Will Coupa Inspire Sourcing?

In our last post we asked whether or not Coupa inspired at Inspire, referencing our post from last week that asked if they would inspire and indicated where inspiration might come from, and determined that, for an average Procurement professional, Coupa most definitely inspired. But it’s Sourcing and Procurement. Did Coupa inspire Sourcing?

The short answer is no (but it’s not the full answer). Sourcing wasn’t really addressed beyond the statement that with the acquisition of Trade Extensions, most likely being (re)named Coupa Sourcing Optimization, Coupa now has the ability to do complex events, as complex as an organization requires. And while this is true, it’s not very inspiring to an organization that might not even know why they truly need advanced sourcing.

But then again, as of now, Coupa does not really understand advanced sourcing. Up until now, for a Coupa customer, sourcing has been very simple RFXs and basic auctions, age old sourcing technology that any strategic buyer would not find very interesting. But, fortunately for us, Coupa readily admits this. And, as of now, have not decided when, or even if, they are going to integrate Trade Extensions or Spend 360 into their core platform. (Their expectation is they will integrate what makes sense at the right time, but recognize that these were Power User Applications acquired primarily to support current, and future, power users.)

This is good news. The main reason every acquisition of a (strategic sourcing decision) optimization company has failed to date is, as pointed out in a previous post, because the acquirer believed they understood the technology, could integrate, and take it further (and immediately preceded to do just that). But, as we pointed out in a previous post, the ultimate procurement application is the exact antithesis of the ultimate sourcing application. For starters, one is no UI and the other is a UI so complex current desktop systems cannot support it yet.

Based on the history of acquisitions of advanced technology in our space, the only chance of success is to allow both of these acquisitions to more or less continue business as usual, independent of Coupa’s primary business, until such time as all parties collectively, and in conjunction with their collective user base, decide what integration makes sense, when, and how. (And, fortunately for us, this is exactly what Coupa plans to do.) Moreover, in most companies, the people that do the complex sourcing are not only a small user base but are not the people who do procurement and requisitioning. As a result, effective integration right now only consists of pushing transactional data to Spend360 for opportunity analysis, kicking off a sourcing event in Trade Extensions, and then pushing a selected award into Coupa for contract creation and tracking.

As long as Coupa takes advantage of the new Trade Extensions 6 capabilities to define workflows and UX that are easy to use and consistent with what a Coupa power user might define, then they can create a transition for their more advanced organizations into true sourcing, and, more importantly, the Trade Extensions 6 platform can offer a path for those customers that need a more modern Procurement application. And letting each unit do what it does best will insure that, for the first time in the history of acquisitions of advanced sourcing technologies, all parties will actually thrive.

In other words, by saying and doing essentially nothing at this point with respect to their recent advanced sourcing acquisitions, Coupa is actually doing the best they could actually do because it’s going to take them a while to figure out what they got, what they can do, and how all of these units will work together. As far as the doctor knows, they haven’t even hired a translator yet who speaks advanced optimization, machine learning, econometrics (and risk), and truly simplistic Procurement and can teach all parties a common language. So, even though they’ve historically advanced as fast as possible in Procurement, the fastest way for them to advance beyond is to actually slam on the breaks and start again in first gear. If they do this and maintain this philosophy, at a future Inspire, they will finally inspire sourcing in a way that their Source-to-Pay brethren have not.

Did Coupa Inspire?

Last week we asked if Coupa would inspire at Inspire, especially given that most big companies, especially in the Procurement arena, have not given us much to be inspired about as of late. In fact, most announcements in the Procurement Arena have consisted of new UIs, name changes, and acquisitions — none of which does anything for the end user who needs to do a job day-in-and-day-out and spends much of it fighting with an outdated processes implemented by an even more outdated technology causing them to go prematurely bald as they rip out their follicles one by one.

We said that in this day and age you don’t get inspired unless the technology brings value and a technology doesn’t bring value unless it offers something more than the same-old same-old which could include an inclusive design (that also provides functionality needed by suppliers to serve buyers), a multi-functional application (that supports the organizational stakeholders that Procurement must serve), or a better user experience (which is not just a fancy-smancy UI).

So, to this end, did Coupa Inspire?

Inclusive Design?

Definitely. In this release they’ve made it easier for suppliers and service providers to acknowledge POs, flip to invoices, enter timesheets upon services completion, and respond to buyers. The suppliers can determine whether or not they want dynamic discounting, and what terms they will accept (and not need to opt in to or opt out from buyer offers one-by-one). They can maintain their catalogs and use the collaboration tools. With Coupa’s new release, it’s a more inclusive design.

Multi-Functional Application?

With InvoicePay integration, AP can now pay suppliers in 31 countries and know they are being fully compliant with local regulations. Their inventory functionality makes it easy for office managers responsible for indirect inventory. Their single data store gives Finance visibility into all spend under management. And their new risk scoring functionality gives buyers visibility into potential risky transactions that are being directed at potentially risky suppliers. It certainly is becoming multi-functional.

User Experience?

Coupa has figured out the most important feature of a Procurement UI, and that is, simply put, no UI. Procurement is a tactical function that is focussed entirely on servicing a need in the most efficient and effective manner possible. A user shouldn’t see any more than they need to see, do any more than they need to do, and, most importantly if they don’t need to see or do anything, they shouldn’t need to see or do anything at all. If all that is required is an approval for a requisition that completely satisfies all the requirements, and the approval is only required because the total amount exceeds an amount where all requisitions must be approved, then if the approver is aware the approval request is coming, is aware of what it’s for, and has given verbal approval already, then all she should need to do is press a button in the approval request email or send a yes response to an SMS — no application entry needed. In Coupa, she can do that, and a number of other processes have been simplified as well.

So, did Coupa inspire?

For an average Procurement professional, most definitely yes!

But what about a sophisticated sourcing professional, who has to do demand consolidation, new supplier identification and strategic supplier management, complex negotiation, and sophisticated contract creation? Well, that all depends on their read of what Coupa’s acquisitions mean … come back tomorrow.

Supply Management Technical Difficulty … Part III

A lot of vendors will tell you a lot of what they do is so hard and took thousands of hours of development and that no one else could do it as good or as fast or as flexible when the reality is that much of what they do is easy, mostly available in open source, and can be replicated in modern Business Process Management (BPM) configuration toolkits in a matter of weeks.

So, to help you understand what’s truly hard and, in the spend master’s words, so easy a high school student with an Access database could do it, the doctor is going to bust out his technical chops that include a PhD in computer science (with deep expertise in algorithms, data structures, databases, big data, computational geometry, and optimization), experience in research / architect / technology officer industry roles, and cross-platform experience across pretty much all of the major OSs and implementation languages of choice. Having covered basic sourcing and basic procurement it’s time to move on to Supplier Management.

But first, what is Supplier Management? Supplier Management, depending on the vendor, is defined as the provision of Supplier Information Management, Supplier Performance Management, and/or Supplier Relationship Management. The question is, do either of these areas contain any technical difficulty.

Supplier Information Management

Technical Challenge: NONE

Let’s face it, supplier information management is just data in, data out. Collect the data, push it in the database, run a report, pull it out. It’s just a database with a pre-defined schema and some fancy, optimized, UI for getting the right data to push in and pull out.

Supplier Performance Management

Technical Challenge: NONE

Supplier performance management is two part — performance tracking, done with software, and performance improvement initiatives, identified and managed by humans. The latter can be complex, but since this series is focussed on technical complexity, we will ignore this aspect. As for performance tracking, this is just tracking computed metrics over time. Essentially information management, but focussed on collected performance data and metrics.

Supplier Relationship Management

Technical Challenge: NONE

Supplier relationship management is all about managing the relationship. It’s usually done with collaboration (and collaboration software is not technically challenging), development management (lean, six sigma, and other programs), and innovation management (goal definition, initiative tracking, and workflow). All human challenges, not technical challenges.

But does this mean there are no challenges? Depends whether you are using old definitions or new definitions. A new definition goes beyond the basics and looks to software to guide the future of Supplier Management. And that’s where the challenges come in.

Technical Challenge: Predictive Analytics

Inventory levels, sales, and costs are relatively easy to predict with high accuracy with enough data using a suite of trend algorithms. They’re not always right, but they’re right more often than human “gut” (unless you happen to have a true expert who’s top of her league and been doing it for 20 years, and those are very rare) and that’s all we can expect.

But predicting a market trend is different than predicting supplier performance as performance shifts can result from a variety of factors that include, but aren’t limited to, worker problems (such as union strikes), financial problems (which can happen overnight as the result of a massive launch failure, loss, etc.), raw material shortages (as the result of a mine failure, etc.) and so on.

Thus, predicting future performance requires not only tracking performance, but also external market indicators of a financial, regulatory, and incident nature. The latter is particularly tricky as incidents are the result of events that can often only be detected by monitoring news feeds and applying semantic algorithms to the data to identify incidents that can affect future performance. Then, all of this data needs to be integrated to paint a picture that can more accurately predict performance than the predictions made from just monitoring internal data sources.

In other words, if all you are being sold is a data collection and monitoring tool, it’s not particularly challenging to build (and a business process management / workflow configurator tool could probably be used to build a prototype with your custom requirements in a week), but if it’s a true, modern, performance management solution with integrated predicted analytics to help you identify those relationships at risk, that’s a completely different story.

Next Up: Analytics!

Supply Management Technical Difficulty … Part II

A lot of vendors will tell you a lot of what they do is so hard and took thousands of hours of development and that no one else could do it as good or as fast or as flexible when the reality is that much of what they do is easy, mostly available in open source, and can be replicated in modern Business Process Management (BPM) configuration toolkits in a matter of weeks.

So, to help you understand what’s truly hard and, in the spend master‘s words, so easy a high school student with an Access database could do it, the doctor is going to bust out his technical chops that include a PhD in computer science (with deep expertise in algorithms, data structures, databases, big data, computational geometry, and optimization), experience in research / architect / technology officer industry roles, and cross-platform experience across pretty much all of the major OSs and implementation languages of choice. We’ll take it area by area in this series. In our first post we tackled standard e-Sourcing, and in this post we’re tackling standard e-procurement.

Requisition, Approval, and Purchase Order Management

Technical Challenge: NOTHING

There’s nothing challenging about creating a requisition, placing it in a, possibly bifurcating and reconnecting, approval stream, getting approvals, and flipping it into a purchase order. It’s literally just adding lines and data to a form, like building a survey or RFX, recording approvals, and generating a purchase order in an appropriate distribution format when the necessary (final) approval(s) have been generated.

Invoice Management

Technical Challenge 1: Automated Error Correction

It’s easy to create and distribute an invoice. It’s easy to run a set of verification rules to verify completeness and correctness and then reject an invoice if data is missing, incomplete, or invalid. It’s harder to determine when data is missing (such as codes, skus, etc.) what that data should be, harder still to figure out which data is likely correct when there is a mismatch between fields that should align, and even harder when data is incomplete and suggests multiple possibilities. The goal should be to not only determine when there are issues with an invoice and flip it back to a supplier for correction (to reduce the number of invoices that need to be manually reviewed and approved) from an average of 15%+ to 1.5%+, but to indicate what the acceptable corrections are / should be so that the supplier can accept and the invoice can be automatically accepted and processed on re-submit. This requires strong AR (Automated Reasoning) technology and it is not easy to not only identify 90% + of the bad data, but 90% + of the correct data to replace the bad / non-existent data with.

Payment Management

Technical Challenge: Working Capital Optimization with Multiple Options

While ACH integration can be a challenge because of the security requirements, it’s not that difficult (as the banks / payment providers did the challenging task of implementing the encryption, secure networks, etc.) and the vendor just needs to plug in, it’s just coding hoops and a well understood process. The challenge is how to optimize the payment schedule against net terms (to prevent penalties), early payment discount options (when it is cheaper to take the discount offered even if the organization has to pay interest at their preferred credit rate), co-factoring (where the organization helps the supplier factor the invoice and agrees to take an early payment cut to cover some of the supplier’s cost of factoring), and investment opportunities to make sure the organization has the cash on hand it needs while minimizing its supply management costs.

Taxation Management

Technical Challenge: NOTHING

While it’s the ultimate pain-in-the-backside to keep up with all of the requirements associated with tax-tracking across multi-level jurisdictions when taxes can be applied at the union, country, state, and city level, especially when the amounts, collection rules, and submission rules are always changing, it’s just data tracking. Nothing more.

IN SUMMARY

With the exception of automated error identification and automated corrective suggestions and of working capital optimization, as with basic e-Sourcing, basic e-Procurement is pretty much common fare today that can be bought off the shelf from dozens (and dozens) of providers, but, as you can see, it’s not all equal. Any provider with AR capabilities for advanced invoice processing and working capital optimization capabilities is leagues ahead of anyone else.

And, as per part I, in this series we’re not discussing the User Experience. While a good User Experience, while not always challenging to code, can be challenging to define, it doesn’t define Technical Difficulty on its own.

Next Up: Supplier Management!

Supply Management Technical Difficulty … Part I

A lot of vendors will tell you a lot of what they do is so hard and took thousands of hours of development and that no one else could do it as good or as fast or as flexible when the reality is that much of what they do is easy, mostly available in open source, and can be replicated in modern Business Process Management (BPM) configuration toolkits in a matter of weeks.

So, to help you understand what’s truly hard and, in the spend master‘s words, so easy a high school student with an Access database could do it, the doctor is going to bust out his technical chops that include a PhD in computer science (with deep expertise in algorithms, data structures, databases, big data, computational geometry, and optimization), experience in research / architect / technology officer industry roles, and cross-platform experience across pretty much all of the major OSs and implementation languages of choice. We’ll take it area by area in this series.

E-SOURCING – RFX

Technical Challenge: NOTHING!

I’m going to burst a lot of bubbles here, but there’s nothing technically sophisticated about the development and implementation of an RFX solution by any stretch of the imagination. In this day of age, one could pretty much implement a basic RFX application in HTML5, javascript, and MySQL with a bunch of open source libraries in a couple of days. Form elements, templates, basic branching workflow, weighting, etc. … you even see most of this in free survey tools. Even bulk file upload is just naming conventions. ( But it’s amazing how many vendors haven’t even figured this out. 🙁 )

E-SOURCING – e-Auction

Technical Challenge: NOTHING!

Again, more bubbles bursting by the dozens. Twenty years ago, implementing an e-Auction was a real challenge with relatively simple web technology, slow internet speeds, and lack of graphical frameworks that could be updated in real time. But today, there’s a host of cheap / freemium solutions that implement basic e-Auction functionality. Unless the vendor is tying in with an optimization solution in real time to create a optimization-backed auction solution, no reason you should pay much for these dime-a-dozen solutions.

E-SOURCING: Optimization

Technical Challenge 1: MODELLING

This comes in two forms.

1. Structured for “Easy” Definition

Creating one or more templates that allow a user to quickly define the entities of interest — suppliers, products, services, locations, lanes, etc., collect the bids, define the constraints, and solve unconstrained, and maybe default constrained scenarios (3 suppliers, geographic split, etc.). Making what’s hard easy is no easy task.

2. Free-Form for Custom Models

Not every model the organization will need to create will fit in a template — especially if the organization wants to optimize working capital, minimize risk, cross-optimize related categories, etc. This requires the ability to allow end users to define the models they want using a modelling interface, which will not be easy to build because how do you hand over all the power but still make it understandable by a non-programmer / non-mathematician.

Technical Challenge 2: SOLVING

It’s hard to build these models, but it’s much harder to solve them. First of all, you have to map them to a system of equations that can be solved by your, hopefully, mixed integer linear programming solver (as you want to use a solver that is mathematically sound and complete), optimize them, optimize the solver settings, and hope that everything was translated consistently and there are no conflicting or unsatisfiable constraints and the model can be solved in a reasonable amount of time. Given that solution time grows exponentially with model size, this can be quite a challenge even for moderate sized models.

E-SOURCING – Contract Management

Technical Challenge 1: Contract Analytics

A simple contract management application, which is nothing more than meta-data based contract indexing and tracking, can literally be built by a high-school student with an Access database and go head-to-head with most of the basic contract management modules out there today! In fact, most of the capabilities of most contract management modules are pretty simplistic and can be built in a matter of days with a good BPM configurator (and companies like Agiloft have done it). The exception is contract analytics (like that provided by the likes of Seal Software).

Using semantic analysis to figure out what contracts contain clauses of a certain type, what contracts are missing clauses that pertain to a certain regulation, and whether a certain clause is close enough to a required / suggested clause is not easy. Not easy at all! Semantic technology is still emerging, and trying to capture a user’s wants, even given a set of sample clauses, is quite computationally difficult!

IN SUMMARY

With the exception of decision optimization and contract analytics, baseline e-Sourcing is pretty much common fare today that can be bought off the shelf from dozens (and dozens) of providers, but, as you can see, it’s not all equal — any provider with true decision optimization or true contract analytics is leagues ahead of anyone else.

And, of course, in this series, we’re not discussing the User Experience, and in some cases, a good User Experience, while not always challenging to code, can be
very challenging to define.

Next up, baseline e-Procurement!