This intelligent application suite incorporates the demand sensing technology from Terra to dramatically improve forecast accuracy. It also includes applications that facilitate forecast collaboration with customers, sales order collaboration, and inventory management to align on a joint understanding of Demand with customers.
- Business Planning — Demand planning, supply planning and response, sales & operations planning, multi-echelon inventory optimization, transportation forecasting, transportation management (from Cloud Logistics), and distribution planning are all a part of this unified application suite.
- Collaborative Manufacturing — E2open already provides orchestration across global manufacturing facilities via manufacturing visibility, contract compliance, manufacturing instructions, and quality and traceability. Industries with controlled manufacturing processes like semiconductors, life sciences, food and beverage can manage their partner facilities.
- Supply Management — This helps ensure global continuity of supply for both material and capacity. A number of pieces come together to make this happen, including forecast collaboration, PO collaboration, inventory collaboration, buy/sell management (for contract manufacturers), drop ship management, and invoicing and cost management. The goal is to ensure there are no surprises within manufacturing (inhouse and multi-tier outsourced) by better sharing of data between partners and collaborative processes to preemptively solve problems.
At the core of E2open’s offerings is a permissions model that provides granular security and access control to be independently set on every field within every object. This type of flexibility and fine-grained control is foundational for supporting multi-enterprise collaborative decision-making processes that span across multiple tiers of the supply chain (from raw materials to end consumer) and are powered by live data from across the end-to-end supply chain.
Cloud Logistics
The acquisition of Cloud Logistics builds on an E2open’s existing transportation forecasting and visibility functionality. About eight years ago, E2open implemented integration to TMS systems (like OTM,3) to facilitate bookings and provide visibility to transportation execution status. Through E2net, customers can cross-connect using existing connections to other logistics networks as well, including GT Nexus, CargoSmart, CHAMP, Descartes, and INTTRA (which E2open recently acquired). They pull in not only logistics data but in some cases procurement data from those and other networks. In addition to bookings and shipment visibility, they also do transportation forecasting to plan future capacity needs by lane/mode/class.
Selling to SMBs
The Cloud Logistics acquisition adds TMS execution capabilities, and just as important, the ability to sell to and support SMBs (Small and Medium Businesses). Prior to this acquisition, E2open’s customer base was almost entirely large enterprises. Designing for, selling to, and supporting smaller companies is an entirely different ball game, requiring extreme ease of use and low-touch, largely self-service sales and support processes. Many companies whose legacy is selling to large enterprises have tried and struggled to go down market and sell to SMBs. E2open is betting that they can absorb some of the culture, tools, UX, and methods that Cloud Logistics has successfully used to sell to SMBs.
Control Tower
Pawan demoed their Control Tower functionality, which showed a map view of distribution centers, highlighting shipments that were good (blue), borderline (yellow), or bad (red). The system highlighted in red only those late shipments that were going to have an impact. Pawan showed one shipment that was two weeks late, but was not red, because it would not cause problems due to demand having been over-forecasted for that location. Without that information, a logistics manager might unnecessarily spend money to expedite a shipment.
A second example showed a shipment that was one week late but highlighted for urgent action. Drilling down it became apparent that the problem happened in the future. The drilldown showed a history of all the shipment tracking milestones, where it currently was (on the ocean), and how late it was. He saw that the planned route had an extra stop before the final destination, since that was the most cost-effective route at the time the plan was generated. He was able to modify the truck’s planned route to deliver to the more urgent location first.
Journey to an Autonomous Supply Chain
This type of control tower functionality goes beyond late shipment alerts to highlighting and focusing attention on only those late shipments that will have actual impacts and it prioritizes those impacts. This can only be done by a solution that has access to end-to-end information about orders, manufacturing status, shipments, inventory, and forecasts, across the network of trading partners. E2open is one of only a handful of solution providers that can bring all of those pieces together.
We expect to see increasing focus and competition amongst these platforms towards these holistic and eventually autonomous supply chain capabilities; network platforms that can spot problems across a supply chain, prioritize the impact, make recommended fixes, and ultimately (over the long run) start to independently implement those fixes. This is going to take a few years for any of the players to fully realize. Each has their own strengths and weaknesses. E2open has particularly rich demand data from both distribution and retail channels as well as an accurate supply picture and the latest logistics status. They also have the double edge sword of their strategy and capabilities to acquire all the pieces, which enables them to move more rapidly to build out a complete suite, but with the burden of more integration efforts.
Activity Stream
With the Cloud Logistics acquisition came a capability called Activity Stream. This integrates unstructured person-to-person dialogs into the structured flow of business processes and transactions. Most business processes consist of a structured flow of transactions, as well as a lot of unstructured discussions and negotiations related to those transactions — these are the back and forth dialog between the people involved, via phone, email, text, meetings, and so forth. For example, a PO process may begin with the buyer sending a purchase order and the seller sending back a confirmation. However, behind the scenes, they often have had numerous back-and-forth dialog and negotiations to reach an agreement about the actual delivery dates and quantities. Typically, those interactions are not captured in any system in an easy to find and useful way.
Activity Stream allows these unstructured activities to be integrated in line with the associated business processes. Cloud Logistics saw the need for a way to manage all the back and forth that happens in logistics processes — negotiating contracts, tendering loads, changing schedules, and so forth. They wanted to capture all that in the system. So they added a way for parties to enter request and replies right there in the transaction itself. If you need to expedite a shipment, you can type it into the shipment order. The system also allows the user viewing a transaction to initiate a phone call, email, or text right there in the transaction. With the click of a button, the system sends email or text and records the conversations right into the transaction flow.
This kind of functionality is broadly useful beyond logistics and E2open has started merging Activity Stream into the rest of its suite.
Automating Processes by Learning from Human Decision Making
The real power of Activity Stream is enabling more advanced robotic process automation (RPA), beyond automating the highly repetitive tasks. By connecting “conversations” with processes, the system will be able to learn what decisions were made, why they were made, and observe the actual outcomes and impacts of those decisions. For example, a PO is late, here is what we did in the past that worked under the same circumstances, so this is what the system is recommending. Using AI/ML, the system can start to see the patterns and make recommendations. Eventually, once enough trust is established, the system may be allowed to directly execute some portion of those decisions (perhaps within certain confidence thresholds) without human intervention. This ties closely to the autonomous supply chain concept described above. E2open is already working on this type of hands-free automation.
Managing a Supply Chain That You Don’t Control
One of the most interesting sessions at the E2open Leaders Forum was given by Julia White, Director of GFPVAN4 and Transformation for the Reproductive Health Supplies Coalition (RHSC). The title of the talk was “You Can’t Manage What You Can’t See,” but I was more impressed by how her group was able to influence and direct a loosely coupled collection of participants across the supply chain.
RHSC has no President or CEO. It is a coalition of companies and public agencies that are collaborating to get reproductive health products to the most remote corners of the world, to those who need it. For them, stockouts can literally have life or death impacts. Managing expiration of drugs is highly challenging, especially for temperature sensitive drugs being sent into remote parts of the world with undeveloped logistical infrastructure. So, here you have extremely challenging supply chain problems, being tackled by a highly decentralized loose-knit volunteer organization. I was really curious how they could pull this off.
Created in 2004, RHSC has 475 member organizations, a ‘coalition of the willing,’ consisting of manufacturers, governments, and NGOs (with substantial support from the Gates Foundation). These are all voluntary members, trying to provide affordable quality products to low income people. They link together procurers, manufacturers, shippers, and central medical stores to understand the flow of goods from the perspective of those buying.
Several years ago, they built the RHSC Interchange to track orders and inventory, and detect stockouts. They put in place a group (she called it their ‘fire brigade’) to take care of crises such as expediting shipments as needed. They realized they needed to understand the consumption and inventory in the country, the flow from factories, and how to prevent the stockouts, rather than doing continual firefighting. They created a coordinated supply planning group, but it was constrained by the siloed tools they were using. Julia said they had essentially had to use ‘human control towers’ looking across and aggregating all the data from different systems to spot the problems and make the decisions. Their logistics was a black box — a request is made, with no idea when it will arrive, or how much shelf life is left before expiration.
So, they started two initiatives — the Global VAN (Global Visibility and Analytics Network) and the Country VAN (Country Visibility and Analytics Network). The countries are the linchpin to meeting their goals, so they start upstream with the buyers, linking back to manufacturers, while thinking about in-country distribution.
Governance Structure
RHSC has a steering committee that sets direction, resolves conflicting priorities, and defines scope. It is made up of the main funders USAID, UK’s aid agency, the Gates Foundation, and UNFPA (United Nations Family Planning Agency). They have task forces for process, data sharing policies, data management, and technology. This way users have skin in the game and take ownership — they do the actual work and then it gets raised up for approval. These have helped them drive collective action. Planners in the countries and procurement groups are brought together to look at the data and make recommendations — to see what countries are asking for, what inventory is available where, and making recommendations about what to ship, when, and where.
Data Sharing Policies and Standards
You can’t have a control tower without the data. The various players across the supply chain have to be willing to share it. They are in the process of moving from many bilateral data sharing policies to a single, multi-lateral, role-based EULA (end user license agreement). That way people who have a particular role are only allowed to see certain data. They expect that effort to be ready around December and they think they may be trailblazers with this kind of agreement. In addition, a control tower needs the data to be in a uniform format. RHSC is piloting the use of GS1 standards and trying to convince countries to move towards that.
Collaborative Technology Selection Process
After years of discussion and “stumbling on issues like what is the scope, where do we start, and what do we need the platform to do,” RHSC started a pretty innovative two-phase RFP process. First, with vendors who had expressed interest, they started with six weeks of a collaborative requirement setting. They had weekly calls, shared the same framework and documents with everyone, and worked collaboratively with the vendors to find out what was realistic to accomplish. Then they packaged all the requirements into a 70-page RFP addendum and went through a selection process. The whole thing took eight months and they ultimately selected E2open.
Progress to Go Live
Implementation has kept up a good pace. They loaded the master data in July of this year, implemented inbound visibility in August, inventory visibility in September, and supply planning and manufacturers in October. This month they are implementing GDSN (Global Data Synchronization Network)5 and plan to go live by the end of the month.
Digital Twins from Land O’ Lakes
Yone Dewberry, SVP Chief Supply Chain Officer, Land O’ Lakes, also gave a fascinating talk on their digital transformation. Land O’ Lakes is the third largest member owned cooperative in the US — with 3,500 members and 10,000 employees. Their mission statement is “Feeding human progress.” One of their services is Answer Plot Insights. They have almost 200 test plots across 32 states and 3 countries where they have done about a million different trials, using different seeds and various applications of fertilizers, pesticides, herbicides, and fungicides. This has generated over six million data points, from which they build a digital twin that models every plot and all the test data.
Then each farmer can divide up their own land into squares of whatever resolution they want (typically an acre or so) and Land O’ Lakes collects data from the farmers’ plots as well. Now the farmer can find out which combination of seeds and treatments will provide the best results for any give region, soil type, and conditions. Each segment of their farm is mapped to find its ‘twin,’ for which test data has already been collected. Farmers can track their yields by fine-grained location, and compare it to previous years, as well as how they are performing versus other location.
Land O’ Lakes also provides a disease map, using a combination of satellite imagery and leaf testing to track the development of disease. They can recommend the proper protective treatments. Because they know the field details for each farm, they can show them the most likely places a given disease is going to hit, and thereby provide much more targeted treatment, rather than having to treat the entire farm. This saves the farmer money and has less impact on the environment.
Demand management is challenging as well. Land O’ Lakes needs to predict which fungicide farmers will want, when and where, what seeds they will want and when. Seasonality is a big factor, but the exact date is also based on the weather. If they could reliably predict weather three months in advance, their forecasting job would be a lot easier. They had already been using E2open for forecasting using POS and shipment data. They ran a 10-city pilot where they added weather data as an input into the E2open Demand Sensing application. This improved forecast accuracy 4%-20% across seven categories. They plan to use those improved short-term predictions to better forward-deploy products.
End-to-End is a Journey — a Very Challenging One!
E2open has a compelling vision and strategy to help their customers achieve end-to-end integrated, demand-driven, digital, automated supply chains. The vast majority of companies have a long, long way to go. There are still many siloed systems, manual processes, and issues with data quality. A platform like E2open can let a company start from where they are today, keep most or all of their existing processes and systems, and begin with that one area that needs immediate improvement. Thus, it becomes a key enabler of starting the journey, and step-by-step moving further along the path to the ultimate end-to-end supply chain. As the saying goes, “the journey of a thousand miles begins with a single step.” For some, that step just might be implementing one of E2open’s capabilities
_____________________________
1 A Networked SaaS platform supports multi-party processes, using a shared data architecture to provide a shared single-version-of-the-truth. — Return to article text above
2 A marchitecture diagram is a high-level depiction of the elements and relationships in a system, intended primarily for non-technical audiences and marketing purposes. Marchitecture is a portmanteau of “marketing” and “architecture.” — Return to article text above
3 OTM (Oracle Transportation Management) is used by many of the world’s largest companies to manage their transportation. Its heritage and strength is in managing ground transportation. — Return to article text above
4 Global Family Planning Visibility and Analytics Network. — Return to article text above
5 GDSN is a GS1 standard-based network of interconnected and synchronized data pools, used for sharing detailed product data provided by manufacturers with retailers and others who need that data. — Return to article text above
To view other articles from this issue of the brief, click here.