deal architect

A perspective on the SAP Capital Markets Day

I am a customer and market analyst and tend to not attend too many earnings calls or Wall Street presentations. But I made time to listen to the SAP presentation yesterday. It was a coming out event for the 2 new co-CEOs, Jen Morgan and Christian Klein and there has been much speculation (with a capital S in my opinion, not fact based) about the activist investor impact on SAP trajectory. The full set of presentations is in the video here - without the breaks its about 3 hours long. The theme was "Growth and Operational Excellence" and I liked quite a bit of what I saw and heard.

To me, growth comes from market differentiation and here are key areas where SAP executives scored well

Momentum in Asia

Both Jen and Adaire Fox-Martin made repeated points about Asia/Pac. SAP is very well positioned in that region as I wrote a couple of months ago "SAP has been in Asia over 30 years and now has over 9,000 employees spread across 31 office in the region. The event in Bangkok showed me just how far the competition has to go to catch up."

Cross-selling

I always shake my head at how little cross-selling has taken place across SAP's product portfolio (listen to Jen starting around 0.16 and Christian around 0.34 about inter-dependence and synergy across SAP products). Most of their competitors can only play in one or two categories.  I am seeing many customers look at application rationalization, and reverse a decade of "ring fencing" their ERP with best of breed cloud applications. As SAP better breaks down its many siloes, starts to show more coherent road maps and delivers better integration, its broad suite should open up many new opportunities within its incumbent customers. 

S/4 adoption

SAP is associated with large, multi-national customers, but the CIO of Chobani, the yogurt company, Parag Agrawal (starting 1:29) was a good representative for the mid-sized, fast growing market that S/4 is doing well in. 40% of S/4 adopters are net new to SAP. The ECC to S/4 conversion market is also gathering steam. As Adaire says, the question for most customers is not 'if" but "when" "why" and "how". As I described in SAP Nation 3.0 - "tilting its bell curve of customers" is its largest single opportunity.

XM and CX

Qualtrics is still a missionary sell to many operationally focused ERP buyers but Ryan Smith humorously reminded the audience that precisely a year ago, he was presenting in a road show which valued the company at nearly $ 5 billion, and that now, as part of SAP, the deal size and customer growth has accelerated. It is another differentiation for SAP. Similarly, SAP has had fits and starts when it comes to the CRM market. But with all the musical chairs in the executive suite at many CRM vendors in the last few weeks you get the sense that market is poised for disruption. SAP's CX portfolio, especially the Commerce piece, and with new leadership under Bob Stutz, will get another shot.    

The Microsoft "embrace"

Microsoft and SAP have been 'frenemies" for decades. At Gartner and since I have written about many joint initiatives like Mondecino and Duet (nearly 15 years ago). The collaboration around Azure opens up support of the vast Microsoft channel, and also simplifies the hyperscaler choice for many undecided customers (while keeping open support for customer loads on AWS, Google and Alibaba clouds). The cloud version of S/4 and SCP should in particular benefit.   

Turning to Operational Excellence, there was quite a bit of focus on margin and cash flow improvement.

Christian has led a digital transformation within SAP for a couple of years and it is accelerating. It's the tagline of "Best Run Business" applied to SAP itself. His points, starting around 0.38, lay out the high level agenda.  Adaire described some of the digital and AI impact on the sales function around 1:26. Thomas Saueressig, starting at 1:09 talks about changes in the development organization. He also addressed the thorny issue of integration, especially across the slew of acquisitions in the last decade,  that customers constantly complain about. In his previous role as global CIO at SAP, he is acutely aware of the pain points (It's also something Bob Stutz in the CX area will have to accelerate). He also touched on centralization of infrastructure. which CFO Luka Mucic quantified (starting at 1:40) as they move from 25 mostly acquired technology stacks to a converged one, and also shifting many cloud loads to the 4 hyperscaler partners.

Luka also touched on improvement of cloud gross margins and marketing and SG&A efficiencies - consolidation of many sponsorships, supplier base among others.

Three areas I would like to understand more:

Margins at the expense of customers?

There is plenty of fat to be removed from within SAP and that should deliver the improved margins SAP is promising to investors. But the temptation in our industry is also to milk the customer (Oracle and the fast growing SaaS vendors are not much better). In the last few years, we have seen quite a backlash to Indirect Access and other SAP audits. I am also seeing SAP propose license conversion fees to move from on-prem to cloud versions. Customers are willing to pay for new functionality, not so much for what they have already paid years in maintenance. So, trying to extract more from older function points, which frankly have been commoditized over the last couple of decades, will only slow down incumbent customer traction.

Focus on total cost to customers?

Historically, SAP has not been very good at benchmarking and managing its partners. That led to my quantifying the huge "GDP" of SAP Nation. As we look at moves from ECC to S/4, I am starting to again see scary SI proposals. Labor intensive and not much automation, with the suspect quality and economics those threaten. Now with the 4 hyperscalers customers could face significant additional exposure. SAP likes to brag it has 000s of partners - without pointing out that cost is borne by customers. I would like to see more SAP detail on how it will better manage its ecosystem, and accelerate its automation and application of ML to its many of the routine conversion, testing, training and other tasks - see more ideas here. Adaire did mention conversion factories - like to hear more about that.

Not enough R&D or Capex to grow the functional footprint?

There was surprisingly little on industry functionality. SAP keeps promising its partners are working on it. For the next few years, it appears there will be little excess R&D or capex for acquisitions. While SAP needs that discipline in the short term, markets don't wait for ever. Beyond verticals, as I wrote earlier this week "the definition of the "enterprise" continues to rapidly expand and morph, and I see a very fragmented market for the next few years". SAP says it has a shot at a TAM of $ 500 billion by 2023. In my opinion, they could already be aiming for a trillion.

Overall, there was much to like in the 3 hours. As an analyst, it is my job to keep pushing vendors to do more for customers, but realize the new executives already have a very full plate to tackle. Should be a very exciting next few years. It is a great way to kick off a new decade!

I will see Josh's five and raise him ten!

Josh Greenbaum has a fascinating post - how Microsoft and SAP plan to take over the enterprise software market. There is much in there I agree with. All the recent executive changes - Jen, Christian, Bill, Bob - are exciting for the industry. Microsoft's infrastructure cloud is on a roll - the SAP "Embrace" and the DoD JEDI contract only add to the momentum. But I see much bigger changes in the enterprise software market beyond what SAP and Microsoft will touch.

To me, Google, Facebook, Amazon and agencies like Accenture with their share of digital advertising and CMO access will continue to significantly influence and morph the CRM space

The customer service segment will be significantly reshaped as IoT, satellite, drone based concepts mature. I expect industrial asset makers like Siemens and Caterpillar and logistics providers like UPS and Amazon to also be major players.

Contract manufacturers like Foxconn and Flex and boutique design firms will have more than their fair share of influence in the product engineering and manufacturing areas.

Amazon, Alibaba and other big commerce sites will continue to commoditize fulfillment functionality for countless SMEs and set a bar for broader supply chain and logistics functionality.

Robotics and other automation vendors will reshape many a business process. Similarly, Google will help many an enterprises rethink processes with machine learning.

A whole bunch of customers are themselves reshaping their industry specific operational systems with custom development and specialist startups - they are tired of waiting for modernized vertical packaged software.

Major economies like China, Japan, Brazil and others have seen so far only modest traction with modern enterprise applications. I have a feeling regional players will have a bigger role in the next few years.  

Josh has a lot of ink about activist investors. To me, they play a vital role in shaking up status quo, but smart as they are, I dont see them as significantly reshaping technology markets. In later, more mature markets, private equity investors have more influence in consolidations and other seismic changes. But we are not at the stage in the current iteration of enterprise software.

Net-net to me : SAP and Microsoft will continue to do well in traditional transactional areas. But the definition of the "enterprise" continues to rapidly expand, and I see a very fragmented market for the next few years.     

Of course, Josh may well be right. My little bet is my ticket to what will be a wild ride. Cannot wait for the 2020s!

A vertical that is growing...well, like a weed

I have been imploring software vendors to globalize and verticalize their applications for a while now - after two decades of cloud apps, less than 20% of the grid of applications across countries and operational processes is filled. I have been increasingly writing about fintech, shop floor, healthcare and other industry functionality. But I wasn't sure when Tom Brennan, CMO of the Salesforce platform-based ERP vendor, Rootstock wanted to discuss what he calls the "Cannabusiness" vertical. I have known Tom for over three decades going back to Coda and more recently, Financial Force. He is a straight shooter, but all kinds of word associations crossed my mind as I thought about roots and weeds.

As background, cannabis legalization has been spreading across North America – 10 states plus Washington, D.C., have legalized recreational marijuana over the last few years, and full legalization came to Canada in October 2018. That made Canada the first G7 nation to do so.  Tom adds "Then you have the hemp side of the market – such as CBD (cannabidiol) oils, which claim to have many health benefits." In many other parts of the world, irrespective of the legal status, enforcement of cannabis and hemp is increasingly lax. While this inconsistency causes nightmares for HR groups in terms of employee drug testing and compliance and for law enforcement, the industry itself is growing rapidly.

According to Rootstock "In just the past three years, industry employment has grown 110%. That’s faster than the ten-year projections for the trendy sectors of solar, wind power, and health care. And it’s why the industry was estimated at $10.4 billion in 2018, accounting for more than 250,000 jobs, and is slated to grow to $21 billion by 2021."

The challenge for the industry is less about selling, much more about quality and consistency. While there are clear exceptions like their professionally run customer, Aphria described later, for many companies in the vertical, the need is one of differentiation earned through reliability.

Tom says "many companies that have entered this market must consider state-by-state regulations. A lot of them implemented entry-level systems or used spreadsheets to manage orders and inventory. But with a booming market, they’re now facing significant challenges around operations, inventory management, compliance, and keeping up to date with changing regulations. Many found themselves in a tough spot."

Rootstock partners with another AppExchange vendor, ComplianceQuest. I had a chance to catch up with Prashanth Rajendran, CEO and Shalini Chowdhary, VP of Marketing and Strategic Alliances.  They have long offered quality, inspection, compliance and risk management to industries like life sciences, cpg, and aerospace. They recently raised a $36 Million Series A round led by Insight Partners. 

The combination with Rootstock offers "seed to sale to service" to the Cannabis industry as the schematic below shows (click on image to enlarge)

According to Tom, the flow is "companies take seeds and cultivate them. And then they have to perform testing and QA to measure potency and ensure consistency in their product. They want to process orders efficiently and compete on both the service and quality fronts."

"On the CBD and recreational fronts, after they’ve packaged their products, they’re fulfilling orders to retailers or selling it through their own store fronts or websites. And all cannabis companies – both recreational and medicinal – have to be able to track incidents, in which customers may have had an adverse reaction. This is a huge challenge, and lot and traceability features can help."

Recent scrutiny of the vaping market is drawing even more interest in quality management and traceability.

"We have track and trace capability where you can drill down from a CRM system. Companies will have the service case. They have the serial number or the lot number and can track it in the ERP system. They can see if it was tested in the ComplianceQuest system. Then figure out, do you need to pull the whole lot? "

Many states are deploying tools like Metrc to track and trace cannabis inventories and products to origin when there are health issues. Besides the compliance issues, there is the inventory shrinkage issue. As Tom says "Companies have cannabis products flying out of their warehouses at a fast pace. When issues arise, they need to be able to track back to where those products came from, from which suppliers and what type of testing was done."

Tom describes some of the customers in this vertical. "NuLeaf is based in Colorado. The company makes CBD oils from hemp. KIVA Connections makes gummies and chocolate bars with THC (tetrahydrocannabinol). Michigan Pure Med is going after the medicinal space and they consider themselves a pharma company"

But their most prominent customer is Aphria which is listed on both the Toronto and New York Stock Exchange, and is one of the largest cannabis companies in the world. In Canada, cannabis is legal for both recreational and medicinal purposes.

Aphria hopes to benefit in a number of ways from the Rootstock/ComplianceQuest functionality and the Salesforce platform including:

• Competitive advantage over other cannabis producers. This gives Aphria an advantage in time-to-market, license approvals and external audit readiness.
• Fast and flexible compliance in a quickly changing, regulated Industry.
• Demonstrable compliance with Health Canada and GMP/GxP Regulations.

Aphria is bringing a professional image to the industry and as the video below shows their "manufacturing" process is increasingly automated. The more this industry matures, the more it should also help the vendors position for other process industry and other regulated verticals.   

Workfront - looking for work productivity in all the right places

Alex Shootman, CEO of Workfront, is versatile on many fronts. He has traveled more of the world than his tech peers. He has a family with adopted kids which deserves to be profiled in a movie. He is very good with financial numbers. But, to me, most impressively he knows the application landscape at many (most?) of his 3,000+ customers.

Workfront positions itself as the "operational book of record" for work. While many tools like Jira aim to provide visibility and productivity across IT programs and projects, Workfront aims much more broadly across the enterprise.

Alex proudly shows slides of customer application landscapes to investors and anyone else interested.  He told me one of his challenges is to keep scaling that customer intimacy culture. For now, the customer-centric gene runs through the organization. As he described in his book Done Right, one of his "laws of leadership” is "What the leader is interested in, people will be fascinated by".

His customers, in turn, say how much they like working with Workfront. Sue Fellows, Chief Customer Officer, interviewed a few of them during the Analyst Day this week that I was invited to. They had stayed on after their Customer Advisory Meeting earlier in the week and joined us for dinner and most of the day. (Next year, I have requested that they consider allowing a few analysts to observe and participate in the entire Advisory meeting.)

Brent Rudewick described how he started in Marketing Operations at Charles Schwab and since has had Risk Management added to his title, and how his group has steadily grown. His Workfront experience has attracted attention of his Investor Relations, Vendor Management and Training groups. Paul Tasker at Sage Software described the role of Workfront as they re-architected and launched a new web site across 25 countries. It involved global, regional and local coordination between Sage teams and those of external agencies.  Think of the coordination and quality assurance challenge. Angela Lapovsky of Lincoln Financial described how they go to market with their annuities, life insurance and other products almost entirely through channels like brokers. Workfront helps with that extended ecosystem and complying with the regulatory issues they are subject to. Since all these executives had marketing roles, I asked if that is Workfront's sweet spot. Alex told me if I had been in the Advisory meeting I would have met many more from other disciplines. When I was first introduced to Workfront they told me they had a "Land and Expand" strategy. Most vendors say that, but I have seen that now in many Workfront customers - they start in a pilot area and in a few years have been adopted widely. 

I had previously seen Steve ZoBell, the CTO present a case study on new product development. It wove together several creative groups in medical product design and product engineering . In the past, an idea for a bionic knee may have been noted but stayed invisible in the CRM system. Each product area would begin design anew. Workfront has changed that.

These buying centers - Marketing Operations, New Product Development, and various corporate shared services - all are increasingly digitized and tend to have a complex web of projects, and most have a high degree of quality and compliance activities to worry about. These buying centers also don't get much love from large technology vendors.

It's not coincidental that we are also seeing growing interest in categories like Robotic Process Automation (RPA) and investments like Salesforce has made in Mulesoft. There are plenty of inefficiencies in enterprise technology "junctions".  

Digital tools like email, instant messages and social media tend to distract workers even more in these hubs. In their annual State of Work survey, Workfront reports "Over the six years we have published the State of Work report, we’ve consistently found knowledge workers on average spend just 40% of their work week on the job they were hired to do." Get your own copy - it's well done.

As Alex says in our digital world "seamwork" is even more important than "teamwork". We heard the term "citizen integration' early and often during the day. They are facilitated by over 150 pre-built integrations Workfront delivers.

For now, Workfront has its hands full in white collar work centers to worry about shop floors, emergency rooms and other operational work areas. Plenty of productivity to deliver in white collar work.

I changed planes at Delta’s Atlanta hub 3 times this week (I love that place, not!). The crowds and the weather and other delays at such hubs are a good metaphor for similar work concentration and inefficiencies at the intersections of enterprise applications. Providing visibility and productivity at these hubs is what Workfront excels at.

SAP Nation 4.0?

I sent Bill McDermott a “please stay in touch” note last night, and sure enough a couple of hours later, even on what must have been an emotional day for him, I got a nice note back. I have always admired his polite grooming and his always-positive vibe.

 

As an analyst, though I have to unemotionally look at technology vendors from a customer pov and my trilogy of SAP Nation books provide a nice summary of McDermott’s legacy at SAP.

On Wednesday, before I knew of his decision to resign, I had the opportunity to summarize that at an ASUG Executive Exchange event. On a panel with CEO Geoff Scott I laid out what I have liked about SAP in the last few years

A) It has been a product launch machine, and I encouraged the audience to not just look at S/4HANA, but also C/4, Leonardo ML and IoT, Qualtrics, IBP, AIN, SCP and other newer products.
B) I told them I liked what Bill and Hasso have done to make “younger” the SAP leadership with Christian, Jennifer, Juergen, Alex, Ryan, Jared and many others.  I suggested the audience similarly think about their recruiting profiles to attract younger tech talent and make that a consideration and not wait till 2025 to consider migrating from the ABAP world
C) I told them Bill, in particular, had made SAP a lot more open and humble. I certainly saw that in writing volume 3.0

I also told them what I would like to see SAP deliver and what will now become Jennifer and Christian’s legacy

A) I want to see SAP become much more serious about the public cloud. I have told Christian I worry that in a few years we will have 000s of single tenant S/4 installations. That will only replicate the implement, upgrade, AMS economics of the previous generation that led to my writing volume 1 which was not flattering to SAP. It will also prevent SAP from gathering substantive data in the cloud for effective cross-customer ML scenarios. SAP has always been better at developing code more than supporting it, so it will take a sea change for SAP to build and importantly scale its public cloud at demanding SLAs. It does have pockets of multi-tenancy experience with BYD and its cloud properties but needs to expand that. 
B) SAP’s “digital core” is already the widest in the industry compared to that of its competitors. However, even that is tiny compared to the industry footprint of the customers it serves. SAP needs to deliver vertical functionality - every industry has gone through seismic changes in the last 15-20 years - and it needs to step up not keep looking for partners to help deliver industry specific operational functionality that its customers have been paying it maintenance for all these years.
C) SAP has to significantly prioritize its customers above its partners. It has to push its SI and outsourcing partners to automate their operations or even cannibalize them with its own IP. Additionally, the market pendulum is swinging back to build technology from the last few decades of buying. SAP needs to view its hyperscale partners - Amazon, Microsoft, Google among them - not just as infrastructure but as platform plays. They have to be viewed as co-optetion, not just as partners and it needs to make its own cloud platform and other build capabilities much more vibrant. 

On the ASUG panel, I was asked when I would write SAP Nation 4.0. Not anytime soon I said. McDermott’s move has me rethinking, but I really want to give Jennifer and Christian time to make their mark. 

 

I look forward to their success, and in the meantime I promise to be tough but fair in my blogs and importantly, in the advice I provide to customers. 

 

Exciting times ahead. 

A tumultuous couple of decades

As I was filing our quarterly Form 941 last week, it hit me that we have entered the last quarter of this decade. As I flew into LaGuardia on Monday, it took me back two decades to my time at Gartner. In October 1999, there was nervousness about the damage Y2K could cause, but also euphoria at the phenomenal few years many tech vendors had enjoyed. The introduction of the euro promised another bonanza for the near future. Enterprise vendors, particularly two sectors that I followed - enterprise applications and service providers - had reason to feel very good about themselves.

In New York this week I heard an optimistic Mike Salvino , the new President and CEO of DXC, at a HfS event. I heard similarly optimistic (for different reasons) Cognizant executives at one of their events the day before - they also have a new CEO. I also kept up with the celebratory mood at HR Tech in Vegas virtually.

At Gartner, I had followed SAP, Oracle, PeopleSoft and their SI partners. CSC, EDS and HP Services were players back then.  They continue to live on in the form of DXC, founded a couple of years ago. Cognizant had revenues of less than $ 100 million in 1999. It has grown like a rocket with over $16 billion in revenues now. HR Tech was only 2 years old in 1999. But the backwaters of the HR function had led the client/server applications market. This time around, it has similarly led the cloud applications market.   

There is lots to appreciate and applaud.

However, the even bigger lesson of the last couple of decades is the many missed opportunities. Outsourcers missed out on the cloud infrastructure market. It was theirs to lose and so far, players from left field like Amazon and Microsoft have benefited. After 20 years of cloud applications (Salesforce was born in 1999), application vendors can barely cover 20% of the country/industry grid. At the HfS event last week, the biggest attention was on robotic process automation (RPA). Hardly any talk of Amazon's Kiva warehouse robots, or UPS's drones or other operational automation. Or automating the labor intensive outsourcing space. CRM vendors completely missed out on the digital advertising market. Google, Facebook, and digital agencies have made out like bandits in the vacuum. SaaS vendors had an amazing business model which bundled multiple contracts - the license, the maintenance, the hosting, the application management - into one. Then they got greedy like their on-premise competitors. I saw Bruce Rogow, my former Gartner colleague in New York and he told me he sees in his travels CIOs struggle with new opex/capex funding paradigms. Instead of delivering a knock out punch, SaaS vendors mostly co-exist with their on-prem competitors. Outsourcers saw every one of their clients develop smart products and let contractor manufacturers like Foxconn and Flex dominate that market. They have let GE, ABB and others generate excitement about the IoT space.

Vendors ask me, what's next? Blockchain, machine learning, quantum computing, sure. But much bigger opportunities will persist from those missed in the last two decades. A new generation of vendors will move in. Vendors will cannibalize each other. And here is the most important nugget. Customers learn from their mistakes - I am seeing that in the new wave of SAP related services. In writing my recent book, I saw much more caution about using SIs as they implement S/4HANA. There is a "fool me twice, shame on me" vibe.

So, forget about doing business as usual. The best thing every vendor can do as we enter a new decade is to be introspective, even if Wall Street is pleased with your performance. Ask the tough questions - how the hell did we miss as many markets as we did? And how do we re-organize to make sure and not miss so many markets again?  

Exciting times ahead.

TAMyopia

When I was at PwC in the early 90s, I remember we used to throw out statements like "we have 40 to 50% market share in the SAP (or PeopleSoft) implementation market" When I got to Gartner I saw their competitors were saying similar stuff. So, I started telling them - "no, actually you have less than 10%" . The tune of at least some of them turned to "we have 50% market share of deals we choose to chase". Choice - certainly their prerogative which segments or verticals they chose to focus on. But it's not a reflection of Total Addressable Market.

One of my favorite advisory projects is helping vendors challenge their previous definitions of TAM.

I have seen plenty of dysfunctional behavior from technology vendors of every shape when their TAM definition is narrow  - salespeople chasing deals convenient to them, not what the leadership would have approved, marketing claims to deliver functionality in areas which their R&D had no intention of developing in any meaningful manner, CFOs and investment committees which don't approve enough R&D and Capex needed to keep their entities competitive.

In my recent book, I outlined several trillion dollar markets that many technology vendors have missed since the late 90s. Those included smart products, digital marketing, IoT, cloud infrastructure, various verticals - - Google, Facebook, Amazon, Foxconn, GE, Siemens and a whole bunch of vertical players have thrived in those markets even as many technology vendors ignored them. Download the chapter from my recent book  in link here if you want to read about all the missed markets.

Many of those markets were brand new ones or what in MBA school we called adjacent markets so you can understand why so many vendors ignored them - even as they now wish they had entered those markets. But even in their core focus markets, I have seen many a myopic definition of TAM. Here are a few examples

a) Automation and Outsourcing

When I wrote Silicon Collar three years ago, I chose the title because it felt like traditional definitions of white, blue, pink collars were dated as technology was reshaping every job, everywhere. We all wear a Silicon collar these days.  I had examples of robotics, drones, machine learning, CGI, exoskeletons, robotic process automation (RPA) and other automation in the oil patch, in accounting, in films, in winemaking, garbage collection and many other sectors. I wasn't dreaming this up - executives at sophisticated companies were in the book describing their experiences.

So, it has been a bit disconcerting to see so many outsourcers have glommed onto RPA as their major (in many cases only) focus when it comes to automation. Even more cynically, most outsourcers do not take seriously repeated customer expectations to automate their own labor intensive services.  

b) Data Centers and networks

I cannot believe how envious many vendors are of AWS's and Azure's recent success. On the other hand, I cringe when I hear software vendors and many outsourcers say "glad we didn't have to invest billions in Capex". The reality is less than a decade ago HP(with EDS), Oracle (with Sun), IBM, Accenture, Verizon, Deutsche Telecom and other telcos populated and ran data centers and networks for the world. The emerging cloud infrastructure market was theirs to lose and they have done so - spectacularly, so far.

Many SaaS vendors view the cloud infrastructure market as a bonanza - let someone invest the Capex, we will just rent it for cheap. It's not really that cheap and increasingly they are being expected to support multiple clouds. Zoho has continued to invest in its own data centers and passing along those savings to customers. The time will come for other software vendors to similarly hedge their bets. They will have to expand, or at least, morph their market definition of infrastructure.

c) "Glo-ver" in SaaS

I have increasingly been pointing out a huge under-achievement in the enterprise applications market. After two decades of cloud and in-memory applications, only about 20% of a grid across countries and industry processes has a decent choice in contemporary architecture. Myths persist such as "verticals are a sink hole". Sure if you dont do it right, but go see how much lock-in revenue and margins SAP has enjoyed with its vertical investments in the early 2000s.

I have accelerated my coverage of applications and executives who are focused on the white spaces - see some of that coverage summarized here in a trend I call Glo-ver  expansion in focus.

Back to helping vendors challenge their previous definitions of TAM.  Those are fun projects when the vendor is willing to open up. I must be honest though - many vendors look to influence analysts, not use them much for intelligence. So, they would rather posture they have "50% market share in deals they choose to chase".  

Seriously, I hope that changes. Too many vendors are defining markets too narrowly and finding their competitors have done the same and a result too many are chasing after the same, narrow markets. That is a sure path to commoditization.  Trust me - I have seen it repeated many a times over the last three decades. There is little safety in those numbers.

Zoho One evolves at rapid clip

Earlier this year, I wrote about Zoho One " ... the bundle priced at $30 a month an employee, or $75 a month on per-user basis. It was launched in 2017 with 35 applications, it now has over 40.  And from the presentations we heard more apps will join the bundle. A dollar a day per employee is a compelling value proposition, when the functionality keeps growing exponentially. "

I compared it to Amazon Prime "The more I heard of Zoho One and the growing apps coverage for the same price, I thought of how Amazon Prime has grown in coverage. It started off a free 2 day shipping, then added free movies and since then countless new features." Prime, of course, has proven to be a rocket ship for Amazon.

This week, Zoho provided an update on One. It's now up to 45 applications. They told me in a side conversation they have 50+ apps right now....just not all announced.  "Some are on alpha programs, and we really do test them before we formally launch them or add them to Zoho One!"

Features keep evolving including its workflow management tool, Orchestly, which lets managers and administrators design and run workflows without the need for coding skills and PhoneBridge, its telephony platform, which integrates over 50 telephony vendors on one side and several Zoho applications on the other side.

When you look at services under the cover, like in the "AI" category, you get a sense for the growing richness of the applications.

The tools in the bundle keep evolving, many aimed at end users which SMEs with small IT staff especially appreciate.

Bear in mind, many vendors promised they would take savings from taking advantage of cloud infrastructure at vendors like Amazon and invest that into application functionality, and either have not done so, or are just not productive. Zoho continues to rapidly grow its application footprint even as it grows its proprietary DC network. They have shown they can run their DCs more efficiently than going with public cloud infrastructure.

No wonder the bundle is popular. There are now 20,000 businesses using Zoho One, up from 12,000 last year. According to Zoho, a quarter of Zoho One customers use more than 25 applications and more than 50 percent use 16 or more. So, clearly, customers are buying into Zoho's version of suite.

Indeed, Zoho tried to jam a 158 slide presentation into an hour's time slot. Hey, it's good to see a vendor continue to invest in new functionality, and not just milk what they have delivered in the past.

Makes you wonder how much further our industry would be if other vendors would similarly reverse their 50 to 60% of revenue investment in Sales and Marketing and plough that into R&D.

 

Plex ML project - something every software vendor/systems integrator could emulate

Machine Learning is a hot topic. Every software vendor is talking about ML and AI in their products. Plex applied ML to something very different - an understanding of how their customers configure parameters in the Plex system. This conversation with Jerry Foster, CTO, describes the project.

It has always bothered me that after millions of ERP, CRM and other enterprise projects and upgrades, we cannot get predictability and savings from automation in these projects. I think every software vendor and SI can build on this exercise around other SaaS products to automate configuration and related templates. Another interesting area Jerry describes is how rethinking customizations and the UX can dramatically reduce configuration change effort.

 

When your competitors talk about machine learning, it’s typically in terms of customer data and different functional areas like accounts payable. Your project looked at implementation configurations, which is a unique approach. Can you share your perspective on Machine Learning?

Just like with any transformative technology, everyone likes to ask, "What is your machine learning strategy?" or "What's your IoT strategy?"  But you can really go down a rabbit trail because those technologies are so big. So we've always tried to ask “What problems are we going to solve for our customers?” which could even mean solving an internal challenge that ultimately helps our customers.

At Plex, we didn't want to just start building Machine Learning models and doing them willy-nilly. Instead, me and my team discussed areas where we could apply machine learning efforts in order to solve a problem while also gaining our own internal knowledge on what a ML project entails.

Can you share some background on how this specific Machine Learning initiative began?

Last year, our VP of Services told me that about 15% of incoming customer care tickets deal exclusively with the configuration setup of our system.

Because the Plex Manufacturing Cloud is a Software-as-a-Service product, everyone is using the same codebase, which means we use configuration settings (also known as customer settings) that determine how the software behaves and functions.

Settings have business flow and visual ramifications. For example, some of those settings determine actual workflow, which may be different for a food and beverage customer versus a discrete manufacturing customer.  While others are as simple as a setting that will show certain columns on the screen and, if it's turned off, they are no longer visible. 

Because the system is so full-featured, there are a large number of settings so it would be unreasonable to expect our customers or even our Customer Care team to fully understand all of the interdependencies of the settings and the unintended side effects from turning settings on and off.

Obviously, when engineers build these systems, they try to take that into account.  They will add a new setting and see how it works with all other similar settings.  But over time, you can't ascertain all of the effects of all of those settings.

Our customers are trying to figure out how to optimize their settings, so when I was pondering that stat – 15% of our tickets are related to configuration settings – I realized this would be a great place to build a machine learning model that could help us understand what the net effect is.

Our desire was to ask “How could a customer use this model to compare it to their settings?” and be able to say, "Here are the settings you should change in order to have the optimum configuration.”

Once you identified your objective, where did you start? What technologies did you use?

We put together a team of engineers and gave them this assumption that there is a data set that could be analyzed to determine whether or not a customer's settings are configured correctly.

The team came up with a framework and a set of tools, which included Azure Machine Learning Studio because, in our opinion, Microsoft has done a really good job providing the necessary tools to very quickly get up to speed on a machine learning presentation.  We also used Neo4j as our graphing database and a system called Notebook, which is a way to put together any Python scripts that you are using to augment your artificial intelligence project. 

From there, the team put these tools together and started building out their models.

Once you had the tools, what was the process like?

 There were actually three phases to the project.  The first is feature discovery, which is basically asking “What are the data sets that we need to test our assumption?”  The second phase was building the knowledge capture system, which actually gathers all of the data from the feature discovery.  Then, of course, the final phase is a building a recommendation system.

The first phase – feature discovery – was determining which data sets were the most important.  One thing I learned is that feature discovery is just as much art as it is science.  You really have to have people on your team that know the data.

We put together a set of data that we could use to test our thesis and we started to build some models around that data.  We found that once we had a set of features and a set of data points, it wasn't enough to determine what data points to use, we needed to establish just how much data to use as well.

For instance, if we were too restrictive in our data sets that we were using, the models couldn't make any inferences from that.  It was basically regurgitating what you put in.  But if we were too broad in the amount of data, data sets, and features that we were putting into the model, it just came up with all sorts of crazy inferences and dependencies that had no meaning.

We had to keep refining, which we is a big part of the machine learning process. We really had to learn to accept that there is an iterative nature to determining which data could best give us the results that we were looking for.

Once we had a set of data points and features that we felt were representative and returned some good initial results, we built a knowledge capture system using Neo4j, Excel spreadsheets, and a data factory from Microsoft. This system pulled in the data, adjusted it, and used the models that we had built in ML Framework to start analyzing the data. 

The next part was actually consolidating and analyzing the data using mainly Azure ML as well as some Power BI and Excel along with Neo4j for some of the visual representation.

Can you explain how this applied to the configuration setting problem you were trying to solve?

As I mentioned, our customers turn settings on and off in order to configure the system to meet their needs.

To make that process easier, Plex has a set of customer setting templates – 12 or 13 primary templates based on size, industry, facility count, etc. – that we use to categorize our customers, which then informs new customer implementations. When a new customer is brought into the fold, it’s more efficient to associate them with an appropriate template and copy the template settings to their instance.

For this project, we wanted the algorithm to pretend we didn't have those templates and determine on its own how it would group our customers. We fed the data into the model and it begain to cluster our customers and match them with a particular customer setting template. Using a mechanism called similarity analysis, the model then made recommendations based on the delta between the customer and neighbors using their common template as a baseline. 

Can you show me how this played out in the model?

The image below is from Neo4j and it actually shows the clustering of customers matched to a particular template.  It first came up with its own clustering, and then it attempted to match those clusters with our templates. 

This next image shows the model’s clusters.  It identified 14 broad groups completely disconnected from our templates. 

Then what we told the model to do on this next screen was to try and match its clusters with our templates, and it actually grouped our templates into those four groups that you see highlighted right there: 0, 7, 9, and 11. 

So the model reclassified your templates?

Yes, the machine learning algorithm essentially said four of the groups it had defined (Group 0, 7, 9, and 11), could match up with one or more of Plex’s existing templates. 

What else does this tell you?

All in all, what you see here is basically six of its groups matched our templates. As you can see on this next image, the model matched two of its groupings – 3 and 8 – with a few templates that we don't typically use as much. The model also has two groupings, 2 and 6, that it couldn't match with any of our templates.

One of the main findings coming out of this is that we are basically missing two templates.  In other words, there is a set of customers that, when they onboard as a new customer, we're assigning them to a template that, although it may not be bad, is probably not the optimum template that they should be getting.

This final graph illustrates a finding that was consistent across the board : the number of times our customer settings were changed dropped off a cliff in about 2015.  That was a completely unexpected finding from the study.

 

This finding validated the work that we've been doing since 2014 to reduce our overall dependence on settings, including our new user interface, which is much less settings-dependent than Classic. It gave us confidence we were on the right path.

This is also really cool because it wasn't actually in our initial hypothesis. It wasn't what my team was trying to find or determine, and it was just an unintended side effect of the work they were doing.  Of course, one of the advantages when you start doing a project like this is you start looking at your data in detail and you come across insights that you had never expected.  This is something that the initial algorithms that we were putting together showed us that wasn't even part of our initial thesis.

Those were the two main results: identifying configuration setting templates that we had missed, and showing us that the work we had done in reducing the number of settings had paid off. 

What action will you take from here?

One of the follow-throughs from this, obviously, will be to look at the model’s groupings that do not identify with a Plex template and ask what are the characteristics of the customers in those groups?  Why did we miss those?  One of the main follow-ups that we're doing right now and in upcoming months is trying to determine which customers are in those groups and what are the characteristics of those groups so we can come up with a proper template for those customers.

The second action is to follow through with the recommendation system because that's really where the productization of this is going.  If I'm an existing customer, I want to know if my settings are optimum. I want to know, based on all the customers that are like me, which of my settings are different, and why do I have those settings different than all of the other customers who are like me?

We already have the "like me" part done.  Now, we actually have to productize that so we can provide a tool, and we want to provide that tool in two different contexts.  One is an internal tool for our support team so that when they get an incoming call that says, "Hey, I've got a problem," they can run this against the settings optimization tool and say, "You know what?  You might want to check these settings.  I've noticed that they're different from you, but all other customers like you have these settings set differently."  Then, of course, the next iteration or the second context would be a self-service tool that the customer could just go onto the Plex menu and analyze their settings. 

Were you happy with the results of this project?

Overall, I was very pleased with the outcome of the project and how the team worked.  It was exciting to see.  One of the interesting things about where machine learning has come from where it was five, ten years ago is that the tools are so strong that you don't need a team of data scientists.  You need a team of smart engineers and a couple people who know the data. 

Yes, a data scientist is obviously going to help, but to get 70% or 80% of the way, in my opinion, you don't really need a data scientist.  All the things we did here, there was no data scientist involved.  That, to me, was a real encouragement.

The last takeaway I'm taking from this is, once you start to realize what's behind machine learning, you actually start to use that as a mechanism to solve problems.  In other words, it's not just this machine learning project that the labs did.  It's how we can analyze any problem moving forward.

It's going to be a long journey, but it's very exciting.  I'm really excited to see where we can go with this.

Where have all the process visionaries gone?

A recent visit to an Amazon Fulfillment Center was eye-opening. They have completely redefined the old warehouse pick, pack, ship mindset. It's now stow, pick, pack, verify, and in their lingo, SLAM and "chaotic storage". Kiva robots and miles of conveyor belts bring work to employees not the other way round. Scanners, scales, sensors monitor and facilitate the worker - what size box to assemble, cut right length of packing tape, apply shipping labels etc. Read more about my visit here. 

I was recounting the visit to a software CEO. He said "Amazon is amazing". And I responded - "it's not just Amazon". I told him about companies I have worked with and written about that have dramatically rethought their processes with all kinds of modern technology - shop floors, cashierless stores, return logistics, transportation routing, payment processing and on and on. He asked me which vendors they work with. And I named a few then stopped and said "In the end, these customers are themselves driving the vision about the new process with a wide range of technologies. Rarely is there a single vendor or product".

The realization sunk in that true process innovation is not really being bought "shrink wrapped". Companies which are taking a "systems approach" - adding sensors, robots, flexible workflow, RPA at front, center, back of any process - are getting stunningly better results.

Of course, plenty of SIs will continue to talk about "best practices" embodied in packaged software, and many customers will say they are satisfied with "good enough" processes.

Most software vendors have in the move to cloud computing hired good architects who have helped them think about multi-tenancy and platform technology. They have not invested similarly in business process experts, especially those unique to individual industries or geographies. And even with the underlying technology they have waited for consumer tech to innovate with UX, AI etc. A few of their "good enough processes" added together could translate to several basis points of margin lost every year.

Jeff Bezos, CEO of Amazon has been known to say "your margin is my opportunity". I used to think it was his way of targeting fat, dumb, high-margin competitors. After visiting his warehouse, I think he also means to target companies which are not focused enough on margins and instead waste them in process inefficiencies.